From xvid-devel-bounces@xvid.org Mon Nov 7 00:15:08 2005 Return-Path: X-Original-To: itdp@fh-biergarten.de Delivered-To: itdp@localhost Received: from edu.bnhof.de (edu.bnhof.de [213.167.167.52]) by mail.kliche.org (Postfix) with ESMTP id F4180126D82 for ; Mon, 7 Nov 2005 00:15:07 +0100 (CET) Received: from edu.bnhof.de (localhost [127.0.0.1]) by edu.bnhof.de (Postfix) with ESMTP id 25BBC18B7E; Mon, 7 Nov 2005 00:14:48 +0100 (CET) X-Original-To: xvid-devel@edu.bnhof.de Delivered-To: xvid-devel@edu.bnhof.de Received: from s6.lansco.de (s6.lansco.de [213.239.218.110]) by edu.bnhof.de (Postfix) with ESMTP id AAA1918B78 for ; Mon, 7 Nov 2005 00:14:45 +0100 (CET) Received: from orchardtech.com (orchardtech.com [207.189.111.32]) by s6.lansco.de (Postfix) with SMTP id 6DD781614C for ; Mon, 7 Nov 2005 00:14:44 +0100 (CET) Received: (qmail 80509 invoked from network); 6 Nov 2005 23:14:36 -0000 Received: from 216-115-114-112.acetekk.com (HELO ?192.168.0.100?) (216.115.114.112) by orchardtech.com with SMTP; 6 Nov 2005 23:14:36 -0000 Mime-Version: 1.0 (Apple Message framework v746.2) Content-Transfer-Encoding: 7bit Message-Id: <4B7F0307-39A8-4805-9F11-BCDA2183DE4D@jrmanes.com> Content-Type: text/plain; charset=US-ASCII; delsp=yes; format=flowed To: xvid-devel@xvid.org From: "J.R. Manes" Date: Sun, 6 Nov 2005 14:14:19 -0900 X-Mailer: Apple Mail (2.746.2) Subject: [XviD-devel] xvid_encraw default bitrate X-BeenThere: xvid-devel@xvid.org X-Mailman-Version: 2.1.4 Precedence: list Reply-To: xvid-devel@xvid.org List-Id: xvid-devel.xvid.org List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Sender: xvid-devel-bounces@xvid.org Errors-To: xvid-devel-bounces@xvid.org Howdy- When you send a series of PPM files through xvid_encraw, what is the default bitrate if you don't specify it on the command line? Thanks, J.R. _______________________________________________ XviD-devel mailing list XviD-devel@xvid.org http://list.xvid.org/mailman/listinfo/xvid-devel From xvid-devel-bounces@xvid.org Wed Nov 9 15:39:07 2005 Return-Path: X-Original-To: itdp@fh-biergarten.de Delivered-To: itdp@localhost Received: from edu.bnhof.de (edu.bnhof.de [213.167.167.52]) by mail.kliche.org (Postfix) with ESMTP id D14E6126D82 for ; Wed, 9 Nov 2005 15:39:07 +0100 (CET) Received: from edu.bnhof.de (localhost [127.0.0.1]) by edu.bnhof.de (Postfix) with ESMTP id 1B6EC18C2E; Wed, 9 Nov 2005 15:38:55 +0100 (CET) X-Original-To: xvid-devel@edu.bnhof.de Delivered-To: xvid-devel@edu.bnhof.de Received: from s6.lansco.de (s6.lansco.de [213.239.218.110]) by edu.bnhof.de (Postfix) with ESMTP id 673AE18D2D for ; Wed, 9 Nov 2005 15:36:36 +0100 (CET) Received: from mail.ln9.net (rrcs-24-73-193-138.se.biz.rr.com [24.73.193.138]) by s6.lansco.de (Postfix) with ESMTP id BE1FA16195 for ; Tue, 8 Nov 2005 19:00:03 +0100 (CET) Received: from [192.168.100.164] (yarr.ad.ln9.net [192.168.100.164]) by mail.ln9.net (Postfix) with ESMTP id 53B674B014B for ; Tue, 8 Nov 2005 12:59:58 -0500 (EST) Message-ID: <4370E79E.8070606@leoninedev.com> Date: Tue, 08 Nov 2005 12:59:58 -0500 From: Bryan Mayland Organization: Leonine Development, Inc. User-Agent: Thunderbird 1.5 (Windows/20051025) MIME-Version: 1.0 To: xvid-devel@xvid.org Subject: Re: [XviD-devel] xvid_encraw default bitrate References: <4B7F0307-39A8-4805-9F11-BCDA2183DE4D@jrmanes.com> In-Reply-To: <4B7F0307-39A8-4805-9F11-BCDA2183DE4D@jrmanes.com> Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 7bit X-BeenThere: xvid-devel@xvid.org X-Mailman-Version: 2.1.4 Precedence: list Reply-To: xvid-devel@xvid.org List-Id: xvid-devel.xvid.org List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Sender: xvid-devel-bounces@xvid.org Errors-To: xvid-devel-bounces@xvid.org J.R. Manes wrote: > When you send a series of PPM files through xvid_encraw, what is the > default bitrate if you don't specify it on the command line? The default is 0 as far as I know. You should specify either a quality or bitrate, as well as -single or -pass1/pass2. _______________________________________________ XviD-devel mailing list XviD-devel@xvid.org http://list.xvid.org/mailman/listinfo/xvid-devel From xvid-devel-bounces@xvid.org Mon Nov 21 06:42:24 2005 Return-Path: X-Original-To: itdp@fh-biergarten.de Delivered-To: itdp@localhost Received: from edu.bnhof.de (edu.bnhof.de [213.167.167.52]) by mail.kliche.org (Postfix) with ESMTP id 0271A126D85 for ; Mon, 21 Nov 2005 06:42:24 +0100 (CET) Received: from edu.bnhof.de (localhost [127.0.0.1]) by edu.bnhof.de (Postfix) with ESMTP id A070918B7A; Mon, 21 Nov 2005 06:42:14 +0100 (CET) X-Original-To: xvid-devel@edu.bnhof.de Delivered-To: xvid-devel@edu.bnhof.de Received: from s6.lansco.de (s6.lansco.de [213.239.218.110]) by edu.bnhof.de (Postfix) with ESMTP id 76E9F18AA0 for ; Mon, 21 Nov 2005 06:42:11 +0100 (CET) Received: from mail-ihug.icp-qv1-irony3.iinet.net.au (ihug-mail.icp-qv1-irony3.iinet.net.au [203.59.1.197]) by s6.lansco.de (Postfix) with ESMTP id B6D0416144 for ; Mon, 21 Nov 2005 06:42:09 +0100 (CET) Received: from 203-173-56-106.dyn.iinet.net.au (HELO [10.1.1.3]) ([203.173.56.106]) by mail-ihug.icp-qv1-irony3.iinet.net.au with ESMTP; 21 Nov 2005 13:42:07 +0800 X-BrightmailFiltered: true X-Brightmail-Tracker: AAAAAQAAA+k= Message-ID: <43815E50.8@syskin.cjb.net> Date: Mon, 21 Nov 2005 16:12:40 +1030 From: Radek Czyz User-Agent: Mozilla Thunderbird 1.0.2 (Windows/20050317) X-Accept-Language: en-us, en MIME-Version: 1.0 To: xvid-devel@xvid.org Content-Type: text/plain; charset=UTF-8; format=flowed Content-Transfer-Encoding: 7bit Subject: [XviD-devel] hey I'm back :) X-BeenThere: xvid-devel@xvid.org X-Mailman-Version: 2.1.4 Precedence: list Reply-To: xvid-devel@xvid.org List-Id: xvid-devel.xvid.org List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Sender: xvid-devel-bounces@xvid.org Errors-To: xvid-devel-bounces@xvid.org Hello everyone, I have summer vacations now, FINALLY :) and this means I'm back to xvid poking, with some fresh ideas and motivations. First thing I'd like to do is release 1.1 final. How about we do this without waiting? I was thinking that we commit GomGom's Sparse patch (the pointer casting for fdct and idct is fugly, everything else is fine), bump version strings and have it behind us. Future work I want to do, in no particular order: - xvidenc - a commandline executable like x264 has. It turns out there's a demand for such executable, we already have a lot of code, and everything else can be taken from x264 - pre-pass ME. During my recent work, I discovered that motion finding over larger distances is often flawed and inefficient, and can be largly improved by pre-pass. I'm going to try and see what happens. - Kopernikus' variable-lambda code. I believe lambda (VHQ and trellis) is the best way to modulate quality for all HVS applications. I want to give it a chance ;) - multithreaded code. I have several ideas how various parts of XviD can be multithreaded, and I intend to implement them, if only to learn how. - everything else that comes to my mind ;) I'll start soon :) Radek _______________________________________________ XviD-devel mailing list XviD-devel@xvid.org http://list.xvid.org/mailman/listinfo/xvid-devel From xvid-devel-bounces@xvid.org Mon Nov 21 09:18:21 2005 Return-Path: X-Original-To: itdp@fh-biergarten.de Delivered-To: itdp@localhost Received: from edu.bnhof.de (edu.bnhof.de [213.167.167.52]) by mail.kliche.org (Postfix) with ESMTP id A23CF126D82 for ; Mon, 21 Nov 2005 09:18:21 +0100 (CET) Received: from edu.bnhof.de (localhost [127.0.0.1]) by edu.bnhof.de (Postfix) with ESMTP id A1CE818BBB; Mon, 21 Nov 2005 09:18:10 +0100 (CET) X-Original-To: xvid-devel@edu.bnhof.de Delivered-To: xvid-devel@edu.bnhof.de Received: from s6.lansco.de (s6.lansco.de [213.239.218.110]) by edu.bnhof.de (Postfix) with ESMTP id C457018BAE for ; Mon, 21 Nov 2005 09:18:07 +0100 (CET) Received: from xproxy.gmail.com (xproxy.gmail.com [66.249.82.206]) by s6.lansco.de (Postfix) with ESMTP id 1ACD816151 for ; Mon, 21 Nov 2005 09:18:05 +0100 (CET) Received: by xproxy.gmail.com with SMTP id h27so706536wxd for ; Mon, 21 Nov 2005 00:18:05 -0800 (PST) DomainKey-Signature: a=rsa-sha1; q=dns; c=nofws; s=beta; d=gmail.com; h=received:message-id:date:from:to:subject:in-reply-to:mime-version:content-type:content-transfer-encoding:content-disposition:references; b=f91vUcZsx7Hgvld+niZf5riQwX/LrOldlspE+7L8xQS0qbI0KNsjeHRHI0AUWPY2OuI69qKSZRDh0zz5eNVfW6jEldSxenSAGbP0vVMwVmlOinR63zmfBpLPoxQFRG1od2y33FyiC/NIxn5gSIhMGZqiXsIBgHUdWbFB6+0ZQrU= Received: by 10.65.15.4 with SMTP id s4mr2575215qbi; Mon, 21 Nov 2005 00:18:05 -0800 (PST) Received: by 10.64.181.13 with HTTP; Mon, 21 Nov 2005 00:18:04 -0800 (PST) Message-ID: Date: Mon, 21 Nov 2005 09:18:04 +0100 From: Hannes Sachsenhofer To: xvid-devel@xvid.org Subject: Re: [XviD-devel] hey I'm back :) In-Reply-To: <43815E50.8@syskin.cjb.net> MIME-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: quoted-printable Content-Disposition: inline References: <43815E50.8@syskin.cjb.net> X-BeenThere: xvid-devel@xvid.org X-Mailman-Version: 2.1.4 Precedence: list Reply-To: xvid-devel@xvid.org List-Id: xvid-devel.xvid.org List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Sender: xvid-devel-bounces@xvid.org Errors-To: xvid-devel-bounces@xvid.org Hi Radek, glad to hear that. I was afraid that Xvid development came to end due to the lack of updates recently ... Keep up the good work. On 11/21/05, Radek Czyz wrote: > Hello everyone, > > I have summer vacations now, FINALLY :) and this means I'm back to xvid > poking, with some fresh ideas and motivations. > > First thing I'd like to do is release 1.1 final. How about we do this > without waiting? I was thinking that we commit GomGom's Sparse patch > (the pointer casting for fdct and idct is fugly, everything else is > fine), bump version strings and have it behind us. > > Future work I want to do, in no particular order: > > - xvidenc - a commandline executable like x264 has. It turns out there's > a demand for such executable, we already have a lot of code, and > everything else can be taken from x264 > > - pre-pass ME. During my recent work, I discovered that motion finding > over larger distances is often flawed and inefficient, and can be largly > improved by pre-pass. I'm going to try and see what happens. > > - Kopernikus' variable-lambda code. I believe lambda (VHQ and trellis) > is the best way to modulate quality for all HVS applications. I want to > give it a chance ;) > > - multithreaded code. I have several ideas how various parts of XviD can > be multithreaded, and I intend to implement them, if only to learn how. > > - everything else that comes to my mind ;) > > I'll start soon :) > > Radek > _______________________________________________ > XviD-devel mailing list > XviD-devel@xvid.org > http://list.xvid.org/mailman/listinfo/xvid-devel > -- regards, Hannes Sachsenhofer < http://www.sachsenhofer.com > _______________________________________________ XviD-devel mailing list XviD-devel@xvid.org http://list.xvid.org/mailman/listinfo/xvid-devel From xvid-devel-bounces@xvid.org Mon Nov 21 10:03:41 2005 Return-Path: X-Original-To: itdp@fh-biergarten.de Delivered-To: itdp@localhost Received: from edu.bnhof.de (edu.bnhof.de [213.167.167.52]) by mail.kliche.org (Postfix) with ESMTP id DBA90126D89 for ; Mon, 21 Nov 2005 10:03:41 +0100 (CET) Received: from edu.bnhof.de (localhost [127.0.0.1]) by edu.bnhof.de (Postfix) with ESMTP id 944D418BBC; Mon, 21 Nov 2005 10:03:33 +0100 (CET) X-Original-To: xvid-devel@edu.bnhof.de Delivered-To: xvid-devel@edu.bnhof.de Received: from s6.lansco.de (s6.lansco.de [213.239.218.110]) by edu.bnhof.de (Postfix) with ESMTP id 3398118BA7 for ; Mon, 21 Nov 2005 10:03:31 +0100 (CET) Received: from guanin.uni-konstanz.de (guanin.uni-konstanz.de [134.34.240.60]) by s6.lansco.de (Postfix) with ESMTP id 3277615FB4 for ; Mon, 21 Nov 2005 10:03:30 +0100 (CET) Received: from unitis.rz.uni-konstanz.de (unitis.rz.uni-konstanz.de [134.34.240.51]) by guanin.uni-konstanz.de (Postfix) with ESMTP id 081561856 for ; Mon, 21 Nov 2005 10:03:25 +0100 (CET) Received: from localhost (localhost [127.0.0.1]) by localhost.unitis.rz.uni-konstanz.de (Postfix) with ESMTP id E9C9C3C007 for ; Mon, 21 Nov 2005 10:03:24 +0100 (CET) Received: from unitis.rz.uni-konstanz.de ([127.0.0.1]) by localhost (unitis.rz.uni-konstanz.de [127.0.0.1]) (amavisd-new, port 10024) with LMTP id 30754-01-68 for ; Mon, 21 Nov 2005 10:03:23 +0100 (CET) Received: from localhost (taxis.rz.uni-konstanz.de [134.34.3.51]) by unitis.rz.uni-konstanz.de (Postfix) with ESMTP id B66DD3C006 for ; Mon, 21 Nov 2005 10:03:23 +0100 (CET) Received: from merry.physik.uni-konstanz.de (merry.physik.uni-konstanz.de [134.34.140.41]) by webmail.uni-konstanz.de (IMP) with HTTP for ; Mon, 21 Nov 2005 10:03:23 +0100 Message-ID: <1132563803.43818d5b79dd2@webmail.uni-konstanz.de> Date: Mon, 21 Nov 2005 10:03:23 +0100 From: Johannes Reinhardt To: xvid-devel@xvid.org Subject: Re: [XviD-devel] hey I'm back :) References: <43815E50.8@syskin.cjb.net> In-Reply-To: <43815E50.8@syskin.cjb.net> MIME-Version: 1.0 Content-Type: text/plain; charset=utf-8 Content-Transfer-Encoding: 8bit User-Agent: Internet Messaging Program (IMP) 3.2.2 X-Originating-IP: 134.34.140.41 X-Virus-Scanned: by amavisd-new at Mailservice RZ Uni-Konstanz X-BeenThere: xvid-devel@xvid.org X-Mailman-Version: 2.1.4 Precedence: list Reply-To: xvid-devel@xvid.org List-Id: xvid-devel.xvid.org List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Sender: xvid-devel-bounces@xvid.org Errors-To: xvid-devel-bounces@xvid.org Hi, I will try do a lambda based hvs plugin as a replacement for the AQ based lumimasking, but I don`t know how much spare time I have the next days. And there are other effects that could be exploited. I recently read some papers about foveating and automatic fixation point detection. Do you think this could be useful? Johannes Quoting Radek Czyz : > Hello everyone, > > I have summer vacations now, FINALLY :) and this means I'm back to > xvid > poking, with some fresh ideas and motivations. > > First thing I'd like to do is release 1.1 final. How about we do > this > without waiting? I was thinking that we commit GomGom's Sparse patch > (the pointer casting for fdct and idct is fugly, everything else is > fine), bump version strings and have it behind us. > > Future work I want to do, in no particular order: > > - xvidenc - a commandline executable like x264 has. It turns out > there's > a demand for such executable, we already have a lot of code, and > everything else can be taken from x264 > > - pre-pass ME. During my recent work, I discovered that motion > finding > over larger distances is often flawed and inefficient, and can be > largly > improved by pre-pass. I'm going to try and see what happens. > > - Kopernikus' variable-lambda code. I believe lambda (VHQ and > trellis) > is the best way to modulate quality for all HVS applications. I want > to > give it a chance ;) > > - multithreaded code. I have several ideas how various parts of XviD > can > be multithreaded, and I intend to implement them, if only to learn > how. > > - everything else that comes to my mind ;) > > I'll start soon :) > > Radek > _______________________________________________ > XviD-devel mailing list > XviD-devel@xvid.org > http://list.xvid.org/mailman/listinfo/xvid-devel > _______________________________________________ XviD-devel mailing list XviD-devel@xvid.org http://list.xvid.org/mailman/listinfo/xvid-devel From xvid-devel-bounces@xvid.org Mon Nov 21 11:27:05 2005 Return-Path: X-Original-To: itdp@fh-biergarten.de Delivered-To: itdp@localhost Received: from edu.bnhof.de (edu.bnhof.de [213.167.167.52]) by mail.kliche.org (Postfix) with ESMTP id 0AC9C126D85 for ; Mon, 21 Nov 2005 11:27:05 +0100 (CET) Received: from edu.bnhof.de (localhost [127.0.0.1]) by edu.bnhof.de (Postfix) with ESMTP id 90BC018BC5; Mon, 21 Nov 2005 11:26:56 +0100 (CET) X-Original-To: xvid-devel@edu.bnhof.de Delivered-To: xvid-devel@edu.bnhof.de Received: from s6.lansco.de (s6.lansco.de [213.239.218.110]) by edu.bnhof.de (Postfix) with ESMTP id 927A218BBF for ; Mon, 21 Nov 2005 11:26:53 +0100 (CET) Received: from s5.lansco.de (s5.lansco.de [212.63.130.170]) by s6.lansco.de (Postfix) with ESMTP id A710C15F0A for ; Mon, 21 Nov 2005 11:26:52 +0100 (CET) Received: from localhost (s5.lansco.de [212.63.130.170]) by s5.lansco.de (Postfix) with ESMTP id 8C56030884 for ; Mon, 21 Nov 2005 11:26:32 +0100 (CET) Received: from p54B5F7F0.dip.t-dialin.net (p54B5F7F0.dip.t-dialin.net [84.181.247.240]) by www.lansco.de (IMP) with HTTP for ; Mon, 21 Nov 2005 11:26:27 +0100 Message-ID: <1132568787.4381a0d36ead6@www.lansco.de> Date: Mon, 21 Nov 2005 11:26:27 +0100 From: Michael Militzer To: xvid-devel@xvid.org Subject: Re: [XviD-devel] hey I'm back :) References: <43815E50.8@syskin.cjb.net> In-Reply-To: <43815E50.8@syskin.cjb.net> MIME-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 8bit User-Agent: Internet Messaging Program (IMP) 3.2.3 X-Originating-IP: 84.181.247.240 X-BeenThere: xvid-devel@xvid.org X-Mailman-Version: 2.1.4 Precedence: list Reply-To: xvid-devel@xvid.org List-Id: xvid-devel.xvid.org List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Sender: xvid-devel-bounces@xvid.org Errors-To: xvid-devel-bounces@xvid.org Hi Radek, welcome back! And good to hear you have ideas for further improvements ;) Note that there is e.g. avs input support in xvidenc now already but we could of course additionally add avi/mp4/mkv output support and should also care about proper implementation of profiles. Also, Skal is working on multithreaded code and has made a lot (good and bad) experiences with this recently. So you should check how your ideas here overlap (or not) just to make sure you don't do things twice... Regards, Michael Quoting Radek Czyz : > Hello everyone, > > I have summer vacations now, FINALLY :) and this means I'm back to xvid > poking, with some fresh ideas and motivations. > > First thing I'd like to do is release 1.1 final. How about we do this > without waiting? I was thinking that we commit GomGom's Sparse patch > (the pointer casting for fdct and idct is fugly, everything else is > fine), bump version strings and have it behind us. > > Future work I want to do, in no particular order: > > - xvidenc - a commandline executable like x264 has. It turns out there's > a demand for such executable, we already have a lot of code, and > everything else can be taken from x264 > > - pre-pass ME. During my recent work, I discovered that motion finding > over larger distances is often flawed and inefficient, and can be largly > improved by pre-pass. I'm going to try and see what happens. > > - Kopernikus' variable-lambda code. I believe lambda (VHQ and trellis) > is the best way to modulate quality for all HVS applications. I want to > give it a chance ;) > > - multithreaded code. I have several ideas how various parts of XviD can > be multithreaded, and I intend to implement them, if only to learn how. > > - everything else that comes to my mind ;) > > I'll start soon :) > > Radek > _______________________________________________ > XviD-devel mailing list > XviD-devel@xvid.org > http://list.xvid.org/mailman/listinfo/xvid-devel > _______________________________________________ XviD-devel mailing list XviD-devel@xvid.org http://list.xvid.org/mailman/listinfo/xvid-devel From xvid-devel-bounces@xvid.org Mon Nov 21 11:38:27 2005 Return-Path: X-Original-To: itdp@fh-biergarten.de Delivered-To: itdp@localhost Received: from edu.bnhof.de (edu.bnhof.de [213.167.167.52]) by mail.kliche.org (Postfix) with ESMTP id 71FA2126D85 for ; Mon, 21 Nov 2005 11:38:27 +0100 (CET) Received: from edu.bnhof.de (localhost [127.0.0.1]) by edu.bnhof.de (Postfix) with ESMTP id B6AE418BCC; Mon, 21 Nov 2005 11:38:19 +0100 (CET) X-Original-To: xvid-devel@edu.bnhof.de Delivered-To: xvid-devel@edu.bnhof.de Received: from s6.lansco.de (s6.lansco.de [213.239.218.110]) by edu.bnhof.de (Postfix) with ESMTP id 1F33F13D9D for ; Mon, 21 Nov 2005 11:38:15 +0100 (CET) Received: from mail-ihug.icp-qv1-irony1.iinet.net.au (ihug-mail.icp-qv1-irony1.iinet.net.au [203.59.1.195]) by s6.lansco.de (Postfix) with ESMTP id BC87A1617C for ; Mon, 21 Nov 2005 11:38:10 +0100 (CET) Received: from 203-173-56-106.dyn.iinet.net.au (HELO [10.1.1.3]) ([203.173.56.106]) by mail-ihug.icp-qv1-irony1.iinet.net.au with ESMTP; 21 Nov 2005 18:38:07 +0800 X-BrightmailFiltered: true X-Brightmail-Tracker: AAAAAA== Message-ID: <4381A3B4.70708@syskin.cjb.net> Date: Mon, 21 Nov 2005 21:08:44 +1030 From: Radek Czyz User-Agent: Mozilla Thunderbird 1.0.2 (Windows/20050317) X-Accept-Language: en-us, en MIME-Version: 1.0 To: xvid-devel@xvid.org Subject: Re: [XviD-devel] hey I'm back :) References: <43815E50.8@syskin.cjb.net> <1132568787.4381a0d36ead6@www.lansco.de> In-Reply-To: <1132568787.4381a0d36ead6@www.lansco.de> Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 7bit X-BeenThere: xvid-devel@xvid.org X-Mailman-Version: 2.1.4 Precedence: list Reply-To: xvid-devel@xvid.org List-Id: xvid-devel.xvid.org List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Sender: xvid-devel-bounces@xvid.org Errors-To: xvid-devel-bounces@xvid.org Hey, Michael Militzer wrote: > Also, Skal is working on multithreaded code and has made a lot (good and > bad) experiences with this recently. So you should check how your ideas > here overlap (or not) just to make sure you don't do things twice... Bah, there goes my fun/learning ;) I mostly want to know *how* to write such code, so even if things repeat, it's not a huge loss. Skal: what have you done? ^_^ Radek _______________________________________________ XviD-devel mailing list XviD-devel@xvid.org http://list.xvid.org/mailman/listinfo/xvid-devel From xvid-devel-bounces@xvid.org Mon Nov 21 14:34:12 2005 Return-Path: X-Original-To: itdp@fh-biergarten.de Delivered-To: itdp@localhost Received: from edu.bnhof.de (edu.bnhof.de [213.167.167.52]) by mail.kliche.org (Postfix) with ESMTP id 69FD0126D82 for ; Mon, 21 Nov 2005 14:34:12 +0100 (CET) Received: from edu.bnhof.de (localhost [127.0.0.1]) by edu.bnhof.de (Postfix) with ESMTP id 35A0C18BAC; Mon, 21 Nov 2005 14:34:01 +0100 (CET) X-Original-To: xvid-devel@edu.bnhof.de Delivered-To: xvid-devel@edu.bnhof.de Received: from s6.lansco.de (s6.lansco.de [213.239.218.110]) by edu.bnhof.de (Postfix) with ESMTP id 9004918BA2 for ; Mon, 21 Nov 2005 14:33:58 +0100 (CET) Received: from ftp.ilog.fr (ftp.ilog.fr [81.80.162.195]) by s6.lansco.de (Postfix) with ESMTP id 8720C16189 for ; Mon, 21 Nov 2005 14:33:57 +0100 (CET) Received: from laposte.ilog.fr (cerbere-qfe0 [81.80.162.193]) by ftp.ilog.fr (8.13.1/8.13.1) with ESMTP id jALDXrnv032533 for ; Mon, 21 Nov 2005 14:33:57 +0100 Received: from marbore.ilog.biz (marbore1.ilog.fr [172.17.2.61]) by laposte.ilog.fr (8.13.1/8.13.1) with ESMTP id jALDXlYL026601 for ; Mon, 21 Nov 2005 14:33:47 +0100 (MET) Received: from parmbx01.ilog.biz ([172.17.2.64]) by marbore.ilog.biz with Microsoft SMTPSVC(6.0.3790.1830); Mon, 21 Nov 2005 14:33:45 +0100 Received: from 6ttgg1j ([172.17.4.249]) by parmbx01.ilog.biz with Microsoft SMTPSVC(6.0.3790.1830); Mon, 21 Nov 2005 14:33:45 +0100 Subject: Re: [XviD-devel] hey I'm back :) From: Skal To: xvid-devel@xvid.org In-Reply-To: <4381A3B4.70708@syskin.cjb.net> References: <43815E50.8@syskin.cjb.net> <1132568787.4381a0d36ead6@www.lansco.de> <4381A3B4.70708@syskin.cjb.net> Content-Type: multipart/mixed; boundary="=-OEzR2cv4KR/ht+vQyepQ" Message-Id: <1132580023.3309.14.camel@localhost.localdomain> Mime-Version: 1.0 X-Mailer: Ximian Evolution 1.4.5 (1.4.5-14) Date: Mon, 21 Nov 2005 14:33:43 +0100 X-OriginalArrivalTime: 21 Nov 2005 13:33:45.0632 (UTC) FILETIME=[3251B200:01C5EEA0] X-Content-Filtered-By: Mailman/MimeDel 2.1.4 X-BeenThere: xvid-devel@xvid.org X-Mailman-Version: 2.1.4 Precedence: list Reply-To: xvid-devel@xvid.org List-Id: xvid-devel.xvid.org List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Sender: xvid-devel-bounces@xvid.org Errors-To: xvid-devel-bounces@xvid.org --=-OEzR2cv4KR/ht+vQyepQ Content-Type: text/plain Content-Transfer-Encoding: 7bit Hi Radek! :) On Mon, 2005-11-21 at 11:38, Radek Czyz wrote: > Hey, > > Michael Militzer wrote: > > Also, Skal is working on multithreaded code and has made a lot (good and > > bad) experiences with this recently. So you should check how your ideas > > here overlap (or not) just to make sure you don't do things twice... > > Bah, there goes my fun/learning ;) > > I mostly want to know *how* to write such code, so even if things > repeat, it's not a huge loss. > > Skal: what have you done? ^_^ Ok, you take me by surprise, but anyway i'll just post here the helpers for MT i was preparing for XviD. It's mainly 2 files (slave.h and slave.c) that belongs to the /utils/ directory. They are not in a 'committable' state, but i think you can play with them. The API for using a 'slave' thread is rather simple: ThreadWorker slave; xvThreadInit(&slave); // creates a sleeping thread, hereby called 'slave' // prepare a workload to give the to the slave when times come ThreadWorkerData work; work._Call = myParallelFunc; work._Int1 = some param to be passed ... // schedule the parallel task xvThreadSetTodo(&slave, &work); // do something else while the slave is working... ... // sync everybody xvThreadSync(&slave); // finish xvThreadEnd(&slave); Now the func to be executed in parallel look something like: void *myParallelFunc(ThreadWorkerData *data) { int myParam = data->_Int1; // etc... /// do something return 0; // success. Return something !=0 to abort MT. } Well, i think it's close to useful, notwithstanding the bugs of course! Haf phun! bye, Skal (for the experiments i made, it'll take another (somewhat longer;) mail) --=-OEzR2cv4KR/ht+vQyepQ Content-Type: text/plain; charset="us-ascii" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit Content-Disposition: inline _______________________________________________ XviD-devel mailing list XviD-devel@xvid.org http://list.xvid.org/mailman/listinfo/xvid-devel --=-OEzR2cv4KR/ht+vQyepQ-- From xvid-devel-bounces@xvid.org Mon Nov 21 14:42:13 2005 Return-Path: X-Original-To: itdp@fh-biergarten.de Delivered-To: itdp@localhost Received: from edu.bnhof.de (edu.bnhof.de [213.167.167.52]) by mail.kliche.org (Postfix) with ESMTP id B4E74126D82 for ; Mon, 21 Nov 2005 14:42:12 +0100 (CET) Received: from edu.bnhof.de (localhost [127.0.0.1]) by edu.bnhof.de (Postfix) with ESMTP id 742DB18BDB; Mon, 21 Nov 2005 14:42:05 +0100 (CET) X-Original-To: xvid-devel@edu.bnhof.de Delivered-To: xvid-devel@edu.bnhof.de Received: from s6.lansco.de (s6.lansco.de [213.239.218.110]) by edu.bnhof.de (Postfix) with ESMTP id 8E48D18BD1 for ; Mon, 21 Nov 2005 14:41:57 +0100 (CET) Received: from ihug-mail.icp-qv1-irony2.iinet.net.au (ihug-mail.icp-qv1-irony2.iinet.net.au [203.59.1.196]) by s6.lansco.de (Postfix) with ESMTP id 95CAF16189 for ; Mon, 21 Nov 2005 14:41:55 +0100 (CET) Received: from 203-173-56-106.dyn.iinet.net.au (HELO [10.1.1.3]) ([203.173.56.106]) by ihug-mail.icp-qv1-irony2.iinet.net.au with ESMTP; 21 Nov 2005 21:41:52 +0800 Message-ID: <4381CEC6.2070106@syskin.cjb.net> Date: Tue, 22 Nov 2005 00:12:30 +1030 From: Radek Czyz User-Agent: Mozilla Thunderbird 1.0.2 (Windows/20050317) X-Accept-Language: en-us, en MIME-Version: 1.0 To: xvid-devel@xvid.org Subject: Re: [XviD-devel] hey I'm back :) References: <43815E50.8@syskin.cjb.net> <1132568787.4381a0d36ead6@www.lansco.de> <4381A3B4.70708@syskin.cjb.net> <1132580023.3309.14.camel@localhost.localdomain> In-Reply-To: <1132580023.3309.14.camel@localhost.localdomain> Content-Type: text/plain; charset=UTF-8; format=flowed Content-Transfer-Encoding: 7bit X-BeenThere: xvid-devel@xvid.org X-Mailman-Version: 2.1.4 Precedence: list Reply-To: xvid-devel@xvid.org List-Id: xvid-devel.xvid.org List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Sender: xvid-devel-bounces@xvid.org Errors-To: xvid-devel-bounces@xvid.org Skal wrote: > Ok, you take me by surprise, but anyway i'll just > post here the helpers for MT i was preparing for > XviD. It's mainly 2 files (slave.h and slave.c) > that belongs to the /utils/ directory. They are > not in a 'committable' state, but i think you > can play with them. The API for using a 'slave' > thread is rather simple: The code was eaten wasn't it? Do I understand correctly that you only create second thread and it doesn't scale up? Thanks, Radek _______________________________________________ XviD-devel mailing list XviD-devel@xvid.org http://list.xvid.org/mailman/listinfo/xvid-devel From xvid-devel-bounces@xvid.org Mon Nov 21 14:48:36 2005 Return-Path: X-Original-To: itdp@fh-biergarten.de Delivered-To: itdp@localhost Received: from edu.bnhof.de (edu.bnhof.de [213.167.167.52]) by mail.kliche.org (Postfix) with ESMTP id 1A5AA126D82 for ; Mon, 21 Nov 2005 14:48:36 +0100 (CET) Received: from edu.bnhof.de (localhost [127.0.0.1]) by edu.bnhof.de (Postfix) with ESMTP id BCA1318BEE; Mon, 21 Nov 2005 14:48:28 +0100 (CET) X-Original-To: xvid-devel@edu.bnhof.de Delivered-To: xvid-devel@edu.bnhof.de Received: from s6.lansco.de (s6.lansco.de [213.239.218.110]) by edu.bnhof.de (Postfix) with ESMTP id B4F1C18BEB for ; Mon, 21 Nov 2005 14:48:26 +0100 (CET) Received: from mail.gmx.net (mail.gmx.net [213.165.64.20]) by s6.lansco.de (Postfix) with SMTP id A1B0A16189 for ; Mon, 21 Nov 2005 14:48:20 +0100 (CET) Received: (qmail invoked by alias); 21 Nov 2005 13:48:19 -0000 Received: from p54A006C6.dip0.t-ipconnect.de (EHLO [84.160.6.198]) [84.160.6.198] by mail.gmx.net (mp007) with SMTP; 21 Nov 2005 14:48:19 +0100 X-Authenticated: #8546565 Message-ID: <4381D0BD.2090902@gmx.at> Date: Mon, 21 Nov 2005 14:50:53 +0100 From: iibot User-Agent: Thunderbird 1.5 (Windows/20051025) MIME-Version: 1.0 To: xvid-devel@xvid.org Subject: Re: [XviD-devel] hey I'm back :) References: <43815E50.8@syskin.cjb.net> <1132568787.4381a0d36ead6@www.lansco.de> <4381A3B4.70708@syskin.cjb.net> In-Reply-To: <4381A3B4.70708@syskin.cjb.net> Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 7bit X-Y-GMX-Trusted: 0 X-BeenThere: xvid-devel@xvid.org X-Mailman-Version: 2.1.4 Precedence: list Reply-To: iibot@gmx.at, xvid-devel@xvid.org List-Id: xvid-devel.xvid.org List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Sender: xvid-devel-bounces@xvid.org Errors-To: xvid-devel-bounces@xvid.org Radek Czyz schrieb: > Michael Militzer wrote: >> Also, Skal is working on multithreaded code and has made a lot (good and >> bad) experiences with this recently. So you should check how your ideas >> here overlap (or not) just to make sure you don't do things twice... > > Bah, there goes my fun/learning ;) Good to have you back. > I mostly want to know *how* to write such code, so even if things > repeat, it's not a huge loss. Nice to see someone volunteer to code that striped MB encoding I suggested a while back. :D bis besser, Tobias _______________________________________________ XviD-devel mailing list XviD-devel@xvid.org http://list.xvid.org/mailman/listinfo/xvid-devel From xvid-devel-bounces@xvid.org Mon Nov 21 15:09:05 2005 Return-Path: X-Original-To: itdp@fh-biergarten.de Delivered-To: itdp@localhost Received: from edu.bnhof.de (edu.bnhof.de [213.167.167.52]) by mail.kliche.org (Postfix) with ESMTP id 8EAFF126D85 for ; Mon, 21 Nov 2005 15:09:04 +0100 (CET) Received: from edu.bnhof.de (localhost [127.0.0.1]) by edu.bnhof.de (Postfix) with ESMTP id C2A0B18B7C; Mon, 21 Nov 2005 15:08:56 +0100 (CET) X-Original-To: xvid-devel@edu.bnhof.de Delivered-To: xvid-devel@edu.bnhof.de Received: from s6.lansco.de (s6.lansco.de [213.239.218.110]) by edu.bnhof.de (Postfix) with ESMTP id B761F18B6E for ; Mon, 21 Nov 2005 15:08:53 +0100 (CET) Received: from mail.ln9.net (rrcs-24-73-193-138.se.biz.rr.com [24.73.193.138]) by s6.lansco.de (Postfix) with ESMTP id 20E691611D for ; Mon, 21 Nov 2005 15:08:47 +0100 (CET) Received: from [192.168.100.164] (yarr.ad.ln9.net [192.168.100.164]) by mail.ln9.net (Postfix) with ESMTP id D1CF14B055C for ; Mon, 21 Nov 2005 09:08:40 -0500 (EST) Message-ID: <4381D4E8.6070003@leoninedev.com> Date: Mon, 21 Nov 2005 09:08:40 -0500 From: Bryan Mayland Organization: Leonine Development, Inc. User-Agent: Thunderbird 1.5 (Windows/20051025) MIME-Version: 1.0 To: xvid-devel@xvid.org Subject: Re: [XviD-devel] hey I'm back :) References: <43815E50.8@syskin.cjb.net> In-Reply-To: <43815E50.8@syskin.cjb.net> Content-Type: text/plain; charset=UTF-8; format=flowed Content-Transfer-Encoding: 7bit X-BeenThere: xvid-devel@xvid.org X-Mailman-Version: 2.1.4 Precedence: list Reply-To: xvid-devel@xvid.org List-Id: xvid-devel.xvid.org List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Sender: xvid-devel-bounces@xvid.org Errors-To: xvid-devel-bounces@xvid.org Welcome back! Radek Czyz wrote: > - multithreaded code. I have several ideas how various parts of XviD can > be multithreaded, and I intend to implement them, if only to learn how. I've also got a multithreaded MotionEstimation() function for P-VOPs that is mostly complete and waiting for Skal's thread abstraction. The algorithm can scale to an arbitrary number of threads is the API will permit. _______________________________________________ XviD-devel mailing list XviD-devel@xvid.org http://list.xvid.org/mailman/listinfo/xvid-devel From xvid-devel-bounces@xvid.org Mon Nov 21 15:14:32 2005 Return-Path: X-Original-To: itdp@fh-biergarten.de Delivered-To: itdp@localhost Received: from edu.bnhof.de (edu.bnhof.de [213.167.167.52]) by mail.kliche.org (Postfix) with ESMTP id 66EA7126D85 for ; Mon, 21 Nov 2005 15:14:32 +0100 (CET) Received: from edu.bnhof.de (localhost [127.0.0.1]) by edu.bnhof.de (Postfix) with ESMTP id E837018BBA; Mon, 21 Nov 2005 15:14:24 +0100 (CET) X-Original-To: xvid-devel@edu.bnhof.de Delivered-To: xvid-devel@edu.bnhof.de Received: from s6.lansco.de (s6.lansco.de [213.239.218.110]) by edu.bnhof.de (Postfix) with ESMTP id D992218BA7 for ; Mon, 21 Nov 2005 15:14:22 +0100 (CET) Received: from weed.lut.ac.uk (weed.lut.ac.uk [158.125.1.226]) by s6.lansco.de (Postfix) with ESMTP id DE57C16129 for ; Mon, 21 Nov 2005 15:14:21 +0100 (CET) Received: from [158.125.1.221] (helo=magic.lut.ac.uk) by weed.lut.ac.uk with esmtp (Exim 4.54) id 1EeCR3-0000Yr-Av; Mon, 21 Nov 2005 14:14:21 +0000 Received: from apache by magic.lut.ac.uk with local (Exim 4.54) id 1EeCR3-0006oc-0D; Mon, 21 Nov 2005 14:14:21 +0000 Received: from elvc-linux3.lut.ac.uk (elvc-linux3.lut.ac.uk [158.125.51.81]) by student-webmail.lboro.ac.uk (IMP) with HTTP for ; Mon, 21 Nov 2005 14:14:20 +0000 Message-ID: <1132582460.4381d63ced3ab@student-webmail.lboro.ac.uk> Date: Mon, 21 Nov 2005 14:14:20 +0000 From: Tom Jacobs To: iibot@gmx.at, xvid-devel@xvid.org Subject: Re: [XviD-devel] hey I'm back :) References: <43815E50.8@syskin.cjb.net> <1132568787.4381a0d36ead6@www.lansco.de> <4381A3B4.70708@syskin.cjb.net> <4381D0BD.2090902@gmx.at> In-Reply-To: <4381D0BD.2090902@gmx.at> MIME-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 8bit User-Agent: Internet Messaging Program (IMP) 3.2.8 X-Originating-IP: 158.125.51.81 X-Scan-Signature: 7b0f4797dd98a006f3ae34a5eb70af8a Cc: X-BeenThere: xvid-devel@xvid.org X-Mailman-Version: 2.1.4 Precedence: list Reply-To: xvid-devel@xvid.org List-Id: xvid-devel.xvid.org List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Sender: xvid-devel-bounces@xvid.org Errors-To: xvid-devel-bounces@xvid.org hi what level of parallelism are you trying to implement skal? MB, frame, GOP? i dont know if any of my work will be of help but here is a link to one of my paper i wrote in the summer about it (im sure the ieee wont mind me posting it ;) ) http://www-staff.lboro.ac.uk/~eltrj/jacobstr_thread-level-parallel-xvid.pdf my work has used a custom multi processor environment so it is more abstract from any system setup for the 'normal' user but i think the therory would remain the same. i know that x264 uses pthread as method of creating different thread etc but im not sure how this works. it was very easy to convert x264 pthread code to work in my system so hope it is just as easy vise vera. Tom Quoting iibot : > Radek Czyz schrieb: > > Michael Militzer wrote: > >> Also, Skal is working on multithreaded code and has made a lot (good > and > >> bad) experiences with this recently. So you should check how your > ideas > >> here overlap (or not) just to make sure you don't do things twice... > > > > Bah, there goes my fun/learning ;) > > Good to have you back. > > > I mostly want to know *how* to write such code, so even if things > > repeat, it's not a huge loss. > > Nice to see someone volunteer to code that striped MB encoding I > suggested a while back. :D > > bis besser, > Tobias > > > _______________________________________________ > XviD-devel mailing list > XviD-devel@xvid.org > http://list.xvid.org/mailman/listinfo/xvid-devel > _______________________________________________ XviD-devel mailing list XviD-devel@xvid.org http://list.xvid.org/mailman/listinfo/xvid-devel From xvid-devel-bounces@xvid.org Mon Nov 21 15:16:48 2005 Return-Path: X-Original-To: itdp@fh-biergarten.de Delivered-To: itdp@localhost Received: from edu.bnhof.de (edu.bnhof.de [213.167.167.52]) by mail.kliche.org (Postfix) with ESMTP id 9F419126D82 for ; Mon, 21 Nov 2005 15:16:48 +0100 (CET) Received: from edu.bnhof.de (localhost [127.0.0.1]) by edu.bnhof.de (Postfix) with ESMTP id 33DC318BC2; Mon, 21 Nov 2005 15:16:41 +0100 (CET) X-Original-To: xvid-devel@edu.bnhof.de Delivered-To: xvid-devel@edu.bnhof.de Received: from s6.lansco.de (s6.lansco.de [213.239.218.110]) by edu.bnhof.de (Postfix) with ESMTP id DABC118BBF for ; Mon, 21 Nov 2005 15:16:37 +0100 (CET) Received: from mail.gmx.net (mail.gmx.net [213.165.64.20]) by s6.lansco.de (Postfix) with SMTP id 418EF16129 for ; Mon, 21 Nov 2005 15:16:36 +0100 (CET) Received: (qmail invoked by alias); 21 Nov 2005 14:16:35 -0000 Received: from u-122-163.adsl.univie.ac.at (HELO carl5f404cvpbw) [131.130.122.163] by mail.gmx.net (mp015) with SMTP; 21 Nov 2005 15:16:35 +0100 X-Authenticated: #11725910 Message-ID: <004a01c5eea6$1801a530$8c00000a@carl5f404cvpbw> From: "bond" To: References: <20051121140900.D67AC18BA0@edu.bnhof.de> Subject: Re: [XviD-devel] hey I'm back :) Date: Mon, 21 Nov 2005 15:15:58 +0100 MIME-Version: 1.0 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: 7bit X-Priority: 3 X-MSMail-Priority: Normal X-Mailer: Microsoft Outlook Express 6.00.2800.1506 X-MimeOLE: Produced By Microsoft MimeOLE V6.00.2800.1506 X-Y-GMX-Trusted: 0 X-BeenThere: xvid-devel@xvid.org X-Mailman-Version: 2.1.4 Precedence: list Reply-To: xvid-devel@xvid.org List-Id: xvid-devel.xvid.org List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Sender: xvid-devel-bounces@xvid.org Errors-To: xvid-devel-bounces@xvid.org > I have summer vacations now, FINALLY :) and this means I'm back to xvid > poking, with some fresh ideas and motivations. mooo!!! ;) > - xvidenc - a commandline executable like x264 has. It turns out there's > a demand for such executable, we already have a lot of code, and > everything else can be taken from x264 there is this encraw version discussed on doom9 there is also a very basic and somewhat old cmdl xvid encoder with .mp4 output in the mpeg4ip cvs which might be interesting to look at? cheers bond _______________________________________________ XviD-devel mailing list XviD-devel@xvid.org http://list.xvid.org/mailman/listinfo/xvid-devel From xvid-devel-bounces@xvid.org Mon Nov 21 15:41:15 2005 Return-Path: X-Original-To: itdp@fh-biergarten.de Delivered-To: itdp@localhost Received: from edu.bnhof.de (edu.bnhof.de [213.167.167.52]) by mail.kliche.org (Postfix) with ESMTP id CE7C6126D82 for ; Mon, 21 Nov 2005 15:41:15 +0100 (CET) Received: from edu.bnhof.de (localhost [127.0.0.1]) by edu.bnhof.de (Postfix) with ESMTP id 5196118BA0; Mon, 21 Nov 2005 15:41:08 +0100 (CET) X-Original-To: xvid-devel@edu.bnhof.de Delivered-To: xvid-devel@edu.bnhof.de Received: from s6.lansco.de (s6.lansco.de [213.239.218.110]) by edu.bnhof.de (Postfix) with ESMTP id 060CB18AA0 for ; Mon, 21 Nov 2005 15:40:58 +0100 (CET) Received: from ftp.ilog.fr (ftp.ilog.fr [81.80.162.195]) by s6.lansco.de (Postfix) with ESMTP id 636BF16198 for ; Mon, 21 Nov 2005 15:40:57 +0100 (CET) Received: from laposte.ilog.fr (cerbere-qfe0 [81.80.162.193]) by ftp.ilog.fr (8.13.1/8.13.1) with ESMTP id jALEevEW003291 for ; Mon, 21 Nov 2005 15:40:57 +0100 Received: from marbore.ilog.biz (marbore1.ilog.fr [172.17.2.61]) by laposte.ilog.fr (8.13.1/8.13.1) with ESMTP id jALEephE000199 for ; Mon, 21 Nov 2005 15:40:52 +0100 (MET) Received: from parmbx01.ilog.biz ([172.17.2.64]) by marbore.ilog.biz with Microsoft SMTPSVC(6.0.3790.1830); Mon, 21 Nov 2005 15:40:49 +0100 Received: from 6ttgg1j ([172.17.4.249]) by parmbx01.ilog.biz with Microsoft SMTPSVC(6.0.3790.1830); Mon, 21 Nov 2005 15:40:49 +0100 Subject: Re: [XviD-devel] hey I'm back :) From: Skal To: xvid-devel@xvid.org In-Reply-To: <4381D4E8.6070003@leoninedev.com> References: <43815E50.8@syskin.cjb.net> <4381D4E8.6070003@leoninedev.com> Content-Type: text/plain Message-Id: <1132584047.3309.18.camel@localhost.localdomain> Mime-Version: 1.0 X-Mailer: Ximian Evolution 1.4.5 (1.4.5-14) Date: Mon, 21 Nov 2005 15:40:47 +0100 Content-Transfer-Encoding: 7bit X-OriginalArrivalTime: 21 Nov 2005 14:40:49.0923 (UTC) FILETIME=[90FBC530:01C5EEA9] X-BeenThere: xvid-devel@xvid.org X-Mailman-Version: 2.1.4 Precedence: list Reply-To: xvid-devel@xvid.org List-Id: xvid-devel.xvid.org List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Sender: xvid-devel-bounces@xvid.org Errors-To: xvid-devel-bounces@xvid.org Hi Bryan, On Mon, 2005-11-21 at 15:08, Bryan Mayland wrote: > Welcome back! > > Radek Czyz wrote: > > - multithreaded code. I have several ideas how various parts of XviD can > > be multithreaded, and I intend to implement them, if only to learn how. > I've also got a multithreaded MotionEstimation() function for P-VOPs > that is mostly complete and waiting for Skal's thread abstraction. The > algorithm can scale to an arbitrary number of threads is the API will > permit. yep, sorry i'm late sending you the stuff. You can get the code here: http://skal.planet-d.net/slave.c http://skal.planet-d.net/slave.h since it got stripped in the previous mail. As said, it's most probably buggy, so be careful... later! Skal _______________________________________________ XviD-devel mailing list XviD-devel@xvid.org http://list.xvid.org/mailman/listinfo/xvid-devel From xvid-devel-bounces@xvid.org Mon Nov 21 15:58:05 2005 Return-Path: X-Original-To: itdp@fh-biergarten.de Delivered-To: itdp@localhost Received: from edu.bnhof.de (edu.bnhof.de [213.167.167.52]) by mail.kliche.org (Postfix) with ESMTP id EA544126D82 for ; Mon, 21 Nov 2005 15:58:04 +0100 (CET) Received: from edu.bnhof.de (localhost [127.0.0.1]) by edu.bnhof.de (Postfix) with ESMTP id 1D51A18BAE; Mon, 21 Nov 2005 15:57:57 +0100 (CET) X-Original-To: xvid-devel@edu.bnhof.de Delivered-To: xvid-devel@edu.bnhof.de Received: from s6.lansco.de (s6.lansco.de [213.239.218.110]) by edu.bnhof.de (Postfix) with ESMTP id 4927C18B90 for ; Mon, 21 Nov 2005 15:57:55 +0100 (CET) Received: from mail.ln9.net (rrcs-24-73-193-138.se.biz.rr.com [24.73.193.138]) by s6.lansco.de (Postfix) with ESMTP id 2DAA716144 for ; Mon, 21 Nov 2005 15:57:54 +0100 (CET) Received: from [192.168.100.164] (yarr.ad.ln9.net [192.168.100.164]) by mail.ln9.net (Postfix) with ESMTP id 385EF4B0194 for ; Mon, 21 Nov 2005 09:57:48 -0500 (EST) Message-ID: <4381E06B.5070108@leoninedev.com> Date: Mon, 21 Nov 2005 09:57:47 -0500 From: Bryan Mayland Organization: Leonine Development, Inc. User-Agent: Thunderbird 1.5 (Windows/20051025) MIME-Version: 1.0 To: xvid-devel@xvid.org Subject: Re: [XviD-devel] hey I'm back :) References: <43815E50.8@syskin.cjb.net> <4381D4E8.6070003@leoninedev.com> <1132584047.3309.18.camel@localhost.localdomain> In-Reply-To: <1132584047.3309.18.camel@localhost.localdomain> Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 7bit X-BeenThere: xvid-devel@xvid.org X-Mailman-Version: 2.1.4 Precedence: list Reply-To: xvid-devel@xvid.org List-Id: xvid-devel.xvid.org List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Sender: xvid-devel-bounces@xvid.org Errors-To: xvid-devel-bounces@xvid.org Skal wrote: > yep, sorry i'm late sending you the stuff. > You can get the code here: > > http://skal.planet-d.net/slave.c > http://skal.planet-d.net/slave.h > Oh great, now I actually have to finish up work on my code? Thanks a lot! :) _______________________________________________ XviD-devel mailing list XviD-devel@xvid.org http://list.xvid.org/mailman/listinfo/xvid-devel From xvid-devel-bounces@xvid.org Mon Nov 21 17:54:32 2005 Return-Path: X-Original-To: itdp@fh-biergarten.de Delivered-To: itdp@localhost Received: from edu.bnhof.de (edu.bnhof.de [213.167.167.52]) by mail.kliche.org (Postfix) with ESMTP id 7503E126D82 for ; Mon, 21 Nov 2005 17:54:32 +0100 (CET) Received: from edu.bnhof.de (localhost [127.0.0.1]) by edu.bnhof.de (Postfix) with ESMTP id A05BD18B7A; Mon, 21 Nov 2005 17:54:23 +0100 (CET) X-Original-To: xvid-devel@edu.bnhof.de Delivered-To: xvid-devel@edu.bnhof.de Received: from s6.lansco.de (s6.lansco.de [213.239.218.110]) by edu.bnhof.de (Postfix) with ESMTP id 31AEA15888 for ; Mon, 21 Nov 2005 17:54:19 +0100 (CET) Received: from mail.ln9.net (rrcs-24-73-193-138.se.biz.rr.com [24.73.193.138]) by s6.lansco.de (Postfix) with ESMTP id B4ABD161A5 for ; Mon, 21 Nov 2005 17:54:18 +0100 (CET) Received: from [192.168.100.164] (yarr.ad.ln9.net [192.168.100.164]) by mail.ln9.net (Postfix) with ESMTP id 9E5904B00A8 for ; Mon, 21 Nov 2005 11:54:17 -0500 (EST) Message-ID: <4381FBB9.70408@leoninedev.com> Date: Mon, 21 Nov 2005 11:54:17 -0500 From: Bryan Mayland Organization: Leonine Development, Inc. User-Agent: Thunderbird 1.5 (Windows/20051025) MIME-Version: 1.0 To: xvid-devel@xvid.org Subject: Re: [XviD-devel] hey I'm back :) References: <43815E50.8@syskin.cjb.net> <4381D4E8.6070003@leoninedev.com> <1132584047.3309.18.camel@localhost.localdomain> In-Reply-To: <1132584047.3309.18.camel@localhost.localdomain> Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 7bit X-BeenThere: xvid-devel@xvid.org X-Mailman-Version: 2.1.4 Precedence: list Reply-To: xvid-devel@xvid.org List-Id: xvid-devel.xvid.org List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Sender: xvid-devel-bounces@xvid.org Errors-To: xvid-devel-bounces@xvid.org Skal wrote: > yep, sorry i'm late sending you the stuff. > You can get the code here: > http://skal.planet-d.net/slave.c > http://skal.planet-d.net/slave.h > > The good news is that I've fixed the deadlock that is in the code I sent you earlier, and am getting a much better distribution of macroblocks processed per thread: (warning, 2051 pixel wide image) http://capnbry.net/~bmayland/xvid/xvid-mpme.png Diff and win32 binary: http://capnbry.net/~bmayland/xvid/multi.diff http://capnbry.net/~bmayland/xvid/xvidcoreMP.zip I guess the big question now is how to best use your abstraction to clean up my code. Currently in CVS, there is one MotionEstimation(...) function. In my code, I've *copied* the internals of that function into a thread function with a basic while(!terminated) loop. The main thread now puts its parameters in variables accessible by the thread, kicks it off, then contributes by running its own ME loop. When out of macroblocks the threads rejoin and return from MotionEstimation(...). I'm thinking the way to do this now would be to remove the entirety of the MotionEstimation function into the thread function. For single-threaded execution, the thread function will just call the thread function directly. For multi-threaded execution, I'm leaning towards: For each thread: copy ME()'s parameters into a structure and put it in a ThreadWorkerData xvThreadSetTodo(&threadN, threadworkerdataN); // at this point all threads should be working on ME For each thread: xvThreadSync(&threadN); Alternatively, we can xvThreadInit numthreads - 1, SetTodo on them and call one thread function directly so the main thread is doing more than waiting for everyone to finish. I don't think that would look that bad in code, and has the benefit of creating one less thread. Of course, this starts to be a big mess if you need to set this up everywhere you want parallelism. So what's the better way? I'm open to suggestion before I dig into making my changes. _______________________________________________ XviD-devel mailing list XviD-devel@xvid.org http://list.xvid.org/mailman/listinfo/xvid-devel From xvid-devel-bounces@xvid.org Fri Nov 25 09:10:53 2005 Return-Path: X-Original-To: itdp@fh-biergarten.de Delivered-To: itdp@localhost Received: from edu.bnhof.de (edu.bnhof.de [213.167.167.52]) by mail.kliche.org (Postfix) with ESMTP id ACD7E126D8B for ; Fri, 25 Nov 2005 09:10:53 +0100 (CET) Received: from edu.bnhof.de (localhost [127.0.0.1]) by edu.bnhof.de (Postfix) with ESMTP id 5B1CF18B6C; Fri, 25 Nov 2005 09:10:45 +0100 (CET) X-Original-To: xvid-devel@edu.bnhof.de Delivered-To: xvid-devel@edu.bnhof.de Received: from s6.lansco.de (s6.lansco.de [213.239.218.110]) by edu.bnhof.de (Postfix) with ESMTP id 5301D158C2 for ; Fri, 25 Nov 2005 09:10:43 +0100 (CET) Received: from customer-domains.icp-qv1-irony7.iinet.net.au (customer-domains.icp-qv1-irony7.iinet.net.au [203.59.1.128]) by s6.lansco.de (Postfix) with ESMTP id 39659176F5 for ; Fri, 25 Nov 2005 09:10:40 +0100 (CET) Received: from 203-173-56-106.dyn.iinet.net.au (HELO [10.1.1.3]) ([203.173.56.106]) by customer-domains.icp-qv1-irony7.iinet.net.au with ESMTP; 25 Nov 2005 16:10:32 +0800 X-BrightmailFiltered: true X-Brightmail-Tracker: AAAAAA== Message-ID: <4386C71D.5080002@syskin.cjb.net> Date: Fri, 25 Nov 2005 18:41:09 +1030 From: Radek Czyz User-Agent: Mozilla Thunderbird 1.0.2 (Windows/20050317) X-Accept-Language: en-us, en MIME-Version: 1.0 To: xvid-devel@xvid.org Subject: Re: [XviD-devel] Lambda for plugins References: <435C0647.5090205@uni-konstanz.de> In-Reply-To: <435C0647.5090205@uni-konstanz.de> Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 7bit X-BeenThere: xvid-devel@xvid.org X-Mailman-Version: 2.1.4 Precedence: list Reply-To: xvid-devel@xvid.org List-Id: xvid-devel.xvid.org List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Sender: xvid-devel-bounces@xvid.org Errors-To: xvid-devel-bounces@xvid.org Hey everyone, I cleaned up Johannes's patch, got rid of some bugs and probably added some new ones. The patch is here: http://syskin.is.dreaming.org/lambda.patch Most of the code is copy&paste of adaptive quantization code. If plugins want to, they request a lambda table, which is six floats for each macroblock big. Initially it's full of 1.0f, but plugins can modify this table to indicate lower-importance and higher-importance blocks. Internally, the floats are then converted to ints, with 1< Hi everybody, > > I tried to expose the lambda values for macroblocks to plugins, to be > able to do some HVS stuff. It seems to compile fine and not bork > anything here. I am at the moment not able to test it more careful. The > patch is against the latest CVS. > > http://web3.gleitz.org/Kopernikus/lambdaplugin.diff > > But for now, I have not tried to code a plugin that uses the lambda > values. Can someone with a better knowledge of the XviD Code look > through it, if I have borked some really important things, propose > improvements or correct me if my style is bad. > > I have some questions on the code: > > Why is dquant_stride in xvid_plg_data_t in xvid.h needed? Is it possible > that its different from mb_width? > > I assumed that the lambda[6] in SearchData in estimation.h is the lambda > for the 6 blocks of a macroblock (4 Luminance and 2 Chrominance blocks). > Is that correct? > > In xvid_plg_data_t there is a remark that deprecated data is kept for > binary compatibility. Does it break binary compatibility if I add the > pointer to the lambda table to xvid_plg_data? Would it make sense, to > replace one of the deprecated ints with the pointer to keep compatibility? > > Does it make sense, to multiplicate the lambda in dct_quantize_trellis_c > in mbtransquant.c with the lambda modificator? I am not sure I > understood trellis completly. Is there any further information about > trellis quantization? > > Why are the dquants available for plugins at PLUGIN_AFTER? Is that > needed for the lambdas? > > And finally, is there a draft of the specs or something similar > available for download. I would like to have precise Information on some > things. > > Thanks in advance > > Johannes Reinhardt > _______________________________________________ > XviD-devel mailing list > XviD-devel@xvid.org > http://list.xvid.org/mailman/listinfo/xvid-devel > > _______________________________________________ XviD-devel mailing list XviD-devel@xvid.org http://list.xvid.org/mailman/listinfo/xvid-devel From xvid-devel-bounces@xvid.org Fri Nov 25 15:08:06 2005 Return-Path: X-Original-To: itdp@fh-biergarten.de Delivered-To: itdp@localhost Received: from edu.bnhof.de (edu.bnhof.de [213.167.167.52]) by mail.kliche.org (Postfix) with ESMTP id 86C9D126D92 for ; Fri, 25 Nov 2005 15:08:06 +0100 (CET) Received: from edu.bnhof.de (localhost [127.0.0.1]) by edu.bnhof.de (Postfix) with ESMTP id 3BD9D13660; Fri, 25 Nov 2005 15:08:01 +0100 (CET) X-Original-To: xvid-devel@edu.bnhof.de Delivered-To: xvid-devel@edu.bnhof.de Received: from s6.lansco.de (s6.lansco.de [213.239.218.110]) by edu.bnhof.de (Postfix) with ESMTP id A8AB0A83B for ; Fri, 25 Nov 2005 15:07:58 +0100 (CET) Received: from mail-ihug.icp-qv1-irony5.iinet.net.au (ihug-mail.icp-qv1-irony5.iinet.net.au [203.59.1.199]) by s6.lansco.de (Postfix) with ESMTP id 9581517817 for ; Fri, 25 Nov 2005 15:07:56 +0100 (CET) Received: from 203-173-56-106.dyn.iinet.net.au (HELO [10.1.1.3]) ([203.173.56.106]) by mail-ihug.icp-qv1-irony5.iinet.net.au with ESMTP; 25 Nov 2005 22:07:54 +0800 X-BrightmailFiltered: true X-Brightmail-Tracker: AAAAAA== Message-ID: <43871AE3.4040409@syskin.cjb.net> Date: Sat, 26 Nov 2005 00:38:35 +1030 From: Radek Czyz User-Agent: Mozilla Thunderbird 1.0.2 (Windows/20050317) X-Accept-Language: en-us, en MIME-Version: 1.0 To: xvid-devel@xvid.org Content-Type: text/plain; charset=UTF-8; format=flowed Content-Transfer-Encoding: 7bit Subject: [XviD-devel] HUGE ME regressions between beta1 and beta2 X-BeenThere: xvid-devel@xvid.org X-Mailman-Version: 2.1.4 Precedence: list Reply-To: xvid-devel@xvid.org List-Id: xvid-devel.xvid.org List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Sender: xvid-devel-bounces@xvid.org Errors-To: xvid-devel-bounces@xvid.org Hi everyone, Apparently our testing community sucks ;)) There was a HUGE quality and speed regression between beta1 and beta2. I narrowed it down to SAD-based ME. The major changes in that time was Isibaar's new lambda code, but apparently we have hardly any logs what happened exactly (well, we've got cvs, I'll check that). For fast settings (no VHQ, no trellis, no chromaME, defaults otherwise): Beta1's ME: time 1:19 filesize 26.2MB psnr 41.19dB Current ME: time 1:24 filesize 28.8MB psnr 41.14dB What the heck happened? ;_; I suggest we postpone 1.1final until we figure this out - looks like all speed improvements over 1.0 are gone (10% slower at these settings, 5% slower at defaults), and 10% more filesize for *lower* quality is... bad. Radek _______________________________________________ XviD-devel mailing list XviD-devel@xvid.org http://list.xvid.org/mailman/listinfo/xvid-devel From xvid-devel-bounces@xvid.org Fri Nov 25 15:36:42 2005 Return-Path: X-Original-To: itdp@fh-biergarten.de Delivered-To: itdp@localhost Received: from edu.bnhof.de (edu.bnhof.de [213.167.167.52]) by mail.kliche.org (Postfix) with ESMTP id EEAF2126D8B for ; Fri, 25 Nov 2005 15:36:41 +0100 (CET) Received: from edu.bnhof.de (localhost [127.0.0.1]) by edu.bnhof.de (Postfix) with ESMTP id AC40E19983; Fri, 25 Nov 2005 15:36:37 +0100 (CET) X-Original-To: xvid-devel@edu.bnhof.de Delivered-To: xvid-devel@edu.bnhof.de Received: from s6.lansco.de (s6.lansco.de [213.239.218.110]) by edu.bnhof.de (Postfix) with ESMTP id B699E1DD7E for ; Fri, 25 Nov 2005 15:35:21 +0100 (CET) Received: from s5.lansco.de (s5.lansco.de [212.63.130.170]) by s6.lansco.de (Postfix) with ESMTP id B1D5F177F5 for ; Fri, 25 Nov 2005 15:35:20 +0100 (CET) Received: from localhost (s5.lansco.de [212.63.130.170]) by s5.lansco.de (Postfix) with ESMTP id 838D61D303 for ; Fri, 25 Nov 2005 15:35:15 +0100 (CET) Received: from p54B5F37D.dip.t-dialin.net (p54B5F37D.dip.t-dialin.net [84.181.243.125]) by www.lansco.de (IMP) with HTTP for ; Fri, 25 Nov 2005 15:35:15 +0100 Message-ID: <1132929315.4387212367133@www.lansco.de> Date: Fri, 25 Nov 2005 15:35:15 +0100 From: Michael Militzer To: xvid-devel@xvid.org Subject: [XviD-devel] HUGE ME regressions between beta1 and beta2 MIME-Version: 1.0 Content-Type: multipart/mixed; boundary="-MOQ1132929315b5738482271ac216dede758390da4fd1" User-Agent: Internet Messaging Program (IMP) 3.2.3 X-Originating-IP: 84.181.243.125 X-Mailman-Approved-At: Fri, 25 Nov 2005 15:36:32 +0100 X-Content-Filtered-By: Mailman/MimeDel 2.1.4 X-BeenThere: xvid-devel@xvid.org X-Mailman-Version: 2.1.4 Precedence: list Reply-To: xvid-devel@xvid.org List-Id: xvid-devel.xvid.org List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Sender: xvid-devel-bounces@xvid.org Errors-To: xvid-devel-bounces@xvid.org This message is in MIME format. ---MOQ1132929315b5738482271ac216dede758390da4fd1 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 8bit Hi Radek, well, I've basically just changed our lambda tables. Before, the values for lambda were borrowed from h.264 it seemed however didn't fit XviD's linear qscale at all. In result, R-D performance especially for higher QPs was absolutely horrible. I've changed it to a more lambda = 0.85*Q scheme (ok, can't recall the exact constant from memory now). I've tested the change for VHQ and non-VHQ modes and had always superior performance, sometimes with really huge gains especially a low bit-rates/high quants. I've attached my original mail to the list at the end of this mails and also reposted some of the R-D curves I did at those times. Your testing case is covered as well (it seems). Sure we can postpone 1.1 final until this problem is fixed. BTW: did you use a standard sequence for your test so we can reproduce the problem? Regards, Michael Quoting Radek Czyz : > Hi everyone, > > Apparently our testing community sucks ;)) > > There was a HUGE quality and speed regression between beta1 and beta2. I > narrowed it down to SAD-based ME. The major changes in that time was > Isibaar's new lambda code, but apparently we have hardly any logs what > happened exactly (well, we've got cvs, I'll check that). > > For fast settings (no VHQ, no trellis, no chromaME, defaults otherwise): > > Beta1's ME: time 1:19 filesize 26.2MB psnr 41.19dB > Current ME: time 1:24 filesize 28.8MB psnr 41.14dB > > > What the heck happened? ;_; > > I suggest we postpone 1.1final until we figure this out - looks like all > speed improvements over 1.0 are gone (10% slower at these settings, 5% > slower at defaults), and 10% more filesize for *lower* quality is... bad. > > Radek > _______________________________________________ > XviD-devel mailing list > XviD-devel@xvid.org > http://list.xvid.org/mailman/listinfo/xvid-devel > ----- Forwarded message from Michael Militzer ----- Date: Mon, 14 Mar 2005 19:59:37 +0100 From: Michael Militzer Reply-To: Michael Militzer Subject: low-bitrate patch To: xvid-devel@xvid.org Hi all, I've committed a patch to cvs that improves XVID's performance at low- bitrates by using better lambda values for R-D vector search. By this also mode decision (SAD-based) is improved at low-bitrates. At mid/higher bitrates there's no change in performance, so only people who aim at really low bitrate (and quality) encodes will benefit from this patch. I've attached some R-D graphs that compare the behavior of patched and unpatched XVID for various sequences and different bit-rates. Note that I've used a logarithmic scale, so that the differences at low bitrates can be better perceived. As can be clearly noticed, the compression performance is enhanced significantly at very low bitrates (up to 2-3 dB PSNR). Well, it's another question how useful this patch actually is because such extremely low bit-rates are barely used. But for completeness, XVID should perform equally great at all bitrates... bye, Michael ----- End forwarded message ----- ---MOQ1132929315b5738482271ac216dede758390da4fd1 Content-Type: text/plain; charset="us-ascii" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit Content-Disposition: inline _______________________________________________ XviD-devel mailing list XviD-devel@xvid.org http://list.xvid.org/mailman/listinfo/xvid-devel ---MOQ1132929315b5738482271ac216dede758390da4fd1-- From xvid-devel-bounces@xvid.org Fri Nov 25 15:39:09 2005 Return-Path: X-Original-To: itdp@fh-biergarten.de Delivered-To: itdp@localhost Received: from edu.bnhof.de (edu.bnhof.de [213.167.167.52]) by mail.kliche.org (Postfix) with ESMTP id D76F0126D92 for ; Fri, 25 Nov 2005 15:39:08 +0100 (CET) Received: from edu.bnhof.de (localhost [127.0.0.1]) by edu.bnhof.de (Postfix) with ESMTP id AAC051DD89; Fri, 25 Nov 2005 15:39:03 +0100 (CET) X-Original-To: xvid-devel@edu.bnhof.de Delivered-To: xvid-devel@edu.bnhof.de Received: from s6.lansco.de (s6.lansco.de [213.239.218.110]) by edu.bnhof.de (Postfix) with ESMTP id D17321DD85 for ; Fri, 25 Nov 2005 15:39:00 +0100 (CET) Received: from s5.lansco.de (s5.lansco.de [212.63.130.170]) by s6.lansco.de (Postfix) with ESMTP id 8DCA5177F6 for ; Fri, 25 Nov 2005 15:38:59 +0100 (CET) Received: from localhost (s5.lansco.de [212.63.130.170]) by s5.lansco.de (Postfix) with ESMTP id 739C02576A for ; Fri, 25 Nov 2005 15:38:59 +0100 (CET) Received: from p54B5F37D.dip.t-dialin.net (p54B5F37D.dip.t-dialin.net [84.181.243.125]) by www.lansco.de (IMP) with HTTP for ; Fri, 25 Nov 2005 15:38:59 +0100 Message-ID: <1132929539.43872203572e6@www.lansco.de> Date: Fri, 25 Nov 2005 15:38:59 +0100 From: Michael Militzer To: xvid-devel@xvid.org Subject: Re: [XviD-devel] HUGE ME regressions between beta1 and beta2 References: <1132929315.4387212367133@www.lansco.de> In-Reply-To: <1132929315.4387212367133@www.lansco.de> MIME-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 8bit User-Agent: Internet Messaging Program (IMP) 3.2.3 X-Originating-IP: 84.181.243.125 X-BeenThere: xvid-devel@xvid.org X-Mailman-Version: 2.1.4 Precedence: list Reply-To: xvid-devel@xvid.org List-Id: xvid-devel.xvid.org List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Sender: xvid-devel-bounces@xvid.org Errors-To: xvid-devel-bounces@xvid.org As always, attachments got stripped. So have a look here: http://www.xvid.org/images/carphone.png http://www.xvid.org/images/highway.png http://www.xvid.org/images/mother.png http://www.xvid.org/images/paris.png http://www.xvid.org/images/tempete.png Regards, Michael Quoting Michael Militzer : > Hi Radek, > > well, I've basically just changed our lambda tables. Before, the values for > lambda were borrowed from h.264 it seemed however didn't fit XviD's linear > qscale at all. In result, R-D performance especially for higher QPs was > absolutely horrible. I've changed it to a more lambda = 0.85*Q scheme (ok, > can't recall the exact constant from memory now). I've tested the change > for VHQ and non-VHQ modes and had always superior performance, sometimes > with really huge gains especially a low bit-rates/high quants. > > I've attached my original mail to the list at the end of this mails and also > reposted some of the R-D curves I did at those times. Your testing case is > covered as well (it seems). > > Sure we can postpone 1.1 final until this problem is fixed. BTW: did you use > a standard sequence for your test so we can reproduce the problem? > > Regards, > Michael > > > Quoting Radek Czyz : > > > Hi everyone, > > > > Apparently our testing community sucks ;)) > > > > There was a HUGE quality and speed regression between beta1 and beta2. I > > narrowed it down to SAD-based ME. The major changes in that time was > > Isibaar's new lambda code, but apparently we have hardly any logs what > > happened exactly (well, we've got cvs, I'll check that). > > > > For fast settings (no VHQ, no trellis, no chromaME, defaults otherwise): > > > > Beta1's ME: time 1:19 filesize 26.2MB psnr 41.19dB > > Current ME: time 1:24 filesize 28.8MB psnr 41.14dB > > > > > > What the heck happened? ;_; > > > > I suggest we postpone 1.1final until we figure this out - looks like all > > speed improvements over 1.0 are gone (10% slower at these settings, 5% > > slower at defaults), and 10% more filesize for *lower* quality is... bad. > > > > Radek > > _______________________________________________ > > XviD-devel mailing list > > XviD-devel@xvid.org > > http://list.xvid.org/mailman/listinfo/xvid-devel > > > > > ----- Forwarded message from Michael Militzer ----- > Date: Mon, 14 Mar 2005 19:59:37 +0100 > From: Michael Militzer > Reply-To: Michael Militzer > Subject: low-bitrate patch > To: xvid-devel@xvid.org > > Hi all, > > I've committed a patch to cvs that improves XVID's performance at low- > bitrates by using better lambda values for R-D vector search. By this also > mode decision (SAD-based) is improved at low-bitrates. At mid/higher > bitrates > there's no change in performance, so only people who aim at really low > bitrate (and quality) encodes will benefit from this patch. > > I've attached some R-D graphs that compare the behavior of patched and > unpatched XVID for various sequences and different bit-rates. Note that I've > used a logarithmic scale, so that the differences at low bitrates can be > better perceived. As can be clearly noticed, the compression performance is > enhanced significantly at very low bitrates (up to 2-3 dB PSNR). > > Well, it's another question how useful this patch actually is because such > extremely low bit-rates are barely used. But for completeness, XVID should > perform equally great at all bitrates... > > bye, > Michael > > ----- End forwarded message ----- > > > > _______________________________________________ XviD-devel mailing list XviD-devel@xvid.org http://list.xvid.org/mailman/listinfo/xvid-devel From xvid-devel-bounces@xvid.org Fri Nov 25 16:12:24 2005 Return-Path: X-Original-To: itdp@fh-biergarten.de Delivered-To: itdp@localhost Received: from edu.bnhof.de (edu.bnhof.de [213.167.167.52]) by mail.kliche.org (Postfix) with ESMTP id 1681A126D8B for ; Fri, 25 Nov 2005 16:12:24 +0100 (CET) Received: from edu.bnhof.de (localhost [127.0.0.1]) by edu.bnhof.de (Postfix) with ESMTP id 1A51C18B7A; Fri, 25 Nov 2005 16:12:20 +0100 (CET) X-Original-To: xvid-devel@edu.bnhof.de Delivered-To: xvid-devel@edu.bnhof.de Received: from s6.lansco.de (s6.lansco.de [213.239.218.110]) by edu.bnhof.de (Postfix) with ESMTP id 8BA769C5A for ; Fri, 25 Nov 2005 16:12:17 +0100 (CET) Received: from cytosin.uni-konstanz.de (cytosin.uni-konstanz.de [134.34.240.61]) by s6.lansco.de (Postfix) with ESMTP id 6F4BB1781A for ; Fri, 25 Nov 2005 16:12:16 +0100 (CET) Received: from [213.6.129.167] (A81a7.a.pppool.de [213.6.129.167]) (using TLSv1 with cipher DHE-RSA-AES256-SHA (256/256 bits)) (No client certificate requested) by cytosin.uni-konstanz.de (Postfix) with ESMTP id 86734D7F for ; Fri, 25 Nov 2005 16:12:14 +0100 (CET) Message-ID: <438729CA.8040201@uni-konstanz.de> Date: Fri, 25 Nov 2005 16:12:10 +0100 From: Johannes Reinhardt User-Agent: Mozilla Thunderbird 1.0.2 (Windows/20050317) X-Accept-Language: en-us, en MIME-Version: 1.0 To: xvid-devel@xvid.org Subject: Re: [XviD-devel] Lambda for plugins References: <435C0647.5090205@uni-konstanz.de> <4386C71D.5080002@syskin.cjb.net> In-Reply-To: <4386C71D.5080002@syskin.cjb.net> Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 7bit X-BeenThere: xvid-devel@xvid.org X-Mailman-Version: 2.1.4 Precedence: list Reply-To: xvid-devel@xvid.org List-Id: xvid-devel.xvid.org List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Sender: xvid-devel-bounces@xvid.org Errors-To: xvid-devel-bounces@xvid.org Hi Radek, Thank you for your effort. I see there were some things you solved more elegant. Can perhaps someone clarify these points from my original mail: -Why is dquant_stride in xvid_plg_data_t in xvid.h needed? Is it possible that its different from mb_width? -Is there a draft of the specs or something similar available for download. I would like to have precise Information on some things. Johannes Radek Czyz wrote: > Hey everyone, > > I cleaned up Johannes's patch, got rid of some bugs and probably added > some new ones. > > The patch is here: > http://syskin.is.dreaming.org/lambda.patch > > Most of the code is copy&paste of adaptive quantization code. If > plugins want to, they request a lambda table, which is six floats for > each macroblock big. Initially it's full of 1.0f, but plugins can > modify this table to indicate lower-importance and higher-importance > blocks. > Internally, the floats are then converted to ints, with 1< being "neutral" lambda. > > This lambda then modifies the real VHQ lambda (the code was originally > meant for 6 independant lambdas, there's hardly any change in ME area) > and in trellis. > > What do you think of this? This is originally what I meant to make, > but I somehow didn't. > If there is no objection by tomorrow, I'll commit this (btw can I > commit without disturbing 1_1_branch already?). > > Thanks, Johannes! > > Radek > > PS, new TODO: see if we can use this lambda to SKIP more in > less-important areas > > > > Johannes Reinhardt wrote: > >> Hi everybody, >> >> I tried to expose the lambda values for macroblocks to plugins, to be >> able to do some HVS stuff. It seems to compile fine and not bork >> anything here. I am at the moment not able to test it more careful. The >> patch is against the latest CVS. >> >> http://web3.gleitz.org/Kopernikus/lambdaplugin.diff >> >> But for now, I have not tried to code a plugin that uses the lambda >> values. Can someone with a better knowledge of the XviD Code look >> through it, if I have borked some really important things, propose >> improvements or correct me if my style is bad. >> >> I have some questions on the code: >> >> Why is dquant_stride in xvid_plg_data_t in xvid.h needed? Is it possible >> that its different from mb_width? >> >> I assumed that the lambda[6] in SearchData in estimation.h is the lambda >> for the 6 blocks of a macroblock (4 Luminance and 2 Chrominance blocks). >> Is that correct? >> >> In xvid_plg_data_t there is a remark that deprecated data is kept for >> binary compatibility. Does it break binary compatibility if I add the >> pointer to the lambda table to xvid_plg_data? Would it make sense, to >> replace one of the deprecated ints with the pointer to keep >> compatibility? >> >> Does it make sense, to multiplicate the lambda in dct_quantize_trellis_c >> in mbtransquant.c with the lambda modificator? I am not sure I >> understood trellis completly. Is there any further information about >> trellis quantization? >> >> Why are the dquants available for plugins at PLUGIN_AFTER? Is that >> needed for the lambdas? >> >> And finally, is there a draft of the specs or something similar >> available for download. I would like to have precise Information on some >> things. >> >> Thanks in advance >> >> Johannes Reinhardt >> _______________________________________________ >> XviD-devel mailing list >> XviD-devel@xvid.org >> http://list.xvid.org/mailman/listinfo/xvid-devel >> >> > _______________________________________________ > XviD-devel mailing list > XviD-devel@xvid.org > http://list.xvid.org/mailman/listinfo/xvid-devel > _______________________________________________ XviD-devel mailing list XviD-devel@xvid.org http://list.xvid.org/mailman/listinfo/xvid-devel From xvid-devel-bounces@xvid.org Fri Nov 25 16:46:15 2005 Return-Path: X-Original-To: itdp@fh-biergarten.de Delivered-To: itdp@localhost Received: from edu.bnhof.de (edu.bnhof.de [213.167.167.52]) by mail.kliche.org (Postfix) with ESMTP id 8FBB6126D8B for ; Fri, 25 Nov 2005 16:46:15 +0100 (CET) Received: from edu.bnhof.de (localhost [127.0.0.1]) by edu.bnhof.de (Postfix) with ESMTP id 3A0BA18B77; Fri, 25 Nov 2005 16:46:11 +0100 (CET) X-Original-To: xvid-devel@edu.bnhof.de Delivered-To: xvid-devel@edu.bnhof.de Received: from s6.lansco.de (s6.lansco.de [213.239.218.110]) by edu.bnhof.de (Postfix) with ESMTP id 70CD3DFAB for ; Fri, 25 Nov 2005 16:46:09 +0100 (CET) Received: from mail-ihug.icp-qv1-irony3.iinet.net.au (ihug-mail.icp-qv1-irony3.iinet.net.au [203.59.1.197]) by s6.lansco.de (Postfix) with ESMTP id 560E41784C for ; Fri, 25 Nov 2005 16:46:06 +0100 (CET) Received: from 203-173-56-106.dyn.iinet.net.au (HELO [10.1.1.3]) ([203.173.56.106]) by mail-ihug.icp-qv1-irony3.iinet.net.au with ESMTP; 25 Nov 2005 23:46:03 +0800 X-BrightmailFiltered: true X-Brightmail-Tracker: AAAAAA== Message-ID: <438731E1.9050905@syskin.cjb.net> Date: Sat, 26 Nov 2005 02:16:41 +1030 From: Radek Czyz User-Agent: Mozilla Thunderbird 1.0.2 (Windows/20050317) X-Accept-Language: en-us, en MIME-Version: 1.0 To: xvid-devel@xvid.org Subject: Re: [XviD-devel] HUGE ME regressions between beta1 and beta2 References: <1132929315.4387212367133@www.lansco.de> In-Reply-To: <1132929315.4387212367133@www.lansco.de> Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 7bit X-BeenThere: xvid-devel@xvid.org X-Mailman-Version: 2.1.4 Precedence: list Reply-To: xvid-devel@xvid.org List-Id: xvid-devel.xvid.org List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Sender: xvid-devel-bounces@xvid.org Errors-To: xvid-devel-bounces@xvid.org Hi again, Michael Militzer wrote: > well, I've basically just changed our lambda tables. Yes, but you also changed much more: the way lambda actually works went back to "traditional" way (which is in theory more correct). The lambda value multiplier changed from 10.5 to 0.6 (and from 40.0 to 0.6 for inter4v). In general, I'd expect that this kind of change has large consequences, in particular to inter/inter4v decisions. Overal, that was a reasonably large patch. The important thing is that I cannot confirm your results. This is what I get at pure defaults: Before: 1:53, 25.6MB, 41.70dB After: 2:03, 26.0MB, 41.78dB I am still unsure what causes the slowdown. Your patch removes one multiplication in important spot so I'd expect is to be faster. I've put the "old" files online, if anyone wants to test: http://syskin.is.dreaming.org/motion_old.zip Radek Before, the values for > lambda were borrowed from h.264 it seemed however didn't fit XviD's linear > qscale at all. In result, R-D performance especially for higher QPs was > absolutely horrible. I've changed it to a more lambda = 0.85*Q scheme (ok, > can't recall the exact constant from memory now). I've tested the change > for VHQ and non-VHQ modes and had always superior performance, sometimes > with really huge gains especially a low bit-rates/high quants. > > I've attached my original mail to the list at the end of this mails and also > reposted some of the R-D curves I did at those times. Your testing case is > covered as well (it seems). > > Sure we can postpone 1.1 final until this problem is fixed. BTW: did you use > a standard sequence for your test so we can reproduce the problem? > > Regards, > Michael > > > Quoting Radek Czyz : > > >>Hi everyone, >> >>Apparently our testing community sucks ;)) >> >>There was a HUGE quality and speed regression between beta1 and beta2. I >>narrowed it down to SAD-based ME. The major changes in that time was >>Isibaar's new lambda code, but apparently we have hardly any logs what >>happened exactly (well, we've got cvs, I'll check that). >> >>For fast settings (no VHQ, no trellis, no chromaME, defaults otherwise): >> >>Beta1's ME: time 1:19 filesize 26.2MB psnr 41.19dB >>Current ME: time 1:24 filesize 28.8MB psnr 41.14dB >> >> >>What the heck happened? ;_; >> >>I suggest we postpone 1.1final until we figure this out - looks like all >>speed improvements over 1.0 are gone (10% slower at these settings, 5% >>slower at defaults), and 10% more filesize for *lower* quality is... bad. >> >>Radek >>_______________________________________________ >>XviD-devel mailing list >>XviD-devel@xvid.org >>http://list.xvid.org/mailman/listinfo/xvid-devel >> > > > > ----- Forwarded message from Michael Militzer ----- > Date: Mon, 14 Mar 2005 19:59:37 +0100 > From: Michael Militzer > Reply-To: Michael Militzer > Subject: low-bitrate patch > To: xvid-devel@xvid.org > > Hi all, > > I've committed a patch to cvs that improves XVID's performance at low- > bitrates by using better lambda values for R-D vector search. By this also > mode decision (SAD-based) is improved at low-bitrates. At mid/higher bitrates > there's no change in performance, so only people who aim at really low > bitrate (and quality) encodes will benefit from this patch. > > I've attached some R-D graphs that compare the behavior of patched and > unpatched XVID for various sequences and different bit-rates. Note that I've > used a logarithmic scale, so that the differences at low bitrates can be > better perceived. As can be clearly noticed, the compression performance is > enhanced significantly at very low bitrates (up to 2-3 dB PSNR). > > Well, it's another question how useful this patch actually is because such > extremely low bit-rates are barely used. But for completeness, XVID should > perform equally great at all bitrates... > > bye, > Michael > > ----- End forwarded message ----- > > > > > > ------------------------------------------------------------------------ > > _______________________________________________ > XviD-devel mailing list > XviD-devel@xvid.org > http://list.xvid.org/mailman/listinfo/xvid-devel _______________________________________________ XviD-devel mailing list XviD-devel@xvid.org http://list.xvid.org/mailman/listinfo/xvid-devel From xvid-devel-bounces@xvid.org Fri Nov 25 17:01:45 2005 Return-Path: X-Original-To: itdp@fh-biergarten.de Delivered-To: itdp@localhost Received: from edu.bnhof.de (edu.bnhof.de [213.167.167.52]) by mail.kliche.org (Postfix) with ESMTP id 52485126D92 for ; Fri, 25 Nov 2005 17:01:45 +0100 (CET) Received: from edu.bnhof.de (localhost [127.0.0.1]) by edu.bnhof.de (Postfix) with ESMTP id 9EC6A18BA2; Fri, 25 Nov 2005 17:01:32 +0100 (CET) X-Original-To: xvid-devel@edu.bnhof.de Delivered-To: xvid-devel@edu.bnhof.de Received: from s6.lansco.de (s6.lansco.de [213.239.218.110]) by edu.bnhof.de (Postfix) with ESMTP id 9C86C18B8C for ; Fri, 25 Nov 2005 17:01:29 +0100 (CET) Received: from s5.lansco.de (s5.lansco.de [212.63.130.170]) by s6.lansco.de (Postfix) with ESMTP id 4FE7517856 for ; Fri, 25 Nov 2005 17:01:28 +0100 (CET) Received: from localhost (s5.lansco.de [212.63.130.170]) by s5.lansco.de (Postfix) with ESMTP id 24B1425870 for ; Fri, 25 Nov 2005 17:01:28 +0100 (CET) Received: from p54B5F37D.dip.t-dialin.net (p54B5F37D.dip.t-dialin.net [84.181.243.125]) by www.lansco.de (IMP) with HTTP for ; Fri, 25 Nov 2005 17:01:28 +0100 Message-ID: <1132934488.4387355800550@www.lansco.de> Date: Fri, 25 Nov 2005 17:01:28 +0100 From: Michael Militzer To: xvid-devel@xvid.org Subject: Re: [XviD-devel] HUGE ME regressions between beta1 and beta2 References: <1132929315.4387212367133@www.lansco.de> <438731E1.9050905@syskin.cjb.net> In-Reply-To: <438731E1.9050905@syskin.cjb.net> MIME-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 8bit User-Agent: Internet Messaging Program (IMP) 3.2.3 X-Originating-IP: 84.181.243.125 X-BeenThere: xvid-devel@xvid.org X-Mailman-Version: 2.1.4 Precedence: list Reply-To: xvid-devel@xvid.org List-Id: xvid-devel.xvid.org List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Sender: xvid-devel-bounces@xvid.org Errors-To: xvid-devel-bounces@xvid.org Hi, well, I'm rather confident that the current ("traditional") way finds better vectors in ME than before, especially at high quants. What I can imagine is that there's a larger influence on mode decision that I expected (and experienced) in non R-D optimized mode. I've tested with lots more sequences than just the ones I posted to the list and never had a regression with the new code. However, I used standard testing sequences from uncompressed input. Perhaps, mode decision with the new lambdas doesn't work well on DVD/MPEG-2 compressed input. That should be easy to find out: if there's no regression with VHQ>0, then the degradation on some sequences likely comes from mode decision. That should be easy to fix then by applying additional weights after vector search. Also, it may explain why the problem slipped through unnoticed - most people use VHQ I guess. BTW: with what sequence do you test? With the same I used and you get different results (that would be very concerning)? Or with another one? Regards, Michael Quoting Radek Czyz : > Hi again, > > Michael Militzer wrote: > > well, I've basically just changed our lambda tables. > > Yes, but you also changed much more: the way lambda actually works went > back to "traditional" way (which is in theory more correct). The lambda > value multiplier changed from 10.5 to 0.6 (and from 40.0 to 0.6 for > inter4v). > > In general, I'd expect that this kind of change has large consequences, > in particular to inter/inter4v decisions. > > Overal, that was a reasonably large patch. The important thing is that I > cannot confirm your results. This is what I get at pure defaults: > > Before: 1:53, 25.6MB, 41.70dB > After: 2:03, 26.0MB, 41.78dB > > I am still unsure what causes the slowdown. Your patch removes one > multiplication in important spot so I'd expect is to be faster. > > I've put the "old" files online, if anyone wants to test: > http://syskin.is.dreaming.org/motion_old.zip > > Radek > > > Before, the values for > > lambda were borrowed from h.264 it seemed however didn't fit XviD's linear > > qscale at all. In result, R-D performance especially for higher QPs was > > absolutely horrible. I've changed it to a more lambda = 0.85*Q scheme (ok, > > can't recall the exact constant from memory now). I've tested the change > > for VHQ and non-VHQ modes and had always superior performance, sometimes > > with really huge gains especially a low bit-rates/high quants. > > > > I've attached my original mail to the list at the end of this mails and > also > > reposted some of the R-D curves I did at those times. Your testing case is > > covered as well (it seems). > > > > Sure we can postpone 1.1 final until this problem is fixed. BTW: did you > use > > a standard sequence for your test so we can reproduce the problem? > > > > Regards, > > Michael > > > > > > Quoting Radek Czyz : > > > > > >>Hi everyone, > >> > >>Apparently our testing community sucks ;)) > >> > >>There was a HUGE quality and speed regression between beta1 and beta2. I > >>narrowed it down to SAD-based ME. The major changes in that time was > >>Isibaar's new lambda code, but apparently we have hardly any logs what > >>happened exactly (well, we've got cvs, I'll check that). > >> > >>For fast settings (no VHQ, no trellis, no chromaME, defaults otherwise): > >> > >>Beta1's ME: time 1:19 filesize 26.2MB psnr 41.19dB > >>Current ME: time 1:24 filesize 28.8MB psnr 41.14dB > >> > >> > >>What the heck happened? ;_; > >> > >>I suggest we postpone 1.1final until we figure this out - looks like all > >>speed improvements over 1.0 are gone (10% slower at these settings, 5% > >>slower at defaults), and 10% more filesize for *lower* quality is... bad. > >> > >>Radek > >>_______________________________________________ > >>XviD-devel mailing list > >>XviD-devel@xvid.org > >>http://list.xvid.org/mailman/listinfo/xvid-devel > >> > > > > > > > > ----- Forwarded message from Michael Militzer ----- > > Date: Mon, 14 Mar 2005 19:59:37 +0100 > > From: Michael Militzer > > Reply-To: Michael Militzer > > Subject: low-bitrate patch > > To: xvid-devel@xvid.org > > > > Hi all, > > > > I've committed a patch to cvs that improves XVID's performance at low- > > bitrates by using better lambda values for R-D vector search. By this also > > mode decision (SAD-based) is improved at low-bitrates. At mid/higher > bitrates > > there's no change in performance, so only people who aim at really low > > bitrate (and quality) encodes will benefit from this patch. > > > > I've attached some R-D graphs that compare the behavior of patched and > > unpatched XVID for various sequences and different bit-rates. Note that > I've > > used a logarithmic scale, so that the differences at low bitrates can be > > better perceived. As can be clearly noticed, the compression performance > is > > enhanced significantly at very low bitrates (up to 2-3 dB PSNR). > > > > Well, it's another question how useful this patch actually is because such > > extremely low bit-rates are barely used. But for completeness, XVID should > > perform equally great at all bitrates... > > > > bye, > > Michael > > > > ----- End forwarded message ----- > > > > > > > > > > > > ------------------------------------------------------------------------ > > > > _______________________________________________ > > XviD-devel mailing list > > XviD-devel@xvid.org > > http://list.xvid.org/mailman/listinfo/xvid-devel > _______________________________________________ > XviD-devel mailing list > XviD-devel@xvid.org > http://list.xvid.org/mailman/listinfo/xvid-devel > _______________________________________________ XviD-devel mailing list XviD-devel@xvid.org http://list.xvid.org/mailman/listinfo/xvid-devel From xvid-devel-bounces@xvid.org Fri Nov 25 17:02:54 2005 Return-Path: X-Original-To: itdp@fh-biergarten.de Delivered-To: itdp@localhost Received: from edu.bnhof.de (edu.bnhof.de [213.167.167.52]) by mail.kliche.org (Postfix) with ESMTP id EBEF5126D92 for ; Fri, 25 Nov 2005 17:02:53 +0100 (CET) Received: from edu.bnhof.de (localhost [127.0.0.1]) by edu.bnhof.de (Postfix) with ESMTP id 0FA9E18BAE; Fri, 25 Nov 2005 17:02:48 +0100 (CET) X-Original-To: xvid-devel@edu.bnhof.de Delivered-To: xvid-devel@edu.bnhof.de Received: from s6.lansco.de (s6.lansco.de [213.239.218.110]) by edu.bnhof.de (Postfix) with ESMTP id AF03C18B96 for ; Fri, 25 Nov 2005 17:02:45 +0100 (CET) Received: from s5.lansco.de (s5.lansco.de [212.63.130.170]) by s6.lansco.de (Postfix) with ESMTP id AAF0717853 for ; Fri, 25 Nov 2005 17:02:39 +0100 (CET) Received: from localhost (s5.lansco.de [212.63.130.170]) by s5.lansco.de (Postfix) with ESMTP id 911D02584A for ; Fri, 25 Nov 2005 17:02:39 +0100 (CET) Received: from p54B5F37D.dip.t-dialin.net (p54B5F37D.dip.t-dialin.net [84.181.243.125]) by www.lansco.de (IMP) with HTTP for ; Fri, 25 Nov 2005 17:02:39 +0100 Message-ID: <1132934559.4387359f82e20@www.lansco.de> Date: Fri, 25 Nov 2005 17:02:39 +0100 From: Michael Militzer To: xvid-devel@xvid.org Subject: Re: [XviD-devel] HUGE ME regressions between beta1 and beta2 References: <1132929315.4387212367133@www.lansco.de> <438731E1.9050905@syskin.cjb.net> <1132934488.4387355800550@www.lansco.de> In-Reply-To: <1132934488.4387355800550@www.lansco.de> MIME-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 8bit User-Agent: Internet Messaging Program (IMP) 3.2.3 X-Originating-IP: 84.181.243.125 X-BeenThere: xvid-devel@xvid.org X-Mailman-Version: 2.1.4 Precedence: list Reply-To: xvid-devel@xvid.org List-Id: xvid-devel.xvid.org List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Sender: xvid-devel-bounces@xvid.org Errors-To: xvid-devel-bounces@xvid.org Ok, well looking at the file sizes you report in your test it's rather likely you used another sequence than me - and likely DVD input... Michael Quoting Michael Militzer : > Hi, > > well, I'm rather confident that the current ("traditional") way finds > better vectors in ME than before, especially at high quants. What I can > imagine is that there's a larger influence on mode decision that I expected > (and experienced) in non R-D optimized mode. I've tested with lots more > sequences than just the ones I posted to the list and never had a > regression with the new code. However, I used standard testing sequences > from uncompressed input. Perhaps, mode decision with the new lambdas doesn't > work well on DVD/MPEG-2 compressed input. > > That should be easy to find out: if there's no regression with VHQ>0, then > the degradation on some sequences likely comes from mode decision. That > should be easy to fix then by applying additional weights after vector > search. Also, it may explain why the problem slipped through unnoticed - > most people use VHQ I guess. > > BTW: with what sequence do you test? With the same I used and you get > different results (that would be very concerning)? Or with another one? > > Regards, > Michael > > > Quoting Radek Czyz : > > > Hi again, > > > > Michael Militzer wrote: > > > well, I've basically just changed our lambda tables. > > > > Yes, but you also changed much more: the way lambda actually works went > > back to "traditional" way (which is in theory more correct). The lambda > > value multiplier changed from 10.5 to 0.6 (and from 40.0 to 0.6 for > > inter4v). > > > > In general, I'd expect that this kind of change has large consequences, > > in particular to inter/inter4v decisions. > > > > Overal, that was a reasonably large patch. The important thing is that I > > cannot confirm your results. This is what I get at pure defaults: > > > > Before: 1:53, 25.6MB, 41.70dB > > After: 2:03, 26.0MB, 41.78dB > > > > I am still unsure what causes the slowdown. Your patch removes one > > multiplication in important spot so I'd expect is to be faster. > > > > I've put the "old" files online, if anyone wants to test: > > http://syskin.is.dreaming.org/motion_old.zip > > > > Radek > > > > > > Before, the values for > > > lambda were borrowed from h.264 it seemed however didn't fit XviD's > linear > > > qscale at all. In result, R-D performance especially for higher QPs was > > > absolutely horrible. I've changed it to a more lambda = 0.85*Q scheme > (ok, > > > can't recall the exact constant from memory now). I've tested the change > > > for VHQ and non-VHQ modes and had always superior performance, sometimes > > > > with really huge gains especially a low bit-rates/high quants. > > > > > > I've attached my original mail to the list at the end of this mails and > > also > > > reposted some of the R-D curves I did at those times. Your testing case > is > > > covered as well (it seems). > > > > > > Sure we can postpone 1.1 final until this problem is fixed. BTW: did you > > use > > > a standard sequence for your test so we can reproduce the problem? > > > > > > Regards, > > > Michael > > > > > > > > > Quoting Radek Czyz : > > > > > > > > >>Hi everyone, > > >> > > >>Apparently our testing community sucks ;)) > > >> > > >>There was a HUGE quality and speed regression between beta1 and beta2. I > > > >>narrowed it down to SAD-based ME. The major changes in that time was > > >>Isibaar's new lambda code, but apparently we have hardly any logs what > > >>happened exactly (well, we've got cvs, I'll check that). > > >> > > >>For fast settings (no VHQ, no trellis, no chromaME, defaults otherwise): > > >> > > >>Beta1's ME: time 1:19 filesize 26.2MB psnr 41.19dB > > >>Current ME: time 1:24 filesize 28.8MB psnr 41.14dB > > >> > > >> > > >>What the heck happened? ;_; > > >> > > >>I suggest we postpone 1.1final until we figure this out - looks like all > > > >>speed improvements over 1.0 are gone (10% slower at these settings, 5% > > >>slower at defaults), and 10% more filesize for *lower* quality is... > bad. > > >> > > >>Radek > > >>_______________________________________________ > > >>XviD-devel mailing list > > >>XviD-devel@xvid.org > > >>http://list.xvid.org/mailman/listinfo/xvid-devel > > >> > > > > > > > > > > > > ----- Forwarded message from Michael Militzer ----- > > > Date: Mon, 14 Mar 2005 19:59:37 +0100 > > > From: Michael Militzer > > > Reply-To: Michael Militzer > > > Subject: low-bitrate patch > > > To: xvid-devel@xvid.org > > > > > > Hi all, > > > > > > I've committed a patch to cvs that improves XVID's performance at low- > > > bitrates by using better lambda values for R-D vector search. By this > also > > > mode decision (SAD-based) is improved at low-bitrates. At mid/higher > > bitrates > > > there's no change in performance, so only people who aim at really low > > > bitrate (and quality) encodes will benefit from this patch. > > > > > > I've attached some R-D graphs that compare the behavior of patched and > > > unpatched XVID for various sequences and different bit-rates. Note that > > I've > > > used a logarithmic scale, so that the differences at low bitrates can be > > > better perceived. As can be clearly noticed, the compression performance > > is > > > enhanced significantly at very low bitrates (up to 2-3 dB PSNR). > > > > > > Well, it's another question how useful this patch actually is because > such > > > extremely low bit-rates are barely used. But for completeness, XVID > should > > > perform equally great at all bitrates... > > > > > > bye, > > > Michael > > > > > > ----- End forwarded message ----- > > > > > > > > > > > > > > > > > > ------------------------------------------------------------------------ > > > > > > _______________________________________________ > > > XviD-devel mailing list > > > XviD-devel@xvid.org > > > http://list.xvid.org/mailman/listinfo/xvid-devel > > _______________________________________________ > > XviD-devel mailing list > > XviD-devel@xvid.org > > http://list.xvid.org/mailman/listinfo/xvid-devel > > > > > > > _______________________________________________ > XviD-devel mailing list > XviD-devel@xvid.org > http://list.xvid.org/mailman/listinfo/xvid-devel > _______________________________________________ XviD-devel mailing list XviD-devel@xvid.org http://list.xvid.org/mailman/listinfo/xvid-devel From xvid-devel-bounces@xvid.org Fri Nov 25 17:28:23 2005 Return-Path: X-Original-To: itdp@fh-biergarten.de Delivered-To: itdp@localhost Received: from edu.bnhof.de (edu.bnhof.de [213.167.167.52]) by mail.kliche.org (Postfix) with ESMTP id 91FD4126D92 for ; Fri, 25 Nov 2005 17:28:23 +0100 (CET) Received: from edu.bnhof.de (localhost [127.0.0.1]) by edu.bnhof.de (Postfix) with ESMTP id A479318BB9; Fri, 25 Nov 2005 17:28:16 +0100 (CET) X-Original-To: xvid-devel@edu.bnhof.de Delivered-To: xvid-devel@edu.bnhof.de Received: from s6.lansco.de (s6.lansco.de [213.239.218.110]) by edu.bnhof.de (Postfix) with ESMTP id 39B7D18B7C for ; Fri, 25 Nov 2005 17:27:48 +0100 (CET) Received: from mail.skviril.net (unknown [84.234.141.4]) by s6.lansco.de (Postfix) with ESMTP id 2FC8517542 for ; Fri, 25 Nov 2005 17:27:42 +0100 (CET) Received: from localhost (localhost.localdomain [127.0.0.1]) by mail.skviril.net (Postfix) with ESMTP id 11CC15000084 for ; Fri, 25 Nov 2005 17:27:52 +0100 (CET) Received: from mail.skviril.net ([127.0.0.1]) by localhost (ryo-ohki.4th-age.com [127.0.0.1]) (amavisd-new, port 10024) with ESMTP id 14617-04 for ; Fri, 25 Nov 2005 17:27:51 +0100 (CET) Received: from [10.0.0.3] (tsunami.lan.4th-age.com [10.0.0.3]) by mail.skviril.net (Postfix) with ESMTP id AFA375000083 for ; Fri, 25 Nov 2005 17:27:51 +0100 (CET) Message-ID: <43873B7D.4080707@nilsen.com> Date: Fri, 25 Nov 2005 17:27:41 +0100 From: Morten Nilsen User-Agent: Mozilla/5.0 (X11; U; Linux i686; en-US; rv:1.7.12) Gecko/20050920 Mnenhy/0.7.2.0 X-Accept-Language: en-us, en MIME-Version: 1.0 To: xvid-devel@xvid.org Subject: Re: [XviD-devel] HUGE ME regressions between beta1 and beta2 References: <1132929315.4387212367133@www.lansco.de> <1132929539.43872203572e6@www.lansco.de> In-Reply-To: <1132929539.43872203572e6@www.lansco.de> Content-Type: text/plain; charset=us-ascii; format=flowed Content-Transfer-Encoding: 7bit X-Virus-Scanned: amavisd-new at skviril.net X-BeenThere: xvid-devel@xvid.org X-Mailman-Version: 2.1.4 Precedence: list Reply-To: xvid-devel@xvid.org List-Id: xvid-devel.xvid.org List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Sender: xvid-devel-bounces@xvid.org Errors-To: xvid-devel-bounces@xvid.org Michael Militzer wrote: > As always, attachments got stripped. So have a look here: > > http://www.xvid.org/images/carphone.png > http://www.xvid.org/images/highway.png > http://www.xvid.org/images/mother.png > http://www.xvid.org/images/paris.png > http://www.xvid.org/images/tempete.png I am unable to reach xvid.org .. connection times out a traceroute stops at: 20 et-1.16.RS3K7.RZ6.hetzner.de (213.239.239.200) 52.180 ms 51.898 ms 52.423 ms someone using a different ISP had no problems.. is hetzner null-routing my net (84.234.141.0/24) ? -- Morten _______________________________________________ XviD-devel mailing list XviD-devel@xvid.org http://list.xvid.org/mailman/listinfo/xvid-devel From xvid-devel-bounces@xvid.org Fri Nov 25 17:42:35 2005 Return-Path: X-Original-To: itdp@fh-biergarten.de Delivered-To: itdp@localhost Received: from edu.bnhof.de (edu.bnhof.de [213.167.167.52]) by mail.kliche.org (Postfix) with ESMTP id 22F94126D8B for ; Fri, 25 Nov 2005 17:42:35 +0100 (CET) Received: from edu.bnhof.de (localhost [127.0.0.1]) by edu.bnhof.de (Postfix) with ESMTP id 91E1218B9C; Fri, 25 Nov 2005 17:42:31 +0100 (CET) X-Original-To: xvid-devel@edu.bnhof.de Delivered-To: xvid-devel@edu.bnhof.de Received: from s6.lansco.de (s6.lansco.de [213.239.218.110]) by edu.bnhof.de (Postfix) with ESMTP id 2E7BB18B73 for ; Fri, 25 Nov 2005 17:42:29 +0100 (CET) Received: from mail-ihug.icp-qv1-irony2.iinet.net.au (ihug-mail.icp-qv1-irony2.iinet.net.au [203.59.1.196]) by s6.lansco.de (Postfix) with ESMTP id 58159176EE for ; Fri, 25 Nov 2005 17:42:26 +0100 (CET) Received: from 203-173-56-106.dyn.iinet.net.au (HELO [10.1.1.3]) ([203.173.56.106]) by mail-ihug.icp-qv1-irony2.iinet.net.au with ESMTP; 26 Nov 2005 00:42:23 +0800 X-BrightmailFiltered: true X-Brightmail-Tracker: AAAAAA== Message-ID: <43873F16.30604@syskin.cjb.net> Date: Sat, 26 Nov 2005 03:13:02 +1030 From: Radek Czyz User-Agent: Mozilla Thunderbird 1.0.2 (Windows/20050317) X-Accept-Language: en-us, en MIME-Version: 1.0 To: xvid-devel@xvid.org Subject: Re: [XviD-devel] HUGE ME regressions between beta1 and beta2 References: <1132929315.4387212367133@www.lansco.de> <438731E1.9050905@syskin.cjb.net> <1132934488.4387355800550@www.lansco.de> <1132934559.4387359f82e20@www.lansco.de> In-Reply-To: <1132934559.4387359f82e20@www.lansco.de> Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 7bit X-BeenThere: xvid-devel@xvid.org X-Mailman-Version: 2.1.4 Precedence: list Reply-To: xvid-devel@xvid.org List-Id: xvid-devel.xvid.org List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Sender: xvid-devel-bounces@xvid.org Errors-To: xvid-devel-bounces@xvid.org Michael Militzer wrote: > Ok, well looking at the file sizes you report in your test it's rather > likely you used another sequence than me - and likely DVD input... Yes it's one of my dvd sequences I keep (actually it's huffyuv, too big to upload). I'll test with bvop-vhq and see if the filesize difference I get is purely because of mode decision. What about speed though? I see 5% slower at defaults, but people at Doom9 forum are reporting up to 23% slower second pass. I still don't even know where it comes from. Radek _______________________________________________ XviD-devel mailing list XviD-devel@xvid.org http://list.xvid.org/mailman/listinfo/xvid-devel From xvid-devel-bounces@xvid.org Fri Nov 25 17:56:22 2005 Return-Path: X-Original-To: itdp@fh-biergarten.de Delivered-To: itdp@localhost Received: from edu.bnhof.de (edu.bnhof.de [213.167.167.52]) by mail.kliche.org (Postfix) with ESMTP id DD609126D8B for ; Fri, 25 Nov 2005 17:56:22 +0100 (CET) Received: from edu.bnhof.de (localhost [127.0.0.1]) by edu.bnhof.de (Postfix) with ESMTP id 82B3B18BCC; Fri, 25 Nov 2005 17:56:18 +0100 (CET) X-Original-To: xvid-devel@edu.bnhof.de Delivered-To: xvid-devel@edu.bnhof.de Received: from s6.lansco.de (s6.lansco.de [213.239.218.110]) by edu.bnhof.de (Postfix) with ESMTP id 377D018BC4 for ; Fri, 25 Nov 2005 17:56:16 +0100 (CET) Received: from smtp6.wanadoo.fr (smtp6.wanadoo.fr [193.252.22.25]) by s6.lansco.de (Postfix) with ESMTP id 255AB177BB for ; Fri, 25 Nov 2005 17:56:15 +0100 (CET) Received: from me-wanadoo.net (localhost [127.0.0.1]) by mwinf0607.wanadoo.fr (SMTP Server) with ESMTP id D66031C001B4 for ; Fri, 25 Nov 2005 17:56:09 +0100 (CET) Received: from localhost.localdomain (ASte-Genev-Bois-151-1-63-245.w82-121.abo.wanadoo.fr [82.121.7.245]) by mwinf0607.wanadoo.fr (SMTP Server) with ESMTP id 8D3C51C0016A for ; Fri, 25 Nov 2005 17:56:09 +0100 (CET) X-ME-UUID: 20051125165609578.8D3C51C0016A@mwinf0607.wanadoo.fr Subject: Re: [XviD-devel] HUGE ME regressions between beta1 and beta2 From: Pascal Massimino To: xvid-devel@xvid.org In-Reply-To: <43873F16.30604@syskin.cjb.net> References: <1132929315.4387212367133@www.lansco.de> <438731E1.9050905@syskin.cjb.net> <1132934488.4387355800550@www.lansco.de> <1132934559.4387359f82e20@www.lansco.de> <43873F16.30604@syskin.cjb.net> Content-Type: text/plain Message-Id: <1132937693.5066.6.camel@localhost.localdomain> Mime-Version: 1.0 X-Mailer: Ximian Evolution 1.4.5 (1.4.5-14) Date: Fri, 25 Nov 2005 17:54:53 +0100 Content-Transfer-Encoding: 7bit X-BeenThere: xvid-devel@xvid.org X-Mailman-Version: 2.1.4 Precedence: list Reply-To: skal@planet-d.net, xvid-devel@xvid.org List-Id: xvid-devel.xvid.org List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Sender: xvid-devel-bounces@xvid.org Errors-To: xvid-devel-bounces@xvid.org Hi Radek and all, On Fri, 2005-11-25 at 17:43, Radek Czyz wrote: > What about speed though? I see 5% slower at defaults, but people at > Doom9 forum are reporting up to 23% slower second pass. > > I still don't even know where it comes from. > Don't forget that the speed scales inversely to bitrate, too. If you experience a higher file size, then it's no real surprise the encoding speed is a little lower. If just to process the additional data. Now, it sure is not the definitive explanation, just a part of, i guess... bye! Skal _______________________________________________ XviD-devel mailing list XviD-devel@xvid.org http://list.xvid.org/mailman/listinfo/xvid-devel From xvid-devel-bounces@xvid.org Sat Nov 26 04:38:45 2005 Return-Path: X-Original-To: itdp@fh-biergarten.de Delivered-To: itdp@localhost Received: from edu.bnhof.de (edu.bnhof.de [213.167.167.52]) by mail.kliche.org (Postfix) with ESMTP id 4699C126D83 for ; Sat, 26 Nov 2005 04:38:45 +0100 (CET) Received: from edu.bnhof.de (localhost [127.0.0.1]) by edu.bnhof.de (Postfix) with ESMTP id 3C96618B9D; Sat, 26 Nov 2005 04:38:41 +0100 (CET) X-Original-To: xvid-devel@edu.bnhof.de Delivered-To: xvid-devel@edu.bnhof.de Received: from s6.lansco.de (s6.lansco.de [213.239.218.110]) by edu.bnhof.de (Postfix) with ESMTP id C33E018B8A for ; Sat, 26 Nov 2005 04:38:38 +0100 (CET) Received: from mail-ihug.icp-qv1-irony3.iinet.net.au (ihug-mail.icp-qv1-irony3.iinet.net.au [203.59.1.197]) by s6.lansco.de (Postfix) with ESMTP id C006A1C708 for ; Sat, 26 Nov 2005 04:38:36 +0100 (CET) Received: from 203-173-56-106.dyn.iinet.net.au (HELO [10.1.1.3]) ([203.173.56.106]) by mail-ihug.icp-qv1-irony3.iinet.net.au with ESMTP; 26 Nov 2005 11:38:33 +0800 X-BrightmailFiltered: true X-Brightmail-Tracker: AAAAAA== Message-ID: <4387D8E7.9080901@syskin.cjb.net> Date: Sat, 26 Nov 2005 14:09:19 +1030 From: Radek Czyz User-Agent: Mozilla Thunderbird 1.0.2 (Windows/20050317) X-Accept-Language: en-us, en MIME-Version: 1.0 To: xvid-devel@xvid.org Subject: Re: [XviD-devel] HUGE ME regressions between beta1 and beta2 References: <1132929315.4387212367133@www.lansco.de> <438731E1.9050905@syskin.cjb.net> <1132934488.4387355800550@www.lansco.de> In-Reply-To: <1132934488.4387355800550@www.lansco.de> Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 7bit X-BeenThere: xvid-devel@xvid.org X-Mailman-Version: 2.1.4 Precedence: list Reply-To: xvid-devel@xvid.org List-Id: xvid-devel.xvid.org List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Sender: xvid-devel-bounces@xvid.org Errors-To: xvid-devel-bounces@xvid.org Hi, Michael Militzer wrote: > That should be easy to find out: if there's no regression with VHQ>0, then > the degradation on some sequences likely comes from mode decision. That > should be easy to fix then by applying additional weights after vector > search. Also, it may explain why the problem slipped through unnoticed - > most people use VHQ I guess. Yes, I just confirmed that at defaults (with VHQ1), the quality is a bit higher accross the board. It doesn't exactly look like it because filesize is bigger, but apparently PSNR offsets it. I get the same graphs you got. I will now try to tweak all constants for best PSNR with VHQ0. I will also try to find why there's such a big slowdown - Skal's theory might be correct, and in that case we're doomed - we can't get this 5..10% speed back. I have a hope that explaination is different and some early-skip thresholds no longer work correctly, with different levels of total SAD. If so, we can return to old speed (hopefully still keeping new psnr). Radek _______________________________________________ XviD-devel mailing list XviD-devel@xvid.org http://list.xvid.org/mailman/listinfo/xvid-devel From xvid-devel-bounces@xvid.org Sat Nov 26 07:49:23 2005 Return-Path: X-Original-To: itdp@fh-biergarten.de Delivered-To: itdp@localhost Received: from edu.bnhof.de (edu.bnhof.de [213.167.167.52]) by mail.kliche.org (Postfix) with ESMTP id 4FD5B126D83 for ; Sat, 26 Nov 2005 07:49:23 +0100 (CET) Received: from edu.bnhof.de (localhost [127.0.0.1]) by edu.bnhof.de (Postfix) with ESMTP id C63E918C22; Sat, 26 Nov 2005 07:49:19 +0100 (CET) X-Original-To: xvid-devel@edu.bnhof.de Delivered-To: xvid-devel@edu.bnhof.de Received: from s6.lansco.de (s6.lansco.de [213.239.218.110]) by edu.bnhof.de (Postfix) with ESMTP id A1A5D18C1F for ; Sat, 26 Nov 2005 07:49:17 +0100 (CET) Received: from customer-domains.icp-qv1-irony7.iinet.net.au (customer-domains.icp-qv1-irony7.iinet.net.au [203.59.1.128]) by s6.lansco.de (Postfix) with ESMTP id 97F1D1C7FC for ; Sat, 26 Nov 2005 07:49:14 +0100 (CET) Received: from 203-173-56-106.dyn.iinet.net.au (HELO [10.1.1.3]) ([203.173.56.106]) by customer-domains.icp-qv1-irony7.iinet.net.au with ESMTP; 26 Nov 2005 14:49:12 +0800 X-BrightmailFiltered: true X-Brightmail-Tracker: AAAAAA== Message-ID: <43880598.7010605@syskin.cjb.net> Date: Sat, 26 Nov 2005 17:20:00 +1030 From: Radek Czyz User-Agent: Mozilla Thunderbird 1.0.2 (Windows/20050317) X-Accept-Language: en-us, en MIME-Version: 1.0 To: xvid-devel@xvid.org Subject: Re: [XviD-devel] HUGE ME regressions between beta1 and beta2 References: <1132929315.4387212367133@www.lansco.de> <438731E1.9050905@syskin.cjb.net> <1132934488.4387355800550@www.lansco.de> <1132934559.4387359f82e20@www.lansco.de> In-Reply-To: <1132934559.4387359f82e20@www.lansco.de> Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 7bit X-BeenThere: xvid-devel@xvid.org X-Mailman-Version: 2.1.4 Precedence: list Reply-To: xvid-devel@xvid.org List-Id: xvid-devel.xvid.org List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Sender: xvid-devel-bounces@xvid.org Errors-To: xvid-devel-bounces@xvid.org Hi again, Michael, I have a question: the lambda tables for 16x16 search are different than for 8x8 search. For 8x8, they are linear: 1, 2, 3, 4, 5, 6, 7, 8, 9... but for 16x16, there's a kink at low quants: 0.5, 1, 1.5, 2, 2.5, 5, 7, 8, 9 ...... Why's that? I'm currently getting much better results for linear tables everywhere. In other news, The slowdown is caused by all motion searches checking many more vectors than before. For some frames, twice as many. Apparenty diamonds like to wander far away now. The filesize difference at fixed quant is caused almost entirely by b-frames being much larger than before. P-frames were same size, and are smaller now with linear tables. Regards, Radek _______________________________________________ XviD-devel mailing list XviD-devel@xvid.org http://list.xvid.org/mailman/listinfo/xvid-devel From xvid-devel-bounces@xvid.org Sat Nov 26 11:34:21 2005 Return-Path: X-Original-To: itdp@fh-biergarten.de Delivered-To: itdp@localhost Received: from edu.bnhof.de (edu.bnhof.de [213.167.167.52]) by mail.kliche.org (Postfix) with ESMTP id 24896126D8B for ; Sat, 26 Nov 2005 11:34:21 +0100 (CET) Received: from edu.bnhof.de (localhost [127.0.0.1]) by edu.bnhof.de (Postfix) with ESMTP id 0D294198AC; Sat, 26 Nov 2005 11:34:17 +0100 (CET) X-Original-To: xvid-devel@edu.bnhof.de Delivered-To: xvid-devel@edu.bnhof.de Received: from s6.lansco.de (s6.lansco.de [213.239.218.110]) by edu.bnhof.de (Postfix) with ESMTP id 04E3B198A7 for ; Sat, 26 Nov 2005 11:34:15 +0100 (CET) Received: from s5.lansco.de (s5.lansco.de [212.63.130.170]) by s6.lansco.de (Postfix) with ESMTP id F3BD01C89D for ; Sat, 26 Nov 2005 11:34:13 +0100 (CET) Received: from localhost (s5.lansco.de [212.63.130.170]) by s5.lansco.de (Postfix) with ESMTP id CF0B9306DC for ; Sat, 26 Nov 2005 11:34:13 +0100 (CET) Received: from p54B5B7E9.dip.t-dialin.net (p54B5B7E9.dip.t-dialin.net [84.181.183.233]) by www.lansco.de (IMP) with HTTP for ; Sat, 26 Nov 2005 11:34:13 +0100 Message-ID: <1133001253.43883a25b075c@www.lansco.de> Date: Sat, 26 Nov 2005 11:34:13 +0100 From: Michael Militzer To: xvid-devel@xvid.org Subject: Re: [XviD-devel] HUGE ME regressions between beta1 and beta2 References: <1132929315.4387212367133@www.lansco.de> <438731E1.9050905@syskin.cjb.net> <1132934488.4387355800550@www.lansco.de> <1132934559.4387359f82e20@www.lansco.de> <43880598.7010605@syskin.cjb.net> In-Reply-To: <43880598.7010605@syskin.cjb.net> MIME-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 8bit User-Agent: Internet Messaging Program (IMP) 3.2.3 X-Originating-IP: 84.181.183.233 X-BeenThere: xvid-devel@xvid.org X-Mailman-Version: 2.1.4 Precedence: list Reply-To: xvid-devel@xvid.org List-Id: xvid-devel.xvid.org List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Sender: xvid-devel-bounces@xvid.org Errors-To: xvid-devel-bounces@xvid.org Hi Radek, Quoting Radek Czyz : > Michael, I have a question: the lambda tables for 16x16 search are > different than for 8x8 search. For 8x8, they are linear: > 1, 2, 3, 4, 5, 6, 7, 8, 9... > > but for 16x16, there's a kink at low quants: > > 0.5, 1, 1.5, 2, 2.5, 5, 7, 8, 9 ...... > > Why's that? > I'm currently getting much better results for linear tables everywhere. Well, the idea was: imho, the lambda*MV_bits cost function bears the danger that lowers perceived quality even though it increases PSNR. That's because PSNR is a metric that does not consider temporal artifacts. For example, the higher the lambda value you use the higher the danger for smearing artifacts. Try some high lambda and watch moving objects leaving trails even though it doesn't hurt R-D performance (because the PSNR reduction is compensated by bit gain). Again imho, temporal artifacts are minimized (and subjective quality with this regard maximized) when your ME finds vectors that correlate with the true motion in the sequence. In this case, MBs will temporally predict from those past areas the human eye expect it. So in theory, lambda=0 should give best results with this regard. However, of course the R-D optimized vector search helps a lot, so we cannot go with lambda=0. But certainly, a lower lambda should result in less smearing or trails, so if we can afford it we should go for it. And for low quants we typically have large textual residuals, which take up most of the output file size. So reducing the MV bits is less a consideration - simply because MV bits will take up just a minor fraction of the total size. So a smaller lambda should hurt R-D performance just very mildly while providing the better visual quality. That was the idea ;) > In other news, > > The slowdown is caused by all motion searches checking many more vectors > than before. For some frames, twice as many. Apparenty diamonds like to > wander far away now. That basically also shouldn't be the case at low quants. At higher quants, yes sure, as the old lambda tables were introducing such a huge cost penalty that each MB was pinned at MV (0,0). Certainly faster but resulted in horrible quality at higher quants. Perhaps, one of the early stop criterias don't work properly anymore. E.g. the early stop to prevent extensive 8x8 search. I've just seen that these early stops use fixed thresholds. Imho, they should be adaptive to be more robust... > The filesize difference at fixed quant is caused almost entirely by > b-frames being much larger than before. P-frames were same size, and are > smaller now with linear tables. Actually, there shouldn't be large file size differences. If you look at the graphs I plotted, the sample points of before and after the lambda change are rather close at low quants. With the new linear lambda tables just performing a bit better. Regards, Michael _______________________________________________ XviD-devel mailing list XviD-devel@xvid.org http://list.xvid.org/mailman/listinfo/xvid-devel From xvid-devel-bounces@xvid.org Sat Nov 26 12:05:13 2005 Return-Path: X-Original-To: itdp@fh-biergarten.de Delivered-To: itdp@localhost Received: from edu.bnhof.de (edu.bnhof.de [213.167.167.52]) by mail.kliche.org (Postfix) with ESMTP id 1FB84126D83 for ; Sat, 26 Nov 2005 12:05:13 +0100 (CET) Received: from edu.bnhof.de (localhost [127.0.0.1]) by edu.bnhof.de (Postfix) with ESMTP id 277B4198BD; Sat, 26 Nov 2005 12:05:10 +0100 (CET) X-Original-To: xvid-devel@edu.bnhof.de Delivered-To: xvid-devel@edu.bnhof.de Received: from s6.lansco.de (s6.lansco.de [213.239.218.110]) by edu.bnhof.de (Postfix) with ESMTP id 339D0198BA for ; Sat, 26 Nov 2005 12:05:08 +0100 (CET) Received: from mail-ihug.icp-qv1-irony5.iinet.net.au (ihug-mail.icp-qv1-irony5.iinet.net.au [203.59.1.199]) by s6.lansco.de (Postfix) with ESMTP id 565E51C8D1 for ; Sat, 26 Nov 2005 12:05:04 +0100 (CET) Received: from 203-173-56-106.dyn.iinet.net.au (HELO [10.1.1.3]) ([203.173.56.106]) by mail-ihug.icp-qv1-irony5.iinet.net.au with ESMTP; 26 Nov 2005 19:05:01 +0800 X-BrightmailFiltered: true X-Brightmail-Tracker: AAAAAA== Message-ID: <43884191.3030502@syskin.cjb.net> Date: Sat, 26 Nov 2005 21:35:53 +1030 From: Radek Czyz User-Agent: Mozilla Thunderbird 1.0.2 (Windows/20050317) X-Accept-Language: en-us, en MIME-Version: 1.0 To: xvid-devel@xvid.org Subject: Re: [XviD-devel] HUGE ME regressions between beta1 and beta2 References: <1132929315.4387212367133@www.lansco.de> <438731E1.9050905@syskin.cjb.net> <1132934488.4387355800550@www.lansco.de> <1132934559.4387359f82e20@www.lansco.de> <43880598.7010605@syskin.cjb.net> <1133001253.43883a25b075c@www.lansco.de> In-Reply-To: <1133001253.43883a25b075c@www.lansco.de> Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 7bit X-BeenThere: xvid-devel@xvid.org X-Mailman-Version: 2.1.4 Precedence: list Reply-To: xvid-devel@xvid.org List-Id: xvid-devel.xvid.org List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Sender: xvid-devel-bounces@xvid.org Errors-To: xvid-devel-bounces@xvid.org Michael Militzer wrote: > Well, the idea was: imho, the lambda*MV_bits cost function bears the danger > that lowers perceived quality even though it increases PSNR. (...) OK that's a good theory. However, isn't "moving walls" the current problem with xvid? Static trails are divx3 problem but we didn't have them since forever. By popular vote, robustness of MVs to noise was one of my goals for 1.2 development. > That basically also shouldn't be the case at low quants. At higher quants, > yes sure, as the old lambda tables were introducing such a huge cost penalty > that each MB was pinned at MV (0,0). Certainly faster but resulted in > horrible quality at higher quants. Yeah I'm puzzled by the slowdown. It doesn't seem to be related to the lambda values directly - if I double lambda, speed is still pretty much the same. A solid 5% loss, corresponding to ~20% more candidate checks. The candidate checks happen because fewer predictors repeat and because diamonds take more iterations to complete (which partially contradicts btw.) > Perhaps, one of the early stop criterias don't work properly anymore. E.g. > the early stop to prevent extensive 8x8 search. I've just seen that these > early stops use fixed thresholds. Imho, they should be adaptive to be more > robust... I turned them off for my investigation. It's not it. > Actually, there shouldn't be large file size differences. If you look at > the graphs I plotted, the sample points of before and after the lambda > change are rather close at low quants. With the new linear lambda tables > just performing a bit better. Yes the differences are 1% - 2% of total size so you can't see them on graphs. But if you'd look at first pass statistics file, every single b-frame is larger with new code. It's non-texture data that is larger, but boosting lambda for b-frames seems to have little effect. It might be the skip (direct_none_mv) decision. Regards, Radek _______________________________________________ XviD-devel mailing list XviD-devel@xvid.org http://list.xvid.org/mailman/listinfo/xvid-devel From xvid-devel-bounces@xvid.org Sat Nov 26 12:30:44 2005 Return-Path: X-Original-To: itdp@fh-biergarten.de Delivered-To: itdp@localhost Received: from edu.bnhof.de (edu.bnhof.de [213.167.167.52]) by mail.kliche.org (Postfix) with ESMTP id B2F16126D83 for ; Sat, 26 Nov 2005 12:30:44 +0100 (CET) Received: from edu.bnhof.de (localhost [127.0.0.1]) by edu.bnhof.de (Postfix) with ESMTP id 8F53C198CC; Sat, 26 Nov 2005 12:30:41 +0100 (CET) X-Original-To: xvid-devel@edu.bnhof.de Delivered-To: xvid-devel@edu.bnhof.de Received: from s6.lansco.de (s6.lansco.de [213.239.218.110]) by edu.bnhof.de (Postfix) with ESMTP id E18BB198C9 for ; Sat, 26 Nov 2005 12:30:38 +0100 (CET) Received: from s5.lansco.de (s5.lansco.de [212.63.130.170]) by s6.lansco.de (Postfix) with ESMTP id E264D1C8EE for ; Sat, 26 Nov 2005 12:30:37 +0100 (CET) Received: from localhost (s5.lansco.de [212.63.130.170]) by s5.lansco.de (Postfix) with ESMTP id CE00130738 for ; Sat, 26 Nov 2005 12:30:37 +0100 (CET) Received: from p54B5B7E9.dip.t-dialin.net (p54B5B7E9.dip.t-dialin.net [84.181.183.233]) by www.lansco.de (IMP) with HTTP for ; Sat, 26 Nov 2005 12:30:37 +0100 Message-ID: <1133004637.4388475da551b@www.lansco.de> Date: Sat, 26 Nov 2005 12:30:37 +0100 From: Michael Militzer To: xvid-devel@xvid.org Subject: Re: [XviD-devel] HUGE ME regressions between beta1 and beta2 References: <1132929315.4387212367133@www.lansco.de> <438731E1.9050905@syskin.cjb.net> <1132934488.4387355800550@www.lansco.de> <1132934559.4387359f82e20@www.lansco.de> <43880598.7010605@syskin.cjb.net> <1133001253.43883a25b075c@www.lansco.de> <43884191.3030502@syskin.cjb.net> In-Reply-To: <43884191.3030502@syskin.cjb.net> MIME-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 8bit User-Agent: Internet Messaging Program (IMP) 3.2.3 X-Originating-IP: 84.181.183.233 X-BeenThere: xvid-devel@xvid.org X-Mailman-Version: 2.1.4 Precedence: list Reply-To: xvid-devel@xvid.org List-Id: xvid-devel.xvid.org List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Sender: xvid-devel-bounces@xvid.org Errors-To: xvid-devel-bounces@xvid.org Hi, Quoting Radek Czyz : > Michael Militzer wrote: > > > Well, the idea was: imho, the lambda*MV_bits cost function bears the > danger > > that lowers perceived quality even though it increases PSNR. (...) > > OK that's a good theory. However, isn't "moving walls" the current > problem with xvid? Static trails are divx3 problem but we didn't have > them since forever. By popular vote, robustness of MVs to noise was one > of my goals for 1.2 development. Yes, certainly moving walls are a big problem. But I wouldn't say that trails are a problem that cannot appear in XviD. E.g. with the old lambda tables, trails were very apparent starting from QP=8 and higher. Just that few people used higher quants and you don't notice in B-frames (due to 90% direct usually). Also, if you want to give it a try: double or triple the current lambda and trails will appear ;) So the difficulty of removing the "wobbling walls" will be to use (0,0) on these really static MBs but not favour (0,0) (or short vectors) in general so much that trails and smearing will appear. > > That basically also shouldn't be the case at low quants. At higher quants, > > yes sure, as the old lambda tables were introducing such a huge cost > penalty > > that each MB was pinned at MV (0,0). Certainly faster but resulted in > > horrible quality at higher quants. > > Yeah I'm puzzled by the slowdown. It doesn't seem to be related to the > lambda values directly - if I double lambda, speed is still pretty much > the same. A solid 5% loss, corresponding to ~20% more candidate checks. > > The candidate checks happen because fewer predictors repeat and because > diamonds take more iterations to complete (which partially contradicts btw.) Hm, that's strange. Less predictors repating is actually a good thing - also I wouldn't expect a huge slow-down from few predictors more. If the diamond doesn't terminate that may have a larger influence. But still it's strange: after all, I believe that the MV penalty isn't much different with the new lambda table than with the old one for QP<=6, say. So there shouldn't be much difference in the search - regarding speed and quality. Which version do you use for speed comparison btw? One directly before the lambda patch applied and the other one with the new lambdas? So is the speed loss really only due to the new lambdas or could there also be other reasons? > > Perhaps, one of the early stop criterias don't work properly anymore. E.g. > > the early stop to prevent extensive 8x8 search. I've just seen that these > > early stops use fixed thresholds. Imho, they should be adaptive to be more > > robust... > > I turned them off for my investigation. It's not it. Hm, the vector search shouldn't be so much different at low quants actually. So that's strange... > > Actually, there shouldn't be large file size differences. If you look at > > the graphs I plotted, the sample points of before and after the lambda > > change are rather close at low quants. With the new linear lambda tables > > just performing a bit better. > > Yes the differences are 1% - 2% of total size so you can't see them on > graphs. But if you'd look at first pass statistics file, every single > b-frame is larger with new code. It's non-texture data that is larger, > but boosting lambda for b-frames seems to have little effect. > It might be the skip (direct_none_mv) decision. Well, you would expect that the new MV bit cost penalty added is smaller than with old lambdas - at least that's what I'd conclude from your report that diamonds don't terminate as early anymore. In this case, it should even be easier for a MB to fulfill the B_SKIP criterion. So it seems odd then that B-frame header/MV data require more bits... Michael _______________________________________________ XviD-devel mailing list XviD-devel@xvid.org http://list.xvid.org/mailman/listinfo/xvid-devel From xvid-devel-bounces@xvid.org Sat Nov 26 13:41:30 2005 Return-Path: X-Original-To: itdp@fh-biergarten.de Delivered-To: itdp@localhost Received: from edu.bnhof.de (edu.bnhof.de [213.167.167.52]) by mail.kliche.org (Postfix) with ESMTP id 988AE126D8B for ; Sat, 26 Nov 2005 13:41:30 +0100 (CET) Received: from edu.bnhof.de (localhost [127.0.0.1]) by edu.bnhof.de (Postfix) with ESMTP id 01CB1198D6; Sat, 26 Nov 2005 13:41:28 +0100 (CET) X-Original-To: xvid-devel@edu.bnhof.de Delivered-To: xvid-devel@edu.bnhof.de Received: from s6.lansco.de (s6.lansco.de [213.239.218.110]) by edu.bnhof.de (Postfix) with ESMTP id 51C21198D3 for ; Sat, 26 Nov 2005 13:41:25 +0100 (CET) Received: from mail-ihug.icp-qv1-irony5.iinet.net.au (ihug-mail.icp-qv1-irony5.iinet.net.au [203.59.1.199]) by s6.lansco.de (Postfix) with ESMTP id 6E6A91C75C for ; Sat, 26 Nov 2005 13:41:23 +0100 (CET) Received: from 203-173-56-106.dyn.iinet.net.au (HELO [10.1.1.3]) ([203.173.56.106]) by mail-ihug.icp-qv1-irony5.iinet.net.au with ESMTP; 26 Nov 2005 20:41:16 +0800 X-BrightmailFiltered: true X-Brightmail-Tracker: AAAAAA== Message-ID: <43885821.801@syskin.cjb.net> Date: Sat, 26 Nov 2005 23:12:09 +1030 From: Radek Czyz User-Agent: Mozilla Thunderbird 1.0.2 (Windows/20050317) X-Accept-Language: en-us, en MIME-Version: 1.0 To: xvid-devel@xvid.org Subject: Re: [XviD-devel] HUGE ME regressions between beta1 and beta2 References: <1132929315.4387212367133@www.lansco.de> <438731E1.9050905@syskin.cjb.net> <1132934488.4387355800550@www.lansco.de> <1132934559.4387359f82e20@www.lansco.de> <43880598.7010605@syskin.cjb.net> <1133001253.43883a25b075c@www.lansco.de> <43884191.3030502@syskin.cjb.net> <1133004637.4388475da551b@www.lansco.de> In-Reply-To: <1133004637.4388475da551b@www.lansco.de> Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 7bit X-BeenThere: xvid-devel@xvid.org X-Mailman-Version: 2.1.4 Precedence: list Reply-To: xvid-devel@xvid.org List-Id: xvid-devel.xvid.org List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Sender: xvid-devel-bounces@xvid.org Errors-To: xvid-devel-bounces@xvid.org Now you see how strange it is :) It would appear that we've got an impressive quality boost from your patch. Initially you only saw that with higher quants, however *if* I make all tables linear (same formula for low and high quants) then there's a detectable gain in lower quants too. I tweaked 8x8 lambda multiplier a bit and got some more. If we keep this, I'll be happy to sacrifice some speed. What I said above is true for no bframes. Example I just did, quant 3, defaults, no bframes: old code, size 43,010,048, psnr 44.1543, time 2:03 new code, size 42,905,600, psnr 44.1544, time 2:07 Funny, heh? It searches more and actually finds more. I'm happy with that. With bframes, I get this: old code: size 35,866,624, psnr 43.1332, time 2:01 new code: size 35,934,208, psnr 43.1822, time 1:52 So new and old are equivalent for psnr (more or less) with a rather major slowdown. Whatever new p-frames got us, b-frames wasted. I'll continue working on it now. If I "fix" bframes, then I will highly recommend a simple patch that changes 16x16 table to be linear. It looks like an unexpected win. Radek _______________________________________________ XviD-devel mailing list XviD-devel@xvid.org http://list.xvid.org/mailman/listinfo/xvid-devel From xvid-devel-bounces@xvid.org Sat Nov 26 15:03:54 2005 Return-Path: X-Original-To: itdp@fh-biergarten.de Delivered-To: itdp@localhost Received: from edu.bnhof.de (edu.bnhof.de [213.167.167.52]) by mail.kliche.org (Postfix) with ESMTP id 63C7B126D83 for ; Sat, 26 Nov 2005 15:03:54 +0100 (CET) Received: from edu.bnhof.de (localhost [127.0.0.1]) by edu.bnhof.de (Postfix) with ESMTP id D882918B92; Sat, 26 Nov 2005 15:03:50 +0100 (CET) X-Original-To: xvid-devel@edu.bnhof.de Delivered-To: xvid-devel@edu.bnhof.de Received: from s6.lansco.de (s6.lansco.de [213.239.218.110]) by edu.bnhof.de (Postfix) with ESMTP id 4068518B6B for ; Sat, 26 Nov 2005 15:03:48 +0100 (CET) Received: from s5.lansco.de (s5.lansco.de [212.63.130.170]) by s6.lansco.de (Postfix) with ESMTP id 41D721C8AC for ; Sat, 26 Nov 2005 15:03:47 +0100 (CET) Received: from localhost (s5.lansco.de [212.63.130.170]) by s5.lansco.de (Postfix) with ESMTP id F2E46872D for ; Sat, 26 Nov 2005 15:03:46 +0100 (CET) Received: from p54B5B7E9.dip.t-dialin.net (p54B5B7E9.dip.t-dialin.net [84.181.183.233]) by www.lansco.de (IMP) with HTTP for ; Sat, 26 Nov 2005 15:03:46 +0100 Message-ID: <1133013826.43886b42c6de9@www.lansco.de> Date: Sat, 26 Nov 2005 15:03:46 +0100 From: Michael Militzer To: xvid-devel@xvid.org Subject: Re: [XviD-devel] HUGE ME regressions between beta1 and beta2 References: <1132929315.4387212367133@www.lansco.de> <438731E1.9050905@syskin.cjb.net> <1132934488.4387355800550@www.lansco.de> <1132934559.4387359f82e20@www.lansco.de> <43880598.7010605@syskin.cjb.net> <1133001253.43883a25b075c@www.lansco.de> <43884191.3030502@syskin.cjb.net> <1133004637.4388475da551b@www.lansco.de> <43885821.801@syskin.cjb.net> In-Reply-To: <43885821.801@syskin.cjb.net> MIME-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 8bit User-Agent: Internet Messaging Program (IMP) 3.2.3 X-Originating-IP: 84.181.183.233 X-BeenThere: xvid-devel@xvid.org X-Mailman-Version: 2.1.4 Precedence: list Reply-To: xvid-devel@xvid.org List-Id: xvid-devel.xvid.org List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Sender: xvid-devel-bounces@xvid.org Errors-To: xvid-devel-bounces@xvid.org Quoting Radek Czyz : > Now you see how strange it is :) > > It would appear that we've got an impressive quality boost from your > patch. Initially you only saw that with higher quants, however *if* I > make all tables linear (same formula for low and high quants) then > there's a detectable gain in lower quants too. > > I tweaked 8x8 lambda multiplier a bit and got some more. > > If we keep this, I'll be happy to sacrifice some speed. > > What I said above is true for no bframes. > Example I just did, quant 3, defaults, no bframes: > old code, size 43,010,048, psnr 44.1543, time 2:03 > new code, size 42,905,600, psnr 44.1544, time 2:07 > > Funny, heh? It searches more and actually finds more. I'm happy with that. Good. That fits my findings. So seems the 8x8/16x16 decision wasn't working perfectly for DVD input and VHQ=0. Oh btw: I had similar gains with the pure linear table (with which I started) but then decided to trade the bit reduction for a smaller lambda - hoping that the perceived quality increase would be worth the 0.25-0.5% bitrate loss. But not sure here of course - you need a lot of testing with your own eyes to decide what's actually better. On the other hand: either way the loss wouldn't be huge... > With bframes, I get this: > > old code: size 35,866,624, psnr 43.1332, time 2:01 > new code: size 35,934,208, psnr 43.1822, time 1:52 > > So new and old are equivalent for psnr (more or less) with a rather > major slowdown. Whatever new p-frames got us, b-frames wasted. Hm, assuming your times are minutes:seconds then it looks as if the new code would be faster?? ;) > I'll continue working on it now. > > If I "fix" bframes, then I will highly recommend a simple patch that > changes 16x16 table to be linear. It looks like an unexpected win. Well, the gains at high QPs are _huge_ (>2 dB) so even if XviD would run 100% slower here it would still be worth it. On low QPs a 23% slow-down (as you mentioned in an earlier mail) isn't justified. But I don't think that comes from the lambda change itself but rather from some unexpected side-effects. So that needs to be fixed, sure. Bye, Michael _______________________________________________ XviD-devel mailing list XviD-devel@xvid.org http://list.xvid.org/mailman/listinfo/xvid-devel From xvid-devel-bounces@xvid.org Mon Nov 28 13:57:39 2005 Return-Path: X-Original-To: itdp@fh-biergarten.de Delivered-To: itdp@localhost Received: from edu.bnhof.de (edu.bnhof.de [213.167.167.52]) by mail.kliche.org (Postfix) with ESMTP id 73667126D82 for ; Mon, 28 Nov 2005 13:57:39 +0100 (CET) Received: from edu.bnhof.de (localhost [127.0.0.1]) by edu.bnhof.de (Postfix) with ESMTP id 7199718BE6; Mon, 28 Nov 2005 13:57:20 +0100 (CET) X-Original-To: xvid-devel@edu.bnhof.de Delivered-To: xvid-devel@edu.bnhof.de Received: from s6.lansco.de (s6.lansco.de [213.239.218.110]) by edu.bnhof.de (Postfix) with ESMTP id 7820818B8C for ; Mon, 28 Nov 2005 13:57:17 +0100 (CET) Received: from web30009.mail.mud.yahoo.com (web30009.mail.mud.yahoo.com [68.142.200.72]) by s6.lansco.de (Postfix) with SMTP id DA7DA21992 for ; Mon, 28 Nov 2005 13:57:15 +0100 (CET) Received: (qmail 62770 invoked by uid 60001); 28 Nov 2005 12:57:14 -0000 DomainKey-Signature: a=rsa-sha1; q=dns; c=nofws; s=s1024; d=yahoo.com; h=Message-ID:Received:Date:From:Subject:To:MIME-Version:Content-Type:Content-Transfer-Encoding; b=cEqrXRj9hEH3oAoYDGMDtzYSa3mBiN3yjOPSPsWC9VzKHRQ7S1fql8CJ6Ot4NoOwJxoMpVgh9wLL8DKAZoNqTuosD434zdjnRvBWIwGXcb+w5K+ppUclNIjosWTvUX0fT7KkWXdxoB5yvCge4sOUxnH907ydv+MrUg0xdRF3ilg= ; Message-ID: <20051128125714.62768.qmail@web30009.mail.mud.yahoo.com> Received: from [15.219.201.68] by web30009.mail.mud.yahoo.com via HTTP; Mon, 28 Nov 2005 04:57:14 PST Date: Mon, 28 Nov 2005 04:57:14 -0800 (PST) From: SHAILESHKUMAR GADHAVI To: xvid-devel@xvid.org MIME-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: 8bit X-Content-Filtered-By: Mailman/MimeDel 2.1.4 Subject: [XviD-devel] MPEG4 generation and playback. X-BeenThere: xvid-devel@xvid.org X-Mailman-Version: 2.1.4 Precedence: list Reply-To: xvid-devel@xvid.org List-Id: xvid-devel.xvid.org List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Sender: xvid-devel-bounces@xvid.org Errors-To: xvid-devel-bounces@xvid.org Hi All, I am using xvid_encraw.c program to generate the test file (from cactus .pgm). I am able to get an output file, but it does not play with Windows Media Player or Apple Quick Time player. Can anybody tell why is that so? Has anybody been able to play this file? Seeking help... Regards, Shailesh --------------------------------- Yahoo! Music Unlimited - Access over 1 million songs. Try it free. _______________________________________________ XviD-devel mailing list XviD-devel@xvid.org http://list.xvid.org/mailman/listinfo/xvid-devel From xvid-devel-bounces@xvid.org Mon Nov 28 16:44:04 2005 Return-Path: X-Original-To: itdp@fh-biergarten.de Delivered-To: itdp@localhost Received: from edu.bnhof.de (edu.bnhof.de [213.167.167.52]) by mail.kliche.org (Postfix) with ESMTP id CB3AA126D82 for ; Mon, 28 Nov 2005 16:44:04 +0100 (CET) Received: from edu.bnhof.de (localhost [127.0.0.1]) by edu.bnhof.de (Postfix) with ESMTP id 13CA51916F; Mon, 28 Nov 2005 16:43:52 +0100 (CET) X-Original-To: xvid-devel@edu.bnhof.de Delivered-To: xvid-devel@edu.bnhof.de Received: from s6.lansco.de (s6.lansco.de [213.239.218.110]) by edu.bnhof.de (Postfix) with ESMTP id B73EB1916C for ; Mon, 28 Nov 2005 16:43:49 +0100 (CET) Received: from web52509.mail.yahoo.com (web52509.mail.yahoo.com [206.190.48.192]) by s6.lansco.de (Postfix) with SMTP id 3D75A16AF0 for ; Mon, 28 Nov 2005 16:43:48 +0100 (CET) Received: (qmail 25595 invoked by uid 60001); 28 Nov 2005 15:43:46 -0000 DomainKey-Signature: a=rsa-sha1; q=dns; c=nofws; s=s1024; d=yahoo.com.ar; h=Message-ID:Received:Date:From:Subject:To:In-Reply-To:MIME-Version:Content-Type:Content-Transfer-Encoding; b=Z+lU36l8dRmyWKdykGFhwghUrVcb/7WTYE6qlEX3hwB2pzlqIf30O9/K4OSOop4/0JVbINERbpZORxIYSdsbTDOKn2hmW2iCIjZ82LQyIAlzVwApOrKaqR4B4jBZm892GzXyqNlj7xnB7Z+/gl6fc1p8SCrJJeXNit0coVXTRYE= ; Message-ID: <20051128154346.25593.qmail@web52509.mail.yahoo.com> Received: from [190.48.226.143] by web52509.mail.yahoo.com via HTTP; Mon, 28 Nov 2005 12:43:46 ART Date: Mon, 28 Nov 2005 12:43:46 -0300 (ART) From: Dark Sylinc Subject: Re: [XviD-devel] MPEG4 generation and playback. To: xvid-devel@xvid.org In-Reply-To: <20051128125714.62768.qmail@web30009.mail.mud.yahoo.com> MIME-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: 8bit X-Content-Filtered-By: Mailman/MimeDel 2.1.4 X-BeenThere: xvid-devel@xvid.org X-Mailman-Version: 2.1.4 Precedence: list Reply-To: xvid-devel@xvid.org List-Id: xvid-devel.xvid.org List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Sender: xvid-devel-bounces@xvid.org Errors-To: xvid-devel-bounces@xvid.org Try the GIMP (http://www.gimp.org/) or use a Linux player, like Kaffeine with Xine and libavcodec or XviD installed (http://kaffeine.sourceforge.net/) Cheers! Dark_Sylinc SHAILESHKUMAR GADHAVI escribió: Hi All, I am using xvid_encraw.c program to generate the test file (from cactus .pgm). I am able to get an output file, but it does not play with Windows Media Player or Apple Quick Time player. Can anybody tell why is that so? Has anybody been able to play this file? Seeking help... Regards, Shailesh --------------------------------- Yahoo! Music Unlimited - Access over 1 million songs. Try it free. _______________________________________________ XviD-devel mailing list XviD-devel@xvid.org http://list.xvid.org/mailman/listinfo/xvid-devel --------------------------------- 1GB gratis, Antivirus y Antispam Correo Yahoo!, el mejor correo web del mundo Abrí tu cuenta aquí _______________________________________________ XviD-devel mailing list XviD-devel@xvid.org http://list.xvid.org/mailman/listinfo/xvid-devel From xvid-devel-bounces@xvid.org Tue Nov 29 07:00:53 2005 Return-Path: X-Original-To: itdp@fh-biergarten.de Delivered-To: itdp@localhost Received: from edu.bnhof.de (edu.bnhof.de [213.167.167.52]) by mail.kliche.org (Postfix) with ESMTP id 489E1126D89 for ; Tue, 29 Nov 2005 07:00:52 +0100 (CET) Received: from edu.bnhof.de (localhost [127.0.0.1]) by edu.bnhof.de (Postfix) with ESMTP id 1E48E18BBA; Tue, 29 Nov 2005 07:00:41 +0100 (CET) X-Original-To: xvid-devel@edu.bnhof.de Delivered-To: xvid-devel@edu.bnhof.de Received: from s6.lansco.de (s6.lansco.de [213.239.218.110]) by edu.bnhof.de (Postfix) with ESMTP id C2FB618BAE for ; Tue, 29 Nov 2005 07:00:37 +0100 (CET) Received: from web32804.mail.mud.yahoo.com (web32804.mail.mud.yahoo.com [68.142.206.34]) by s6.lansco.de (Postfix) with SMTP id 38FB51CAF0 for ; Tue, 29 Nov 2005 07:00:36 +0100 (CET) Received: (qmail 74248 invoked by uid 60001); 29 Nov 2005 06:00:34 -0000 DomainKey-Signature: a=rsa-sha1; q=dns; c=nofws; s=s1024; d=yahoo.com; h=Message-ID:Received:Date:From:Subject:To:MIME-Version:Content-Type:Content-Transfer-Encoding; b=ijO1i4IZjx8sYTXJ0/Bl94nZBrQ2f8SzoCwDZCVy5CLeHy8F4rmVyISEW23N8MdES9PIdDyReZ6c4SMM6sSxCusnsMax6hnKQEklfQ5L1v7vNv/eeH5r7zL2dzeSEH7l7OwUJ6S0Dqqnc5FCrBfBLCQzLDr/Di+rBRsJ4PA50Gg= ; Message-ID: <20051129060034.74246.qmail@web32804.mail.mud.yahoo.com> Received: from [221.134.65.200] by web32804.mail.mud.yahoo.com via HTTP; Mon, 28 Nov 2005 22:00:34 PST Date: Mon, 28 Nov 2005 22:00:34 -0800 (PST) From: "M.R.Gavaskar" To: xvid-devel@xvid.org MIME-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: 8bit X-Content-Filtered-By: Mailman/MimeDel 2.1.4 Subject: [XviD-devel] Reg - Xvid features X-BeenThere: xvid-devel@xvid.org X-Mailman-Version: 2.1.4 Precedence: list Reply-To: xvid-devel@xvid.org List-Id: xvid-devel.xvid.org List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Sender: xvid-devel-bounces@xvid.org Errors-To: xvid-devel-bounces@xvid.org Hi team, I am very new to this Xvid media codec. I am in the learning stage. Can any one tell me what are all the features it is currently supporting ? So that I can go for using this codec for development. Eagerly waiting for your replay… Cheers, Gavaskar M.R. --------------------------------- Yahoo! Personals Single? There's someone we'd like you to meet. Lots of someones, actually. Yahoo! Personals _______________________________________________ XviD-devel mailing list XviD-devel@xvid.org http://list.xvid.org/mailman/listinfo/xvid-devel From xvid-devel-bounces@xvid.org Tue Nov 29 09:04:58 2005 Return-Path: X-Original-To: itdp@fh-biergarten.de Delivered-To: itdp@localhost Received: from edu.bnhof.de (edu.bnhof.de [213.167.167.52]) by mail.kliche.org (Postfix) with ESMTP id 88EEF126D83 for ; Tue, 29 Nov 2005 09:04:58 +0100 (CET) Received: from edu.bnhof.de (localhost [127.0.0.1]) by edu.bnhof.de (Postfix) with ESMTP id 6BB5F18B7A; Tue, 29 Nov 2005 09:04:39 +0100 (CET) X-Original-To: xvid-devel@edu.bnhof.de Delivered-To: xvid-devel@edu.bnhof.de Received: from s6.lansco.de (s6.lansco.de [213.239.218.110]) by edu.bnhof.de (Postfix) with ESMTP id 6DF3D15888 for ; Tue, 29 Nov 2005 09:04:36 +0100 (CET) Received: from s5.lansco.de (s5.lansco.de [212.63.130.170]) by s6.lansco.de (Postfix) with ESMTP id 478B721A9A for ; Tue, 29 Nov 2005 09:04:35 +0100 (CET) Received: from [192.168.1.66] (ppp64-78.lns1.mel2.internode.on.net [59.167.64.78]) by s5.lansco.de (Postfix) with ESMTP id E383A306EF for ; Tue, 29 Nov 2005 09:04:33 +0100 (CET) Date: Tue, 29 Nov 2005 19:04:46 +1100 From: pete To: xvid-devel@xvid.org Subject: Re: [XviD-devel] Reg - Xvid features In-Reply-To: <20051129060034.74246.qmail@web32804.mail.mud.yahoo.com> References: <20051129060034.74246.qmail@web32804.mail.mud.yahoo.com> Message-Id: <20051129190324.0626.PROSS@xvid.org> MIME-Version: 1.0 Content-Type: text/plain; charset="US-ASCII" Content-Transfer-Encoding: 7bit X-Mailer: Becky! ver. 2.21.01 [en] X-BeenThere: xvid-devel@xvid.org X-Mailman-Version: 2.1.4 Precedence: list Reply-To: xvid-devel@xvid.org List-Id: xvid-devel.xvid.org List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Sender: xvid-devel-bounces@xvid.org Errors-To: xvid-devel-bounces@xvid.org On Mon, 28 Nov 2005 22:00:34 -0800 (PST) "M.R.Gavaskar" wrote: > Hi team, > I am very new to this Xvid media codec. I am in the learning stage. > Can any one tell me what are all the features it is currently supporting ? see: http://forum.doom9.org/showthread.php?t=16935 http://ronald.vslcatena.nl/docs/xvidfaq.html http://en.wikipedia.org/wiki/XviD cheers, -- pete _______________________________________________ XviD-devel mailing list XviD-devel@xvid.org http://list.xvid.org/mailman/listinfo/xvid-devel From xvid-devel-bounces@xvid.org Tue Nov 29 09:07:16 2005 Return-Path: X-Original-To: itdp@fh-biergarten.de Delivered-To: itdp@localhost Received: from edu.bnhof.de (edu.bnhof.de [213.167.167.52]) by mail.kliche.org (Postfix) with ESMTP id 98DA2126D89 for ; Tue, 29 Nov 2005 09:07:16 +0100 (CET) Received: from edu.bnhof.de (localhost [127.0.0.1]) by edu.bnhof.de (Postfix) with ESMTP id 9058718BA7; Tue, 29 Nov 2005 09:07:05 +0100 (CET) X-Original-To: xvid-devel@edu.bnhof.de Delivered-To: xvid-devel@edu.bnhof.de Received: from s6.lansco.de (s6.lansco.de [213.239.218.110]) by edu.bnhof.de (Postfix) with ESMTP id B51C618B8D for ; Tue, 29 Nov 2005 09:07:02 +0100 (CET) Received: from s5.lansco.de (s5.lansco.de [212.63.130.170]) by s6.lansco.de (Postfix) with ESMTP id 9131721AA2 for ; Tue, 29 Nov 2005 09:07:01 +0100 (CET) Received: from [192.168.1.66] (ppp64-78.lns1.mel2.internode.on.net [59.167.64.78]) by s5.lansco.de (Postfix) with ESMTP id C863D306FA; Tue, 29 Nov 2005 09:06:59 +0100 (CET) Date: Tue, 29 Nov 2005 19:07:12 +1100 From: pete To: xvid-devel@xvid.org Subject: Re: [XviD-devel] Variable Quant + Variable Bitrate (Can it happen?) In-Reply-To: <097CA5D0-352E-4FFC-A010-06967F36532A@comcast.net> References: <097CA5D0-352E-4FFC-A010-06967F36532A@comcast.net> Message-Id: <20051129190638.062B.PROSS@xvid.org> MIME-Version: 1.0 Content-Type: text/plain; charset="US-ASCII" Content-Transfer-Encoding: 7bit X-Mailer: Becky! ver. 2.21.01 [en] Cc: leeamorgan@comcast.net X-BeenThere: xvid-devel@xvid.org X-Mailman-Version: 2.1.4 Precedence: list Reply-To: pross@xvid.org, xvid-devel@xvid.org List-Id: xvid-devel.xvid.org List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Sender: xvid-devel-bounces@xvid.org Errors-To: xvid-devel-bounces@xvid.org hi, [cleaning out my draft mbox...] On Fri, 28 Jan 2005 22:25:31 -0500 Lee Morgan wrote: > Would someone please take a bit of time to explain to me why there > can't be a variable quant with variable bitrate mode? > I know, I know - it exists (2 pass right). But that's not what I mean. > To my understanding 2 pass takes the input bitrate as a avg it intends > to meet. It then analyzes the video determining what parts of the video > need what bitrate then on the second pass varies the quantizers as need > to achieve that overall bitrate (more or less correct?). > It's been stated that 2 pass has a method of determining what level of > quantization a macro block should have (to retain high quality). > Could this not be determined "on the fly" during a single pass encode > to allow instead of a fixed quant but rather a variable quant + > variable bitrate mode? Which would use the lowest needed quantization > for a macro block while still allowing for higher ones to be used when > possible. > > This way it is possible to have a high quality single pass fixed quant > 2 without wasting bits. the 1pass and 2pass rate control algorithms manipulate the quantizer in effectively the same manner. ========= mpeg-4 compression is achieved by quantizing the video. the level of quantization is dictated by the quantizer value. here is a crude ascii diagram showing a black box view of xvid: input video -> [ xvidcore ] -> bitstream output ^ | quantizer notice how there is no mention of "bitrate" in this diagram... that is because we cannot directly control bitrate! instead, we employ a rate control algorithm that instructs xvidcore what quantizer to use for each frame. * the 1pass algorithm measures the bits for each frame compressed by the encoder, and adjusts the next frame's quantizer in order to achieve the target bitrate. the problem with 1pass is that it must react to the input video. so if there is a high-motion scene, followed by a low-motion scene, it will adjust the quantizer such that each scene is allocated the same number of bits. this causes the high-motion scene to look ugly. * the 2pass algorithm overcomes this reactionary problem. the first pass measures the bits for each frame compressed by the encoder. the locations of high- and low-motion scenes are identified, so that the quantizer can be adjusted in the second pass in order to allocate more bits to high-motion scenes and less bits to low-motion scenes. > P.S. Sorry if this double posts (I've been having problems with my mail > today - and I didn't see it come through on the list). yep. several people have been reported problems. -- pete _______________________________________________ XviD-devel mailing list XviD-devel@xvid.org http://list.xvid.org/mailman/listinfo/xvid-devel From xvid-devel-bounces@xvid.org Tue Nov 29 09:08:26 2005 Return-Path: X-Original-To: itdp@fh-biergarten.de Delivered-To: itdp@localhost Received: from edu.bnhof.de (edu.bnhof.de [213.167.167.52]) by mail.kliche.org (Postfix) with ESMTP id CB00B126D8B for ; Tue, 29 Nov 2005 09:08:25 +0100 (CET) Received: from edu.bnhof.de (localhost [127.0.0.1]) by edu.bnhof.de (Postfix) with ESMTP id BFB0218BBB; Tue, 29 Nov 2005 09:08:07 +0100 (CET) X-Original-To: xvid-devel@edu.bnhof.de Delivered-To: xvid-devel@edu.bnhof.de Received: from s6.lansco.de (s6.lansco.de [213.239.218.110]) by edu.bnhof.de (Postfix) with ESMTP id 633D118B8D for ; Tue, 29 Nov 2005 09:08:05 +0100 (CET) Received: from s5.lansco.de (s5.lansco.de [212.63.130.170]) by s6.lansco.de (Postfix) with ESMTP id 4457521AA9 for ; Tue, 29 Nov 2005 09:08:04 +0100 (CET) Received: from [192.168.1.66] (ppp64-78.lns1.mel2.internode.on.net [59.167.64.78]) by s5.lansco.de (Postfix) with ESMTP id 999F428C60; Tue, 29 Nov 2005 09:08:02 +0100 (CET) Date: Tue, 29 Nov 2005 19:08:15 +1100 From: pete To: xvid-devel@xvid.org Subject: Re: [XviD-devel] mpeg_quant In-Reply-To: <64DEC4FC-5CCF-11D9-8997-000A95772E90@kbw.ch> References: <64DEC4FC-5CCF-11D9-8997-000A95772E90@kbw.ch> Message-Id: <20051129190757.062E.PROSS@xvid.org> MIME-Version: 1.0 Content-Type: text/plain; charset="ISO-8859-1" Content-Transfer-Encoding: quoted-printable X-Mailer: Becky! ver. 2.21.01 [en] Cc: chn@kbw.ch X-BeenThere: xvid-devel@xvid.org X-Mailman-Version: 2.1.4 Precedence: list Reply-To: pross@xvid.org, xvid-devel@xvid.org List-Id: xvid-devel.xvid.org List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Sender: xvid-devel-bounces@xvid.org Errors-To: xvid-devel-bounces@xvid.org hi, [cleaning out my draft mbox...] On Sun, 2 Jan 2005 16:03:00 +0100 Christoph N=E4geli wrote: > I want to create some altivec routines for mpeg quantization. great. > Can anyone tell me how many bits I need to process mpeg quantization? > Is 16-Bit Arithmetic enough or do I need 32-Bit? 16-bits is insufficient for mpeg quantization when using large quantizers. i can't recall exactly what constitues large, but it=20 is alteast q=3D10. cheers, -- pete _______________________________________________ XviD-devel mailing list XviD-devel@xvid.org http://list.xvid.org/mailman/listinfo/xvid-devel From xvid-devel-bounces@xvid.org Tue Nov 29 09:08:59 2005 Return-Path: X-Original-To: itdp@fh-biergarten.de Delivered-To: itdp@localhost Received: from edu.bnhof.de (edu.bnhof.de [213.167.167.52]) by mail.kliche.org (Postfix) with ESMTP id 1282C126D83 for ; Tue, 29 Nov 2005 09:08:59 +0100 (CET) Received: from edu.bnhof.de (localhost [127.0.0.1]) by edu.bnhof.de (Postfix) with ESMTP id 9719118BC2; Tue, 29 Nov 2005 09:08:45 +0100 (CET) X-Original-To: xvid-devel@edu.bnhof.de Delivered-To: xvid-devel@edu.bnhof.de Received: from s6.lansco.de (s6.lansco.de [213.239.218.110]) by edu.bnhof.de (Postfix) with ESMTP id 67D4418B86 for ; Tue, 29 Nov 2005 09:08:41 +0100 (CET) Received: from chen (unknown [222.70.220.124]) by s6.lansco.de (Postfix) with SMTP id 656CC21AAC for ; Tue, 29 Nov 2005 09:08:38 +0100 (CET) Date: Tue, 29 Nov 2005 16:08:48 +0800 From: "chenm002@163.com" To: "xvid-devel" X-mailer: Foxmail 5.0 [cn] Mime-Version: 1.0 Message-Id: <20051129080838.656CC21AAC@s6.lansco.de> Subject: [XviD-devel] [fix bug for decode with intra_dc_threshold] X-BeenThere: xvid-devel@xvid.org X-Mailman-Version: 2.1.4 Precedence: list Reply-To: xvid-devel@xvid.org List-Id: xvid-devel.xvid.org List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Content-Type: multipart/mixed; boundary="===============0094857111==" Sender: xvid-devel-bounces@xvid.org Errors-To: xvid-devel-bounces@xvid.org --===============0094857111== Content-Type: text/plain; charset="gb2312" Content-Transfer-Encoding: base64 SGksDQoNCiAgICBJIGZpeCBzb21lIGJ1ZyBmb3IgZGVjb2Rlci4NCiAJCQkJDQpJbmRleDogZGVj b2Rlci5jDQo9PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09 PT09PT09PT09PT09PT09PT09DQpSQ1MgZmlsZTogL3h2aWQveHZpZGNvcmUvc3JjL2RlY29kZXIu Yyx2DQpyZXRyaWV2aW5nIHJldmlzaW9uIDEuNzUNCmRpZmYgLXIxLjc1IGRlY29kZXIuYw0KMjQz YTI0NA0KPiAgICAgICAgIGNvbnN0IHVpbnQzMl90IHByZXZfcXVhbnQsIC8qIHVzZSBmb3IgaW50 cmFfZGNfdGhyZXNob2xkICovDQoyNzdjMjc4DQo8ICAgICBpZiAocXVhbnQgPCBpbnRyYV9kY190 aHJlc2hvbGQpIHsNCi0tLQ0KPiAgICAgaWYgKHByZXZfcXVhbnQgPCBpbnRyYV9kY190aHJlc2hv bGQpIHsNCjcyNWE3MjcNCj4gICBpbnQgcHJldl9xdWFudDsNCjc2MGE3NjMNCj4gICAgICAgcHJl dl9xdWFudCA9IHF1YW50Ow0KNzgwYzc4Mw0KPCAgICAgICBkZWNvZGVyX21iaW50cmEoZGVjLCBt YiwgeCwgeSwgYWNwcmVkX2ZsYWcsIGNicCwgYnMsIHF1YW50LA0KLS0tDQo+ICAgICAgIGRlY29k ZXJfbWJpbnRyYShkZWMsIG1iLCB4LCB5LCBhY3ByZWRfZmxhZywgY2JwLCBicywgcXVhbnQsIHBy ZXZfcXVhbnQsDQo5MzlhOTQzDQo+ICAgaW50IHByZXZfcXVhbnQ7DQoxMDAzYTEwMDgNCj4gICAg ICAgICBwcmV2X3F1YW50ID0gcXVhbnQ7DQoxMDYwYzEwNjUNCjwgICAgICAgICAgICAgICAgICAg aW50cmFfZGNfdGhyZXNob2xkLCBib3VuZCk7DQotLS0NCj4gICAgICAgICAgICAgICAgICAgcHJl dl9xdWFudCwgaW50cmFfZGNfdGhyZXNob2xkLCBib3VuZCk7DQoNCqGhoaGhoaGhoaGhoaGhoaFj aGVuDQqhoaGhoaGhoaGhoaGhoaGhY2hlbm0wMDFAMTYzLmNvbQ0KoaGhoaGhoaGhoaGhoaGhoaGh oaEyMDA1LTExLTI5DQo= --===============0094857111== Content-Type: text/plain; charset="us-ascii" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit Content-Disposition: inline _______________________________________________ XviD-devel mailing list XviD-devel@xvid.org http://list.xvid.org/mailman/listinfo/xvid-devel --===============0094857111==-- From xvid-devel-bounces@xvid.org Tue Nov 29 09:14:31 2005 Return-Path: X-Original-To: itdp@fh-biergarten.de Delivered-To: itdp@localhost Received: from edu.bnhof.de (edu.bnhof.de [213.167.167.52]) by mail.kliche.org (Postfix) with ESMTP id 71AC7126D89 for ; Tue, 29 Nov 2005 09:14:31 +0100 (CET) Received: from edu.bnhof.de (localhost [127.0.0.1]) by edu.bnhof.de (Postfix) with ESMTP id 9F8C518BC7; Tue, 29 Nov 2005 09:14:20 +0100 (CET) X-Original-To: xvid-devel@edu.bnhof.de Delivered-To: xvid-devel@edu.bnhof.de Received: from s6.lansco.de (s6.lansco.de [213.239.218.110]) by edu.bnhof.de (Postfix) with ESMTP id 8CB8318B9C for ; Tue, 29 Nov 2005 09:14:18 +0100 (CET) Received: from chen (unknown [222.70.220.124]) by s6.lansco.de (Postfix) with SMTP id F36A921997 for ; Tue, 29 Nov 2005 09:14:15 +0100 (CET) Date: Tue, 29 Nov 2005 16:14:24 +0800 From: "=?utf-7?B?Y2hlbm0wMDIrQUVBLTE2My5jb20=?=" To: "=?utf-7?B?eHZpZC1kZXZlbA==?=" X-mailer: Foxmail 5.0 [cn] Mime-Version: 1.0 Content-Type: text/plain; charset="utf-7" Content-Transfer-Encoding: 7bit Message-Id: <20051129081415.F36A921997@s6.lansco.de> Subject: [XviD-devel] [fix bug for decode with intra_dc_threshold] X-BeenThere: xvid-devel@xvid.org X-Mailman-Version: 2.1.4 Precedence: list Reply-To: xvid-devel@xvid.org List-Id: xvid-devel.xvid.org List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Sender: xvid-devel-bounces@xvid.org Errors-To: xvid-devel-bounces@xvid.org Hi, I fix some bug for decoder. Index: decoder.c +AD0APQA9AD0APQA9AD0APQA9AD0APQA9AD0APQA9AD0APQA9AD0APQA9AD0APQA9AD0APQA9AD0APQA9AD0APQA9AD0APQA9AD0APQA9AD0APQA9AD0APQA9AD0APQA9AD0APQA9AD0APQA9AD0APQA9AD0APQA9AD0APQA9AD0APQA9AD0- RCS file: /xvid/xvidcore/src/decoder.c,v retrieving revision 1.75 diff -r1.75 decoder.c 243a244 +AD4- const uint32+AF8-t prev+AF8-quant, /+ACo- use for intra+AF8-dc+AF8-threshold +ACo-/ 277c278 +ADw- if (quant +ADw- intra+AF8-dc+AF8-threshold) +AHs- --- +AD4- if (prev+AF8-quant +ADw- intra+AF8-dc+AF8-threshold) +AHs- 725a727 +AD4- int prev+AF8-quant+ADs- 760a763 +AD4- prev+AF8-quant +AD0- quant+ADs- 780c783 +ADw- decoder+AF8-mbintra(dec, mb, x, y, acpred+AF8-flag, cbp, bs, quant, --- +AD4- decoder+AF8-mbintra(dec, mb, x, y, acpred+AF8-flag, cbp, bs, quant, prev+AF8-quant, 939a943 +AD4- int prev+AF8-quant+ADs- 1003a1008 +AD4- prev+AF8-quant +AD0- quant+ADs- 1060c1065 +ADw- intra+AF8-dc+AF8-threshold, bound)+ADs- --- +AD4- prev+AF8-quant, intra+AF8-dc+AF8-threshold, bound)+ADs- +MAAwADAAMAAwADAAMAAwAA-chen +MAAwADAAMAAwADAAMAAwAA-chenm001+AEA-163.com +MAAwADAAMAAwADAAMAAwADAAMAA-2005-11-29 _______________________________________________ XviD-devel mailing list XviD-devel@xvid.org http://list.xvid.org/mailman/listinfo/xvid-devel From xvid-devel-bounces@xvid.org Tue Nov 29 09:43:32 2005 Return-Path: X-Original-To: itdp@fh-biergarten.de Delivered-To: itdp@localhost Received: from edu.bnhof.de (edu.bnhof.de [213.167.167.52]) by mail.kliche.org (Postfix) with ESMTP id E8389126D89 for ; Tue, 29 Nov 2005 09:43:30 +0100 (CET) Received: from edu.bnhof.de (localhost [127.0.0.1]) by edu.bnhof.de (Postfix) with ESMTP id 09A2C18BCE; Tue, 29 Nov 2005 09:43:19 +0100 (CET) X-Original-To: xvid-devel@edu.bnhof.de Delivered-To: xvid-devel@edu.bnhof.de Received: from s6.lansco.de (s6.lansco.de [213.239.218.110]) by edu.bnhof.de (Postfix) with ESMTP id 3EEAA18BC3 for ; Tue, 29 Nov 2005 09:43:14 +0100 (CET) Received: from 163.com (m13-38.163.com [220.181.13.38]) by s6.lansco.de (Postfix) with SMTP id 1389221966 for ; Tue, 29 Nov 2005 09:43:11 +0100 (CET) Received: from web-app38 ( web-app38 [192.168.193.38] ) by webmail (Coremail) ; Tue, 29 Nov 2005 16:26:47 +0800 (CST) MIME-Version: 1.0 Message-ID: <438C10C7.000033.02578@bj163app38.163.com> Date: Tue, 29 Nov 2005 16:26:47 +0800 (CST) From: "=?gb2312?B?s8LD9A==?=" To: xvid-devel@xvid.org X-Priority: 3 X-Originating-IP: [192.168.208.157(222.70.220.124)] X-Mailer: 163com X-Content-Filtered-By: Mailman/MimeDel 2.1.4 Subject: [XviD-devel] [fix bug for decode with intra_dc_threshold] X-BeenThere: xvid-devel@xvid.org X-Mailman-Version: 2.1.4 Precedence: list Reply-To: xvid-devel@xvid.org List-Id: xvid-devel.xvid.org List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Content-Type: multipart/mixed; boundary="===============0585392599==" Sender: xvid-devel-bounces@xvid.org Errors-To: xvid-devel-bounces@xvid.org --===============0585392599== Content-Type: text/plain; charset="gb2312" Content-Transfer-Encoding: base64 SGksDQogICAgSSBmaXggc29tZSBidWcgZm9yIGRlY29kZXIuDQoNCkluZGV4OiBkZWNvZGVy LmMNCj09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09PT09 PT09PT09PT09PT09PT09PT0NClJDUyBmaWxlOiAveHZpZC94dmlkY29yZS9zcmMvZGVjb2Rl ci5jLHYNCnJldHJpZXZpbmcgcmV2aXNpb24gMS43NQ0KZGlmZiAtcjEuNzUgZGVjb2Rlci5j DQoyNDNhMjQ0DQo+ICAgICAgICAgY29uc3QgdWludDMyX3QgcHJldl9xdWFudCwgLyogdXNl IGZvciBpbnRyYV9kY190aHJlc2hvbGQgKi8NCjI3N2MyNzgNCjwgICAgIGlmIChxdWFudCA8 IGludHJhX2RjX3RocmVzaG9sZCkgew0KLS0tDQo+ICAgICBpZiAocHJldl9xdWFudCA8IGlu dHJhX2RjX3RocmVzaG9sZCkgew0KNzI1YTcyNw0KPiAgIGludCBwcmV2X3F1YW50Ow0KNzYw YTc2Mw0KPiAgICAgICBwcmV2X3F1YW50ID0gcXVhbnQ7DQo3ODBjNzgzDQo8ICAgICAgIGRl Y29kZXJfbWJpbnRyYShkZWMsIG1iLCB4LCB5LCBhY3ByZWRfZmxhZywgY2JwLCBicywgcXVh bnQsDQotLS0NCj4gICAgICAgZGVjb2Rlcl9tYmludHJhKGRlYywgbWIsIHgsIHksIGFjcHJl ZF9mbGFnLCBjYnAsIGJzLCBxdWFudCwgcHJldl9xdWFudCwNCjkzOWE5NDMNCj4gICBpbnQg cHJldl9xdWFudDsNCjEwMDNhMTAwOA0KPiAgICAgICAgIHByZXZfcXVhbnQgPSBxdWFudDsN CjEwNjBjMTA2NQ0KPCAgICAgICAgICAgICAgICAgICBpbnRyYV9kY190aHJlc2hvbGQsIGJv dW5kKTsNCi0tLQ0KPiAgICAgICAgICAgICAgICAgICBwcmV2X3F1YW50LCBpbnRyYV9kY190 aHJlc2hvbGQsIGJvdW5kKTsNCg0KoaGhoaGhoaGhoaGhoaGhoWNoZW4NCqGhoaGhoaGhoaGh oaGhoaFjaGVubTAwMUAxNjMuY29tDQqhoaGhoaGhoaGhoaGhoaGhoaGhoTIwMDUtMTEtMjkN Cg== --===============0585392599== Content-Type: text/plain; charset="us-ascii" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit Content-Disposition: inline _______________________________________________ XviD-devel mailing list XviD-devel@xvid.org http://list.xvid.org/mailman/listinfo/xvid-devel --===============0585392599==--