Windows - theora 0.27 stable biuld crash

can you provide access to that input file, with that we might be able to
figure out whats causing the problem, can you also try:

ffmpeg2theora-0.27.exe --optimize --speedlevel 0 -a 1 -- aspect 16:9
 -x 720 -y 480 --nosubtitles Input_file_720p_x264_AAC_SRT_in_mkv

and let me know if that also crashes.

That file you can get at this site :
http://www.animetake.com/hime-chen-otogi-chikku-idol-lilpri-episode-11/

I try :
ffmpeg2theora-0.27.exe --optimize --speedlevel 0 -a 1 --aspect 16:9 -x 720 -y 480 --nosubtitles Input_file_720p_x264_AAC_SRT_in_mkv

And I re-encoded without error and subtitles of course, I suspect chapters (or time stamps labels ) - i remove them and cut small clip to test. I have one suggestion why ffmpeg2theora when file have one path (or Stream) didn't mark him as default, this is really piss me of when I must select subtitles -cy

And on console apeered this:

  Metadata:
    doctype : matroska
    title : Hime Chen! Otogi Chikku Idol Lilpri - 11
  Duration: 00:20:59.81, start: 0.000000, bitrate: N/A
    Chapter #0.0: start 0.000000, end 64.898000
    Metadata:
      title : Prologue
    Chapter #0.1: start 0.000000, end 90.885000
    Metadata:
      title : Opening
    Chapter #0.2: start 64.939000, end 1209.792000
    Metadata:
      title : Episode
    Chapter #0.3: start 1209.834000, end 1259.707000
    Metadata:
      title : Preview
    Stream #0.0: Video: h264, yuv420p, 1280x720 [PAR 1:1 DAR 16:9], 23.98 fps, 2
3.98 tbr, 1k tbn, 47.95 tbc
    Metadata:
      title : Hime Chen! Otogi Chikku Idol Lilpri - 11
    Stream #0.1(jpn): Audio: aac, 48000 Hz, stereo, s16
    Metadata:
      title : 2.0 AAC
    Stream #0.2(eng): Subtitle: 0x0000
    Metadata:
      title : English
    Stream #0.3: Attachment: 0x0000
    Metadata:
      filename : ITCAvantGardeStd-Bk.otf
    Stream #0.4: Attachment: 0x0000
    Metadata:
      filename : Essai.ttf
    Stream #0.5: Attachment: 0x0000
    Metadata:
      filename : Vera.ttf
  [subtitles disabled].
  Pixel Aspect Ratio: 1.19/1 Frame Aspect Ratio: 1.78/1
  Resize: 1280x720 => 720x480
  0:00:00.92 audio: 43kbps video: 1138kbps, ET: 00:45:20, est. size: 129.7 MB
  0:00:01.58 audio: 59kbps video: 1117kbps, ET: 00:52:54, est. size: 114.9 MB
  0:00:04.21 audio: 66kbps video: 1035kbps, ET: 00:39:44, est. size: 165.1 MB
  0:20:59.29 audio: 72kbps video: 746kbps, ET: 00:00:00, est. size: 122.9 MB [
matroska @ 0x3ebf70]Unknown entry 0x2D1441
  0:20:59.81 audio: 72kbps video: 746kbps, time elapsed: 00:38:32
  0:20:59.81 audio: 72kbps video: 746kbps, time elapsed: 00:38:32

Hi,

I've tried your command parameters and your input file, and it converts fine for me. Subtitles seem to be included in the output properly, and the resulting file plays fine in VLC.
I don't suppose Windows gives any kind of useful information about the crash, like a stack trace ?

Thanks I try search for something more useful :

1) I always get error exacly on 14,63ms:
  0:00:00.92 audio: 43kbps video: 1138kbps, ET: 00:45:20, est. size: 129.7 MB
  0:00:01.58 audio: 59kbps video: 1117kbps, ET: 00:39:41, est. size: 114.9 MB
  0:00:03.04 audio: 65kbps video: 584kbps, ET: 00:41:16, est. size: 97.3 MB Fa
iled to encode kate data packet (3.640000 --> 4.980000, [A wedding?!]): -2
  0:00:04.21 audio: 66kbps video: 1035kbps, ET: 00:39:44, est. size: 165.1 MB
  0:00:04.87 audio: 67kbps video: 988kbps, ET: 00:38:34, est. size: 157.5 MB F
ailed to encode kate data packet (5.350000 --> 10.940000, [Yeah. The lady who us
ed to be my neighbor is getting married.]): -2
  0:00:10.55 audio: 70kbps video: 549kbps, ET: 00:35:31, est. size: 92.8 MB Fa
iled to encode kate data packet (10.940000 --> 13.450000, [She invited me to the
 ceremony.
]): -2
  0:00:13.68 audio: 71kbps video: 472kbps, ET: 00:36:24, est. size: 80.6 MB Fa
iled to encode kate data packet (14.570000 --> 15.570000, [How nice.]): -2
  0:00:14.63 audio: 71kbps video: 453kbps, ET: 00:35:26, est. size: 78.3 MB

How can i compile this version and run i debug mode to provide useful information - Visual Studio i enought or Dev-C++??

Hi I remember i install Debugging Tools for Windows (x86) for BSOD ( btw. who is caused by motherboard Gigashit witch I successful touted and relieve with Asrock). I run open executable and get without any flags and with file and flags - in zip i place rapports and screenshots of windows witch registers (there too many to rewrite and you can't copy them all :/) for try i run also previous version and i catch one difrents :

0.26
Executable search path is:
ModLoad: 00400000 00e15000 image00400000
ModLoad: 7c900000 7c9b2000 ntdll.dll
ModLoad: 7c800000 7c8f6000 C:\WINDOWS\system32\kernel32.dll
ModLoad: 77c10000 77c68000 C:\WINDOWS\system32\msvcrt.dll
(c84.4b0): Break instruction exception - code 80000003 (first chance)
eax=00341eb4 ebx=7ffdf000 ecx=00000004 edx=00000010 esi=00341f48 edi=00341eb4
eip=7c90120e esp=0022fb20 ebp=0022fc94 iopl=0 nv up ei pl nz na po nc
cs=001b ss=0023 ds=0023 es=0023 fs=003b gs=0000 efl=00000202
*** ERROR: Symbol file could not be found. Defaulted to export symbols for ntdll.dll -
ntdll!DbgBreakPoint:
7c90120e cc int 3

this version:
Executable search path is:
ModLoad: 00400000 011f9000 image00400000
ModLoad: 7c900000 7c9b2000 ntdll.dll
ModLoad: 7c800000 7c8f6000 C:\WINDOWS\system32\kernel32.dll
ModLoad: 73b50000 73b67000 C:\WINDOWS\system32\AVIFIL32.DLL
ModLoad: 77c10000 77c68000 C:\WINDOWS\system32\msvcrt.dll
ModLoad: 7e410000 7e4a1000 C:\WINDOWS\system32\USER32.dll
ModLoad: 77f10000 77f59000 C:\WINDOWS\system32\GDI32.dll
ModLoad: 76b40000 76b6d000 C:\WINDOWS\system32\WINMM.dll
ModLoad: 77dd0000 77e6b000 C:\WINDOWS\system32\ADVAPI32.dll
ModLoad: 77e70000 77f02000 C:\WINDOWS\system32\RPCRT4.dll
ModLoad: 77fe0000 77ff1000 C:\WINDOWS\system32\Secur32.dll
ModLoad: 774e0000 7761d000 C:\WINDOWS\system32\ole32.dll
ModLoad: 77be0000 77bf5000 C:\WINDOWS\system32\MSACM32.dll
ModLoad: 75a70000 75a91000 C:\WINDOWS\system32\MSVFW32.dll
ModLoad: 7c9c0000 7d1d7000 C:\WINDOWS\system32\SHELL32.dll
ModLoad: 77f60000 77fd6000 C:\WINDOWS\system32\SHLWAPI.dll
ModLoad: 5d090000 5d12a000 C:\WINDOWS\system32\COMCTL32.dll
(f80.9c4): Break instruction exception - code 80000003 (first chance)
eax=00341eb4 ebx=7ffdf000 ecx=00000001 edx=00000002 esi=00341f48 edi=00341eb4
eip=7c90120e esp=0022fb20 ebp=0022fc94 iopl=0 nv up ei pl nz na po nc
cs=001b ss=0023 ds=0023 es=0023 fs=003b gs=0000 efl=00000202
*** ERROR: Symbol file could not be found. Defaulted to export symbols for ntdll.dll -
ntdll!DbgBreakPoint:
7c90120e cc int 3

I also add Analyse - i try use the same thing i was doing when i try check system with BSOD - cy

Do you get other errors before that ?
It looks like the encoder for subtitles fails to initialize, but there should be a message before that if this is the case, which would help me work out why. As for building ffmpeg2theora on Windows, I don't know how, I think j cross compiles from Linux.

I will try build this in VC but i need complete code with all lib statics - because in windows add this libs is much harder -> console is poor and i don't have configure - and for first question i didn't get any errors before that -cy.

On 07/01/2010 09:05 AM, Adam B. wrote:
> I will try build this in VC but i need complete code with all lib
> statics - because in windows add this libs is much harder -> console is
> poor and i don't have configure - and for first question i didn't get
> any errors before that -cy.
>

i dont think you can build this in VC, using mingw32 you might be able
to get a build done, but even that is rather hard.

can you try this:
 http://firefogg.org/nightly/ffmpeg2theora-debug.exe
and see if you get better error messages or tracebacks

A valgrind run finds nothing of interest, apart from the usual machine specific things, and the usual ffmpeg bugs. I'm on a 64 bit architecture though, are you ? My 32 bit machine is out of commission though :(

I have 64 bit processor but old windows XP 32 bit - and Ubuntu 64 bit (for AMD 10.4 :) - now i run i Windows Debugger this version from post j^ - but befor that i try this release and i re-encoded all this file in console with all flags but in VLC 1.1.0 i have subtitles but chapters or time stamps labels disappeared , in Media Player Classic i have only audio and video (perhaps new format libkate).

In console:
C:\Documents and Settings\xer\Desktop>ffmpeg2theora-debug.exe --optimize --speed
level 0 -a 1 --aspect 16:9 -x 720 -y 480 "[Chihiro]_Hime_Chen!_Otogi_Chikku_Id
ol_Lilpri_-_11_[1280x720_H.264_AAC][10115E11].mkv"
[matroska @ 0x3ebf40]Unknown entry 0x6E67
    Last message repeated 1 times
[matroska @ 0x3ebf40]max_analyze_duration reached
[matroska @ 0x3ebf40]Estimating duration from bitrate, this may be inaccurate
Input #0, matroska, from '[Chihiro]_Hime_Chen!_Otogi_Chikku_Idol_Lilpri_-_11_[12
80x720_H.264_AAC][10115E11].mkv':
  Metadata:
    doctype : matroska
    title : Hime Chen! Otogi Chikku Idol Lilpri - 11
  Duration: 00:20:59.81, start: 0.000000, bitrate: N/A
    Chapter #0.0: start 0.000000, end 64.898000
    Metadata:
      title : Prologue
    Chapter #0.1: start 0.000000, end 90.885000
    Metadata:
      title : Opening
    Chapter #0.2: start 64.939000, end 1209.792000
    Metadata:
      title : Episode
    Chapter #0.3: start 1209.834000, end 1259.707000
    Metadata:
      title : Preview
    Stream #0.0: Video: h264, yuv420p, 1280x720 [PAR 1:1 DAR 16:9], 23.98 fps, 2
3.98 tbr, 1k tbn, 47.95 tbc
    Metadata:
      title : Hime Chen! Otogi Chikku Idol Lilpri - 11
    Stream #0.1(jpn): Audio: aac, 48000 Hz, stereo, s16
    Metadata:
      title : 2.0 AAC
    Stream #0.2(eng): Subtitle: 0x0000
    Metadata:
      title : English
    Stream #0.3: Attachment: 0x0000
    Metadata:
      filename : ITCAvantGardeStd-Bk.otf
    Stream #0.4: Attachment: 0x0000
    Metadata:
      filename : Essai.ttf
    Stream #0.5: Attachment: 0x0000
    Metadata:
      filename : Vera.ttf
  Pixel Aspect Ratio: 1.19/1 Frame Aspect Ratio: 1.78/1
  Resize: 1280x720 => 720x480
  0:20:59.29 audio: 72kbps video: 746kbps, ET: 00:00:00, est. size: -0.0 MB [m
atroska @ 0x3ebf40]Unknown entry 0x2D1441
  0:20:59.81 audio: 72kbps video: 746kbps, time elapsed: 00:38:27
  0:20:59.81 audio: 72kbps video: 746kbps, time elapsed: 00:38:27

It seams i get full rapport from debugger a few hours later - must do something -cy

You could use -x 72 -y 48 for testing. It should make things much faster.

And yes, ffmeg2theora does not do chapters. I don't know what you call timestamp labels though.

All tracebacks seems to be identical to previous - log and WEW file is on attachment -cy

And yes, ffmeg2theora does not do chapters. I don't know what you call timestamp labels though.

It seems i suppose call them chapters, and good idea a to set this small resolutions - if you need any other information or data i will try provide to you -cy

I've added logs of relevant information which hopefully should help track down the problem. j^ will build a Windows binary when he's got a moment. Thanks for helping to find the reason for that issue.

No problem - i can test other stable releases or rc i future - personally i using yours great work for my younger sister - she got my odl PC ( Athlon 1,73 + 725 MB ram and GF 6200N +...) and for her configuration theora is the best for me and her (performance & re-encode time + and quality+ and free licence) - cy

new debug build up at
 http://firefogg.org/nightly/ffmpeg2theora-debug.exe

can you test that and post the details here.

I tested this build and I get this:

1> In console:

....>ffmpeg2theora-debug.exe --optimize --spee
dlevel 0 -a 1 --aspect 16:9 -x 72 -y 48 "[Chihiro]_Hime_Chen!_Otogi_Chikku_Ido
l_Lilpri_-_11_[1280x720_H.264_AAC][10115E11].mkv"

[matroska @ 0x3ebf40]Unknown entry 0x6E67
    Last message repeated 1 times
[matroska @ 0x3ebf40]max_analyze_duration reached
[matroska @ 0x3ebf40]Estimating duration from bitrate, this may be inaccurate
Input #0, matroska, from '[Chihiro]_Hime_Chen!_Otogi_Chikku_Idol_Lilpri_-_11_[12
80x720_H.264_AAC][10115E11].mkv':
  Metadata:
    doctype : matroska
    title : Hime Chen! Otogi Chikku Idol Lilpri - 11
  Duration: 00:20:59.81, start: 0.000000, bitrate: N/A
    Chapter #0.0: start 0.000000, end 64.898000
    Metadata:
      title : Prologue
    Chapter #0.1: start 0.000000, end 90.885000
    Metadata:
      title : Opening
    Chapter #0.2: start 64.939000, end 1209.792000
    Metadata:
      title : Episode
    Chapter #0.3: start 1209.834000, end 1259.707000
    Metadata:
      title : Preview
    Stream #0.0: Video: h264, yuv420p, 1280x720 [PAR 1:1 DAR 16:9], 23.98 fps, 2
3.98 tbr, 1k tbn, 47.95 tbc
    Metadata:
      title : Hime Chen! Otogi Chikku Idol Lilpri - 11
    Stream #0.1(jpn): Audio: aac, 48000 Hz, stereo, s16
    Metadata:
      title : 2.0 AAC
    Stream #0.2(eng): Subtitle: 0x0000
    Metadata:
      title : English
    Stream #0.3: Attachment: 0x0000
    Metadata:
      filename : ITCAvantGardeStd-Bk.otf
    Stream #0.4: Attachment: 0x0000
    Metadata:
      filename : Essai.ttf
    Stream #0.5: Attachment: 0x0000
    Metadata:
      filename : Vera.ttf
  Pixel Aspect Ratio: 1.19/1 Frame Aspect Ratio: 1.78/1
  Resize: 1280x720 => 72x48
src/theorautils.c:861: Initializing 1 Kate streams
src/theorautils.c:864: Kate stream 0: 0x2adec40
src/theorautils.c:866: init kate stream
src/kate_encode.c:39: kate_encode_init: state 02ADEC40, info 02ADEC60
src/theorautils.c:868: kate_encode_init (state 0x2adec40, info 0x2adec60): 0
src/theorautils.c:874: kate_comment_init: 0
src/theorautils.c:879: Adding tag: ENCODER=ffmpeg2theora-0.27
src/theorautils.c:994: Encoding Kate headers
src/theorautils.c:997: Kate headers for stream 0 (0x2adec40): state 0x2adec40, i
nfo 0x2adec60
src/theorautils.c:1001: kate_ogg_encode_headers: ret 0
src/theorautils.c:1001: kate_ogg_encode_headers: ret 0
src/theorautils.c:1001: kate_ogg_encode_headers: ret 0
src/theorautils.c:1001: kate_ogg_encode_headers: ret 0
src/theorautils.c:1001: kate_ogg_encode_headers: ret 0
src/theorautils.c:1001: kate_ogg_encode_headers: ret 0
src/theorautils.c:1001: kate_ogg_encode_headers: ret 0
src/theorautils.c:1001: kate_ogg_encode_headers: ret 0
src/theorautils.c:1001: kate_ogg_encode_headers: ret 0
src/theorautils.c:1001: kate_ogg_encode_headers: ret 1
src/theorautils.c:1065: Flushing Kate streams
src/theorautils.c:1068: Flushing Kate stream 0 (0x2adec40)
  0:00:02.52 audio: 64kbps video: 15kbps, ET: 00:24:55, est. size: 0.0 MB src/
theorautils.c:1320: Adding text to Kate stream 0 (0x2adec40): "A wedding?!" (len
gth 11) from 3.640000 to 4....

j^ uploaded another binary at the same place, with more logs around the place of interest.
Can you please post the logs from this new binary ?

****************************************************
Console:
***************************************************

C:\Documents and Settings\xer\Desktop>"ffmpeg2theora-debug(2).exe" --optimize --speedlevel 0 -a 1 --aspect 16:9 -x 72 -y 48 "[Chihiro]_Hime_Chen!_Otogi_Chikku_Idol_Lilpri_-_11_[1280x720_H.264_AAC][10115E11].mkv"
[matroska @ 0x3ebf40]Unknown entry 0x6E67
    Last message repeated 1 times
[matroska @ 0x3ebf40]max_analyze_duration reached
[matroska @ 0x3ebf40]Estimating duration from bitrate, this may be inaccurate
Input #0, matroska, from '[Chihiro]_Hime_Chen!_Otogi_Chikku_Idol_Lilpri_-_11_[12
80x720_H.264_AAC][10115E11].mkv':
  Metadata:
    doctype : matroska
    title : Hime Chen! Otogi Chikku Idol Lilpri - 11
  Duration: 00:20:59.81, start: 0.000000, bitrate: N/A
    Chapter #0.0: start 0.000000, end 64.898000
    Metadata:
      title : Prologue
    Chapter #0.1: start 0.000000, end 90.885000
    Metadata:
      title : Opening
    Chapter #0.2: start 64.939000, end 1209.792000
    Metadata:
      title : Episode
    Chapter #0.3: start 1209.834000, end 1259.707000
    Metadata:
      title : Preview
    Stream #0.0: Video: h264, yuv420p, 1280x720 [PAR 1:1 DAR 16:9], 23.98 fps, 2
3.98 tbr, 1k tbn, 47.95 tbc
    Metadata:
      title : Hime Chen! Otogi Chikku Idol Lilpri - 11
    Stream #0.1(jpn): Audio: aac, 48000 Hz, stereo, s16
    Metadata:
      title : 2.0 AAC
    Stream #0.2(eng): Subtitle: 0x0000
    Metadata:
      title : English
    Stream #0.3: Attachment: 0x0000
    Metadata:
      filename : ITCAvantGardeStd-Bk.otf
    Stream #0.4: Attachment: 0x0000
    Metadata:
      filename : Essai.ttf
    Stream #0.5: Attachment: 0x0000
    Metadata:
      filename : Vera.ttf
  Pixel Aspect Ratio: 1.19/1 Frame Aspect Ratio: 1.78/1
  Resize: 1280x720 => 72x48
src/theorautils.c:861: Initializing 1 Kate streams
src/theorautils.c:864: Kate stream 0: 0x29f9af0
src/theorautils.c:866: init kate stream
src/kate_encode.c:39: kate_encode_init: state 029F9AF0, info 029F9B10
src/theorautils.c:868: kate_encode_init (state 0x29f9af0, info 0x29f9b10): 0
src/theorautils.c:874: kate_comment_init: 0
src/theorautils.c:879: Adding tag: ENCODER=ffmpeg2theora-0.27
src/theorautils.c:994: Encoding Kate headers
src/theorautils.c:997: Kate headers for stream 0 (0x29f9af0): state 0x29f9af0, i
nfo 0x29f9b10
src/kate_ogg.c:36: Copying zu bytes from 00000040 to 02AF1FC0
src/theorautils.c:1001: kate_ogg_encode_headers: ret 0
src/kate_ogg.c:36: Copying zu bytes from 00000044 to 032814A0
src/theorautils.c:1001: kate_ogg_encode_headers: ret 0
src/kate_ogg.c:36: Copying zu bytes from 0000000A to 03271290
src/theorautils.c:1001: kate_ogg_encode_headers: ret 0
src/kate_ogg.c:36: Copying zu bytes from 0000000A to 02AF1FA8
src/theorautils.c:1001: kate_ogg_encode_headers: ret 0
src/kate_ogg.c:36: Copying zu bytes from 0000000A to 03271290
src/theorautils.c:1001: kate_ogg_encode_headers: ret 0
src/kate_ogg.c:36: Copying zu bytes from 0000000A to 02AF1FA8
src/theorautils.c:1001: kate_ogg_encode_headers: ret 0
src/kate_ogg.c:36: Copying zu bytes from 00000...

Missing log from windows debuger.

Thanks - another build is ready for followup, same place again.

Ok i will test that - i just return from garden => hot whether => you must give some water on flowers -i will post result after i download this and test -cy

C:\Documents and Settings\xer\Desktop>ffmpeg2theora-debug.exe --optimize --speed
level 0 -a 1 --aspect 16:9 -x 72 -y 48 "[Chihiro]_Hime_Chen!_Otogi_Chikku_Idol
_Lilpri_-_11_[1280x720_H.264_AAC][10115E11].mkv"
[matroska @ 0x3ebf40]Unknown entry 0x6E67
    Last message repeated 1 times
[matroska @ 0x3ebf40]max_analyze_duration reached
[matroska @ 0x3ebf40]Estimating duration from bitrate, this may be inaccurate
Input #0, matroska, from '[Chihiro]_Hime_Chen!_Otogi_Chikku_Idol_Lilpri_-_11_[12
80x720_H.264_AAC][10115E11].mkv':
  Metadata:
    doctype : matroska
    title : Hime Chen! Otogi Chikku Idol Lilpri - 11
  Duration: 00:20:59.81, start: 0.000000, bitrate: N/A
    Chapter #0.0: start 0.000000, end 64.898000
    Metadata:
      title : Prologue
    Chapter #0.1: start 0.000000, end 90.885000
    Metadata:
      title : Opening
    Chapter #0.2: start 64.939000, end 1209.792000
    Metadata:
      title : Episode
    Chapter #0.3: start 1209.834000, end 1259.707000
    Metadata:
      title : Preview
    Stream #0.0: Video: h264, yuv420p, 1280x720 [PAR 1:1 DAR 16:9], 23.98 fps, 2
3.98 tbr, 1k tbn, 47.95 tbc
    Metadata:
      title : Hime Chen! Otogi Chikku Idol Lilpri - 11
    Stream #0.1(jpn): Audio: aac, 48000 Hz, stereo, s16
    Metadata:
      title : 2.0 AAC
    Stream #0.2(eng): Subtitle: 0x0000
    Metadata:
      title : English
    Stream #0.3: Attachment: 0x0000
    Metadata:
      filename : ITCAvantGardeStd-Bk.otf
    Stream #0.4: Attachment: 0x0000
    Metadata:
      filename : Essai.ttf
    Stream #0.5: Attachment: 0x0000
    Metadata:
      filename : Vera.ttf
  Pixel Aspect Ratio: 1.19/1 Frame Aspect Ratio: 1.78/1
  Resize: 1280x720 => 72x48
src/theorautils.c:861: Initializing 1 Kate streams
src/theorautils.c:864: Kate stream 0: 0x2adec40
src/theorautils.c:866: init kate stream
src/kate_encode.c:39: kate_encode_init: state 02ADEC40, info 02ADEC60
src/theorautils.c:868: kate_encode_init (state 0x2adec40, info 0x2adec60): 0
src/theorautils.c:874: kate_comment_init: 0
src/theorautils.c:879: Adding tag: ENCODER=ffmpeg2theora-0.27
src/theorautils.c:994: Encoding Kate headers
src/theorautils.c:997: Kate headers for stream 0 (0x2adec40): state 0x2adec40, i
nfo 0x2adec60
src/kate_ogg.c:36: Copying 64 bytes from 00000000 to 032AA760
src/theorautils.c:1001: kate_ogg_encode_headers: ret 0
src/kate_ogg.c:36: Copying 68 bytes from 00000000 to 032B9708
src/theorautils.c:1001: kate_ogg_encode_headers: ret 0
src/kate_ogg.c:36: Copying 10 bytes from 00000000 to 0329E520
src/theorautils.c:1001: kate_ogg_encode_headers: ret 0
src/kate_ogg.c:36: Copying 10 bytes from 00000000 to 0329E328
src/theorautils.c:1001: kate_ogg_encode_headers: ret 0
src/kate_ogg.c:36: Copying 10 bytes from 00000000 to 0329E520
src/theorautils.c:1001: kate_ogg_encode_headers: ret 0
src/kate_ogg.c:36: Copying 10 bytes from 00000000 to 0329E328
src/theorautils.c:1001: kate_ogg_encode_headers: ret 0
src/kate_ogg.c:36: Copying 10 bytes from 00000000 to 0329E520
src/theorautils.c:1001: kate_ogg_encode_headers: ret 0
src/kate_ogg.c:36: Copying 10 bytes from 000000...

Hmm, it looks like this libc somehow doesn't like printing either size_t nor long long values :/

I won't need more Windows debug reports, thanks. Just the logs for now.

New binary again, same place, I'll need only the logs please.

Sorry this late test - I don't have Internet for few days, my ISP have one big break (3 days) and after 24 of this mouth i will alsow don't have net - my agreement is expired but i will find new ISp (this is to often nonoperational) but i try test others build using connection of my friend -cy.

C:\Documents and Settings\xer\Desktop>"ffmpeg2theora-debug(2).exe" --optimize --
speedlevel 0 -a 1 --aspect 16:9 -x 72 -y 48 "[Chihiro]_Hime_Chen!_Otogi_Chikku
_Idol_Lilpri_-_11_[1280x720_H.264_AAC][10115E11].mkv"
[matroska @ 0x3bf00]Unknown entry 0x6E67
    Last message repeated 1 times
[matroska @ 0x3bf00]max_analyze_duration reached
[matroska @ 0x3bf00]Estimating duration from bitrate, this may be inaccurate
Input #0, matroska, from '[Chihiro]_Hime_Chen!_Otogi_Chikku_Idol_Lilpri_-_11_[12
80x720_H.264_AAC][10115E11].mkv':
  Metadata:
    doctype : matroska
    title : Hime Chen! Otogi Chikku Idol Lilpri - 11
  Duration: 00:20:59.81, start: 0.000000, bitrate: N/A
    Chapter #0.0: start 0.000000, end 64.898000
    Metadata:
      title : Prologue
    Chapter #0.1: start 0.000000, end 90.885000
    Metadata:
      title : Opening
    Chapter #0.2: start 64.939000, end 1209.792000
    Metadata:
      title : Episode
    Chapter #0.3: start 1209.834000, end 1259.707000
    Metadata:
      title : Preview
    Stream #0.0: Video: h264, yuv420p, 1280x720 [PAR 1:1 DAR 16:9], 23.98 fps, 2
3.98 tbr, 1k tbn, 47.95 tbc
    Metadata:
      title : Hime Chen! Otogi Chikku Idol Lilpri - 11
    Stream #0.1(jpn): Audio: aac, 48000 Hz, stereo, s16
    Metadata:
      title : 2.0 AAC
    Stream #0.2(eng): Subtitle: 0x0000
    Metadata:
      title : English
    Stream #0.3: Attachment: 0x0000
    Metadata:
      filename : ITCAvantGardeStd-Bk.otf
    Stream #0.4: Attachment: 0x0000
    Metadata:
      filename : Essai.ttf
    Stream #0.5: Attachment: 0x0000
    Metadata:
      filename : Vera.ttf
  Pixel Aspect Ratio: 1.19/1 Frame Aspect Ratio: 1.78/1
  Resize: 1280x720 => 72x48
src/theorautils.c:861: Initializing 1 Kate streams
src/theorautils.c:864: Kate stream 0: 0x2c87838
src/theorautils.c:866: init kate stream
src/kate_encode.c:39: kate_encode_init: state 02C87838, info 02C87858
src/theorautils.c:868: kate_encode_init (state 0x2c87838, info 0x2c87858): 0
src/theorautils.c:874: kate_comment_init: 0
src/theorautils.c:879: Adding tag: ENCODER=ffmpeg2theora-0.27
src/theorautils.c:994: Encoding Kate headers
src/theorautils.c:997: Kate headers for stream 0 (0x2c87838): state 0x2c87838, i
nfo 0x2c87858
src/kate_ogg.c:36: Copying 64 bytes from 00000000 to 02A89FC0
src/theorautils.c:1001: kate_ogg_encode_headers: ret 0
src/kate_ogg.c:36: Copying 68 bytes from 00000000 to 02B39B38
src/theorautils.c:1001: kate_ogg_encode_headers: ret 0
src/kate_ogg.c:36: Copying 10 bytes from 00000000 to 02B3E340
src/theorautils.c:1001: kate_ogg_encode_headers: ret 0
src/kate_ogg.c:36: Copying 10 bytes from 00000000 to 02B3E568
src/theorautils.c:1001: kate_ogg_encode_headers: ret 0
src/kate_ogg.c:36: Copying 10 bytes from 00000000 to 02B3E340
src/theorautils.c:1001: kate_ogg_en...

Thanks, and aopologies, this build was actually not up to date. This is now fixed, if you can download and run again - sorry!

This is not very big problem - i have the same problems when i have more that one copy of file or catalog (witch one is latest/correct ) - I tested this version and something one big flood (windows cmd can only display 100 or 180 lines so i found solution using redirect "> ffmpeg.log 2>&1" ) and created this relay long log so i compressed this using zip and posted in attachment :D -cy.

Well, it looks like this log is much larger because it did not crash this time.
This log and the previous ones don't show anything that appears wrong, so I'm at a loss thinking of a reason for the crash. I'll post again if I get a new idea. Thanks for the help so far!

But when i encoding material twice i hear beep - i will tray witch another file and encode this with 480p resolution to test (icon is too small to see subtitles :P)-cy.

Ok i tested this with full 480p with the same source as previews and in the end i get near 14 sec error :

src/kate_granule.c:117: Result float 240710.506714, int64_t 240710
src/kate_encode.c:890: kate_encode_text_raw_times: Stack: 0023E744
src/kate_encode.c:891: kate_encode_text_raw_times: 02C8AAC0, 240710 --> 244010,
I forgot my aunts were coming today., 36, 0023E844
src/kate_encode.c:898: kate_text_validate: 0
src/kate_encode.c:902: kate_encode_state_trim_events: 0
src/kate_encode.c:905: kate_encode_state_add_event: 0
src/kate_encode.c:908: kate_encode_state_get_earliest_event: 0
src/kate_ogg.c:78: kate_ogg_encode_text: 0
src/kate_ogg.c:80: kate_ogg_encode_text: kes 02A8EF60
src/kate_ogg.c:36: Copying 70 bytes from 03226EC0 to 02B40E88
src/kate_ogg.c:82: kate_ogg_encode_text done
src/theorautils.c:1323: kate_ogg_encode_text: ret 0
  0:04:03.58 audio: 72kbps video: 444kbps, ET: 00:33:10, est. size: 0.0 MB src
/theorautils.c:1320: Adding text to Kate stream 0 (0x2c8aac0): "Aunts? Is that i
nconvenient?" (length 28) from 244.009995 to 248.219995
src/kate_ogg.c:76: kate_ogg_encode_text: 02C8AAC0, 244.009995 --> 248.220001, Au
nts? Is that inconvenient?, 28, 0023E8B8, stack kp 0023E844
src/kate_encode.c:991: size_t: 4, kate_int64_t: 8, int: 4, long: 4, long long: 8

src/kate_encode.c:992: kate_encode_text: 02C8AAC0, 244.009995 --> 248.220001, Au
nts? Is that inconvenient?, 28, 0023E844, info 02C8AAE0
src/kate_granule.c:115: kate_duration_granule: ki 02C8AAE0, d 248.220001, gps 10
00 / 1
src/kate_granule.c:117: Result float 248220.501221, int64_t 248220
src/kate_granule.c:115: kate_duration_granule: ki 02C8AAE0, d 244.009995, gps 10
00 / 1
src/kate_granule.c:117: Result float 244010.494507, int64_t 244010
src/kate_encode.c:993: kate_encode_text: granules 244010, 248220
src/kate_encode.c:994: Stack: 0023E814
src/kate_granule.c:115: kate_duration_granule: ki 02C8AAE0, d 248.220001, gps 10
00 / 1
src/kate_granule.c:117: Result float 248220.501221, int64_t 248220
src/kate_granule.c:115: kate_duration_granule: ki 02C8AAE0, d 244.009995, gps 10
00 / 1
src/kate_granule.c:117: Result float 244010.494507, int64_t 244010
src/kate_encode.c:890: kate_encode_text_raw_times: Stack: 0023E744
src/kate_encode.c:891: kate_encode_text_raw_times: 02C8AAC0, 244010 --> 248220,
Aunts? Is that inconvenient?, 28, 0023E844
src/kate_encode.c:898: kate_text_validate: 0
src/kate_encode.c:902: kate_encode_state_trim_events: 0
src/kate_encode.c:905: kate_encode_state_add_event: 0
src/kate_encode.c:908: kate_encode_state_get_earliest_event: 0
src/kate_ogg.c:78: kate_ogg_encode_text: 0
src/kate_ogg.c:80: kate_ogg_encode_text: kes 02A8EF60
src/kate_ogg.c:36: Copying 62 bytes from 02BDBE58 to 02BEB5C8
src/kate_ogg.c:82: kate_ogg_encode_text done
src/theorautils.c:1323: kate_ogg_encode_text: ret 0
  0:04:07.95 audio: 72kbps video: 437kbps, ET: 00:32:51, est. size: 0.0 MB src
/theorautils.c:1320: Adding text to Kate stream 0 (0x2c8aac0): "

Cy

I this whole ogv i found its only 4 mints and a few seconds - i exports subtitles patch and compare them (only 31 rows in ogv ) and stopped at this line :

ogv
M»ŚS«„C¶uS¬‚!PM»ŚS«„C¶uS¬‚!ĽM»ŚS«„C¶uS¬‚"M»ŚS«„C¶uS¬‚"qM»ŚS«„C¶uS¬‚"ĽM»ŚS«„C¶uS¬‚#M»ŚS«„C¶uS¬‚#hM»ŚS«„C¶uS¬‚#¬M»ŚS«„C¶uS¬‚#üM»ŚS«„C¶uS¬‚$eM»ŚS«„C¶uS¬‚$ŘM»ŚS«„C¶uS¬‚%JM»ŚS«„C¶uS¬‚%ŔM»ŚS«„C¶uS¬‚&=M»ŚS«„C¶uS¬‚&´M»ŚS«„C¶uS¬‚'
M»ŚS«„C¶uS¬‚'rM»ŚS«„C¶uS¬‚(M»ŚS«„C¶uS¬‚(WM»ŚS«„C¶uS¬‚(¦M»ŚS«„C¶uS¬‚)+M»ŚS«„C¶uS¬‚)vM»ŚS«„C¶uS¬‚*M»ŚS«„C¶uS¬‚*bM»ŚS«„C¶uS¬‚*ˇM»ŚS«„C¶uS¬‚*éM»ŚS«„C¶uS¬‚+E

mkv (in attachment as screenshot+ two files)

I watch this whole ogv i found its only 4 mints and a few seconds - i exports subtitles patch and compare them (only 31 rows in ogv ) and stopped at this line :

ogv
M»ŚS«„C¶uS¬‚!PM»ŚS«„C¶uS¬‚!ĽM»ŚS«„C¶uS¬‚"M»ŚS«„C¶uS¬‚"qM»ŚS«„C¶uS¬‚"ĽM»ŚS«„C¶uS¬‚#M»ŚS«„C¶uS¬‚#hM»ŚS«„C¶uS¬‚#¬M»ŚS«„C¶uS¬‚#üM»ŚS«„C¶uS¬‚$eM»ŚS«„C¶uS¬‚$ŘM»ŚS«„C¶uS¬‚%JM»ŚS«„C¶uS¬‚%ŔM»ŚS«„C¶uS¬‚&=M»ŚS«„C¶uS¬‚&´M»ŚS«„C¶uS¬‚'
M»ŚS«„C¶uS¬‚'rM»ŚS«„C¶uS¬‚(M»ŚS«„C¶uS¬‚(WM»ŚS«„C¶uS¬‚(¦M»ŚS«„C¶uS¬‚)+M»ŚS«„C¶uS¬‚)vM»ŚS«„C¶uS¬‚*M»ŚS«„C¶uS¬‚*bM»ŚS«„C¶uS¬‚*ˇM»ŚS«„C¶uS¬‚*éM»ŚS«„C¶uS¬‚+E

mkv (in attachment as screenshot+ two files)

Hi,

there's a new build at the same place. This one has a change that might fix the issue, rather unlikely but it's the only thing I can see that might be wrong.

Thanks

Hi i back on line - my new ISP is small but seems much more reliable that TPSA - this version give me the same result as previous only few part subtitles and 2 beeps in moments when encodings (and 1 funny thing => when i tray use redirecting all cmd and this shit is Called Operating System pretty hang up with Task manager) - maybe the fonts in attachment of mkv causes problems - cy (i will try another day test this propely).

Hi i think fund root of this problem -> i was trying use recent build and get info about msvcrt.dll

http://stackoverflow.com/questions/135296/is-msvcrt-under-windows-like-glibc-libc-under-nix

i think this is M$ guilty or to slow porting - for now i will wait for next version - cy

Hi,

since ffmpeg2theora does not crash when subtitles are disabled, I'd be inclined to think
this is not the problem, but rather a bug in my code. The place where it crashes does
not call any libc API. It could be a previous memory scribbling though, so still possible.

Idon't suppose you have MSVC or another debugger on your machine ?
Running ffmpeg2theora inside MSVC might help find better info about the crash.

I will today install MSVC and tray run (I have from MSDN'a 2005 Pro version ) or jut i will install 2008 express - i think will be enough -cy.

Thanks, your continued help is much appreciated :)

Ok when i attach to process i get :

Unhandled exception at 0x008f229c in ffmpeg2theora-0.27.exe: 0xC0000005: Access violation reading location 0x9e000000.\

Diassembler show me this break point :

008F229C mov al,byte ptr [eax]
And i can't compile this because i don't heave not linked sources, or i don't know how properly add all need code to new project. I can't post in attachment debug log here, because it is too big, but you can find it at posted links below -cy.

http://www.megaupload.com/?d=YZ2C4P1M
or
http://www.mediafire.com/?945y5va11wzwa3t

Thanks.
Another attempt is ready, binary's at the same place again.
If it crashes again, it'd help if you'd give me a few lines of disassembly before and after the instruction where it breaks (this time was OK as I could find it).
Thanks for your help :)

Hi i have good news seams bug was squashed, this version recompenses properly and subtitles was in full episode - only Media Home Player Classic latest don't read it probably because he have too old libkate support, but VLC runs it flawless (only fonts was incorrect) - i think you can close this threats thanks for your hard work - cy.

That's great!
If you don't mind, I'd like to remove all the logs and checks I've added so far, and leave just
what I think is the fix, and have you try the 'normal' non-debug build, just to make sure this is
actually fixed. I'll post again when it's ready.

About Media Home Player Classic: I doubt this has support for kate streams. Feel free to
ask them for it though :)

Thanks

Ready:
http://firefogg.org/nightly/ffmpeg2theora-debug.exe
Despite the name, it's a normal build, not a debug one, this time :)
If this one also works, then I'll be very happy :)

This version is broken you forget to compile in static libkate-1.dll :P.

On 08/25/2010 04:25 PM, Adam B. wrote:
> This version is broken you forget to compile in static libkate-1.dll :P.
>
ups, uploaded a new version with static libkate

This time you forget liboggkate-1.dll.

not my day, i hope this time without any dll dependencies

This version is bugy :

008F21CB mov eax,0FFFFFFFFh
008F21D0 leave
008F21D1 ret
008F21D2 push ebp
008F21D3 mov ebp,esp
008F21D5 sub esp,10h
008F21D8 cmp dword ptr [ebp+8],0
008F21DC jne 008F21E8
008F21DE mov eax,0FFFFFFFEh
008F21E3 jmp 008F2416
008F21E8 mov dword ptr [ebp-4],0
008F21EF mov eax,dword ptr [ebp+8]
008F21F2 mov al,byte ptr [eax] <- break point
008F21F4 test al,al
008F21F6 js 008F2215
008F21F8 mov eax,dword ptr [ebp+8]
008F21FB mov al,byte ptr [eax]
008F21FD movsx eax,al
008F2200 mov dword ptr [ebp-4],eax
008F2203 mov eax,dword ptr [ebp+0Ch]
008F2206 mov edx,dword ptr [ebp-4]
008F2209 mov dword ptr [eax],edx
008F220B mov eax,1
008F2210 jmp 008F2416
008F2215 mov eax,dword ptr [ebp+8]
008F2218 mov al,byte ptr [eax]
008F221A movsx eax,al
008F221D and eax,0E0h

I think this must be old bulding.

And show (print) variables - so this must be old :P.

OK, can you try this, new version again:

http://firefogg.org/nightly/ffmpeg2theora-debug.exe

Wait, this binary doesn't have symbols either. Something's stripping them out :(

Tray add like this "_0.27.123" :] (this must be one- don't use space ).

008EF57C leave
008EF57D ret
008EF57E push ebp
008EF57F mov ebp,esp
008EF581 sub esp,10h
008EF584 cmp dword ptr [ebp+8],0
008EF588 jne 008EF594
008EF58A mov eax,0FFFFFFFEh
008EF58F jmp 008EF7C2
008EF594 mov dword ptr [ebp-4],0
008EF59B mov eax,dword ptr [ebp+8]
008EF59E mov al,byte ptr [eax]
008EF5A0 test al,al
008EF5A2 js 008EF5C1
008EF5A4 mov eax,dword ptr [ebp+8]
008EF5A7 mov al,byte ptr [eax]
008EF5A9 movsx eax,al
008EF5AC mov dword ptr [ebp-4],eax
008EF5AF mov eax,dword ptr [ebp+0Ch]
008EF5B2 mov edx,dword ptr [ebp-4]
008EF5B5 mov dword ptr [eax],edx
008EF5B7 mov eax,1
008EF5BC jmp 008EF7C2
008EF5C1 mov eax,dword ptr [ebp+8]
008EF5C4 mov al,byte ptr [eax]
008EF5C6 movsx eax,al
008EF5C9 and eax,0E0h
008EF5CE cmp eax,0C0h
008EF5D3 jne 008EF637
008EF5D5 mov eax,dword ptr [ebp+8]

and show

C:\Documents and Settings\xer\Desktop>ffmpeg2theora-debug.exe --optimize --speed
level 0 -a 1 --aspect 16:9 -x 72 -y 48 "[Chihiro]_Hime_Chen!_Otogi_Chikku_Idol
_Lilpri_-_11_[1280x720_H.264_AAC][10115E11].mkv"
[matroska @ 0x3bfe0]Unknown entry 0x6E67
    Last message repeated 1 times
[matroska @ 0x3bfe0]max_analyze_duration reached
[matroska @ 0x3bfe0]Estimating duration from bitrate, this may be inaccurate
Input #0, matroska, from '[Chihiro]_Hime_Chen!_Otogi_Chikku_Idol_Lilpri_-_11_[12
80x720_H.264_AAC][10115E11].mkv':
  Metadata:
    doctype : matroska
    title : Hime Chen! Otogi Chikku Idol Lilpri - 11
  Duration: 00:20:59.81, start: 0.000000, bitrate: N/A
    Chapter #0.0: start 0.000000, end 64.898000
    Metadata:
      title : Prologue
    Chapter #0.1: start 0.000000, end 90.885000
    Metadata:
      title : Opening
    Chapter #0.2: start 64.939000, end 1209.792000
    Metadata:
      title : Episode
    Chapter #0.3: start 1209.834000, end 1259.707000
    Metadata:
      title : Preview
    Stream #0.0: Video: h264, yuv420p, 1280x720 [PAR 1:1 DAR 16:9], 23.98 fps, 2
3.98 tbr, 1k tbn, 47.95 tbc
    Metadata:
      title : Hime Chen! Otogi Chikku Idol Lilpri - 11
    Stream #0.1(jpn): Audio: aac, 48000 Hz, stereo, s16
    Metadata:
      title : 2.0 AAC
    Stream #0.2(eng): Subtitle: 0x0000
    Metadata:
      title : English
    Stream #0.3: Attachment: 0x0000
    Metadata:
      filename : ITCAvantGardeStd-Bk.otf
    Stream #0.4: Attachment: 0x0000
    Metadata:
      filename : Essai.ttf
    Stream #0.5: Attachment: 0x0000
    Metadata:
      filename : Vera.ttf
  Pixel Aspect Ratio: 1.19/1 Frame Aspect Ratio: 1.78/1
  Resize: 1280x720 => 72x48
  0:00:02.52 audio: 64kbps video: 15kbps, ET: 00:24:55, est. size: 11.3 MB Fai
led to encode kate data packet (3.640000 --> 4.980000, [A wedding?!]): -2
  0:00:04.76 audio: 67kbps video: 25kbps, ET: 00:17:33, est. si...

and in this position is break point : 008EF59E mov al,byte ptr [eax].

After the crash, in the debugger, can you try to see what's in memory near the address pointed to by eax ?
If I recall from a previous company when I was working with MSVC, there's a "view memory" window, and you can drag/drop an address on it, so it should show what's there. It'd help a lot if you can make a screenshot of that window showing the data before and after the address in eax, as well as giving the address in eax.
Based on the dissassembly, I'm now suspecting some kind of off by one on the subtitle text, even if valgrind doesn't find one here.
Thanks

Ok i post this window here (i didn't find "view memory" but memory it self window - i copy address and post here screenshot - if this isn't you want i will try find another -cy.

That's the window I wanted - however, it seems to point to the code which crashed.
What I want to see is what the memory contains around the point eax points to.
One of your previous crashes showed:
eax=00341eb4
The value could have changed on this new build though.

It only show "question marks" and dot's - around values have the same lines.

something like this

0x003410EE ?? ?? <some columns> ?? ?? .. .. <some columns> .. ..
0x003410EE ?? ?? <some columns> ?? ?? .. .. <some columns> .. ..
0x00341117 ?? ?? <some columns> ?? ?? .. .. <some columns> .. ..

looks like this was related to my workaround with the printf,
found a better way for that problem,

can you test
 http://firefogg.org/nightly/ffmpeg2theora.exe

and see if that works for you, it seamed to work here with your sample in wine and on windows 7.

Good news work here too - first i reencode to 72x48 and subtitles was unreadable (strange symptoms, i use 1,20 VLC), but after reencode to 480p i watch this episode, and subtitles show correctly :P so i think this build finally works -cy.

Great! :)
As for subtitles in 72x48, that's expected as if you scale the video, I think the subtitles are scaled with it.
Different players do different things there, some will draw subtitles after scaling.
Thanks much for staying with us till it got fixed!