Debian Bug report logs - #369134
ffmpeg: start time specified after input files is applied wrongly

Toggle useless messages

Message #5 received at submit@bugs.debian.org (full text, mbox, reply):

From: Ben Hutchings <ben@decadentplace.org.uk>

To: Debian Bug Tracking System <submit@bugs.debian.org>

Subject: ffmpeg: start time specified after input files is applied wrongly

Date: Sat, 27 May 2006 20:01:31 +0100

Package: ffmpeg
Version: 0.cvs20050313-2sarge1
Severity: normal
Tags: patch

If a start time (-ss option) is specified after all input files, the
specified time period is correctly discarded from input streams, but a
video output stream will be delayed by that time again with respect to
an audio output stream; i.e. the output streams will be out-of-synch.

Here is a fix for the version in sarge (0.cvs20050313-2sarge1):

--- ffmpeg.c.orig	2005-03-12 17:24:20.000000000 +0000
+++ ffmpeg.c	2006-05-26 20:37:36.000000000 +0100
@@ -474,7 +474,7 @@
         return;               /* Should signal an error ! */
 
     if(audio_sync_method){
-        double delta = ost->sync_ipts * enc->sample_rate - ost->sync_opts 
+        double delta = (ost->sync_ipts - (double)start_time / AV_TIME_BASE) * enc->sample_rate - ost->sync_opts 
                 - fifo_size(&ost->fifo, ost->fifo.rptr)/(ost->st->codec.channels * 2);
         double idelta= delta*ist->st->codec.sample_rate / enc->sample_rate;
         int byte_delta= ((int)idelta)*2*ist->st->codec.channels;
@@ -517,7 +517,7 @@
             }
         } 
     }else
-        ost->sync_opts= lrintf(ost->sync_ipts * enc->sample_rate)
+        ost->sync_opts= lrintf((ost->sync_ipts - (double)start_time / AV_TIME_BASE) * enc->sample_rate)
                         - fifo_size(&ost->fifo, ost->fifo.rptr)/(ost->st->codec.channels * 2); //FIXME wrong
 
     if (ost->audio_resample) {
@@ -703,7 +703,7 @@
 
     if(video_sync_method){
         double vdelta;
-        vdelta = ost->sync_ipts * enc->frame_rate / enc->frame_rate_base - ost->sync_opts;
+        vdelta = (ost->sync_ipts - (double)start_time / AV_TIME_BASE) * enc->frame_rate / enc->frame_rate_base - ost->sync_opts;
         //FIXME set to 0.5 after we fix some dts/pts bugs like in avidec.c
         if (vdelta < -1.1)
             nb_frames = 0;
@@ -720,7 +720,7 @@
                 fprintf(stderr, "*** %d dup!\n", nb_frames-1);
         }
     }else
-        ost->sync_opts= lrintf(ost->sync_ipts * enc->frame_rate / enc->frame_rate_base);
+        ost->sync_opts= lrintf((ost->sync_ipts - (double)start_time / AV_TIME_BASE) * enc->frame_rate / enc->frame_rate_base);
 
     nb_frames= FFMIN(nb_frames, max_frames[CODEC_TYPE_VIDEO] - ost->frame_number);
     if (nb_frames <= 0) 
-- END --

It appears that the version in etch and sid (0.cvs20060329-4) has had
a similar change applied to it.  However, I have not been able to
install and test that version.

This bug left me with a lot of useless XviD files at the end of
DebConf and I'm inclined to consider it a serious data-loss bug that
merits an update in stable.  On the other hand, sensible people should
check the output by eye earlier on...

Ben.

-- System Information:
Debian Release: 3.1
  APT prefers unstable
  APT policy: (500, 'unstable')
Architecture: i386 (i686)
Kernel: Linux 2.6.8-2-686
Locale: LANG=en_GB.UTF-8, LC_CTYPE=en_GB.UTF-8 (charmap=UTF-8)

Versions of packages ffmpeg depends on:
ii  libc6         2.3.6-4                    GNU C Library: Shared libraries an
ii  libfreetype6  2.1.7-2.4                  FreeType 2 font engine, shared lib
ii  libimlib2     1.2.0-2.2                  powerful image loading and renderi
ii  libsdl1.2debi 1.2.7+1.2.8cvs20041007-4.1 Simple DirectMedia Layer
ii  libx11-6      4.3.0.dfsg.1-14sarge1      X Window System protocol client li
ii  xlibs         4.3.0.dfsg.1-14sarge1      X Keyboard Extension (XKB) configu
ii  zlib1g        1:1.2.2-4.sarge.2          compression library - runtime

-- no debconf information

Message #10 received at 369134@bugs.debian.org (full text, mbox, reply):

From: Fabian Greffrath <greffrath@leat.rub.de>

To: control@bugs.debian.org

Cc: 369134@bugs.debian.org, 420230@bugs.debian.org, 426707@bugs.debian.org, 433287@bugs.debian.org, 447478@bugs.debian.org, 294422@bugs.debian.org, 298095@bugs.debian.org, 370333@bugs.debian.org, 364487@bugs.debian.org, 374931@bugs.debian.org, 448072@bugs.debian.org, 369127@bugs.debian.org, 370599@bugs.debian.org, 418231@bugs.debian.org, 420231@bugs.debian.org, 430923@bugs.debian.org, 476644@bugs.debian.org, 418228@bugs.debian.org

Subject: reassign ffmpeg bugs to ffmpeg-free

Date: Fri, 23 May 2008 10:49:45 +0200

# reassign ffmpeg bugs to ffmpeg-free
reassign 369134 ffmpeg-free
reassign 420230 ffmpeg-free
reassign 426707 ffmpeg-free
reassign 433287 ffmpeg-free
reassign 447478 ffmpeg-free
reassign 294422 ffmpeg-free
reassign 298095 ffmpeg-free
reassign 370333 ffmpeg-free
reassign 364487 ffmpeg-free
reassign 374931 ffmpeg-free
reassign 448072 ffmpeg-free
reassign 369127 ffmpeg-free
reassign 370599 ffmpeg-free
reassign 418231 ffmpeg-free
reassign 420231 ffmpeg-free
reassign 430923 ffmpeg-free
reassign 476644 ffmpeg-free
reassign 418228 ffmpeg-free
tag 418228 - pending
thanks

Message #21 received at 369134-done@bugs.debian.org (full text, mbox, reply):

From: Reinhard Tartler <siretart@tauware.de>

To: 369134-done@bugs.debian.org

Cc: Ben Hutchings <ben@decadentplace.org.uk>

Subject: Re: ffmpeg: start time specified after input files is applied wrongly

Date: Sun, 31 Jan 2010 17:06:21 +0100

Version: 0.svn20080206-1

Sorry for taking so long to get back to you.

On Sa, Mai 27, 2006 at 21:01:31 (CEST), Ben Hutchings wrote:

> Package: ffmpeg
> Version: 0.cvs20050313-2sarge1
> Severity: normal
> Tags: patch
>
> If a start time (-ss option) is specified after all input files, the
> specified time period is correctly discarded from input streams, but a
> video output stream will be delayed by that time again with respect to
> an audio output stream; i.e. the output streams will be out-of-synch.
>
> Here is a fix for the version in sarge (0.cvs20050313-2sarge1):

This patch does no longer apply to our package at all. By looking at it,
I'm fairly confident that this has been fixed now anyway, which is why
I'm closing this report.

In case that you can confirm this is not fixed for you, please try to
forward port that patch, ideally to current ffmpeg trunk.

thanks for your understanding.

-- 
Gruesse/greetings,
Reinhard Tartler, KeyID 945348A4

Send a report that this bug log contains spam.