VideoHelp Forum

Our website is made possible by displaying online advertisements to our visitors.
Please consider supporting us by disabling your ad blocker or buy a VSO converter software :)
+ Reply to Thread
Page 11 of 40
FirstFirst ... 9 10 11 12 13 21 ... LastLast
Results 301 to 330 of 1189
Thread
  1. Member
    Join Date
    Mar 2010
    Location
    Italy
    Search PM
    tnx Chris K.
    another question:

    if i set -sameq i also must set the bitrate of video/audio or there is no need to be?
    Quote Quote  
  2. With -sameq, FFmpeg goes entirely its own way. What bitrate it will use depends on the source and may differ widely. It doesn't matter what bitrate you set because it's totally ignored. The FFmpeg docs say about -sameq; "use same video quality as source (implies VBR)".

    You can't even force a higher bitrate than the suggested by simly setting it.

    In my xvid test setup, when I add -sameq, FFmpeg goes up to about 4600 kbps for the given source. When I remove -sameq and set the bitrate higher, it still uses approx. 4600 kbps. I can only "reduce" the bitrate FFmpeg uses by setting it lower (at the cost of lower quality).

    AFAIK, -sameq only works on the video stream so you have to set a audio bitrate.
    Quote Quote  
  3. Member
    Join Date
    Mar 2010
    Location
    Italy
    Search PM
    Ok, tnx2much
    Quote Quote  
  4. Member
    Join Date
    Feb 2010
    Location
    USA
    Search Comp PM
    Is there a way to keep Avanti from copying ffmpeg.exe to ffmp_avanti.exe in the ffmpeg folder when it's started. Doing so creates a significant delay when running from a flash drive, as well as increased wear.

    Also, if I UPX ffmpeg.exe (reduces the size by 2/3) then Avanti thinks it's a 'special build' and won't initialize it. Is there any way around this?

    Many thanks.
    Quote Quote  
  5. There's a history about why Avanti copies and renames FFmpeg.

    Early versions didn't but were restricted to run FFmpeg from the ffmpeg folder at your Avanti install directory. Then there was a user request for the ability to run FFmpeg from any other location because a lot of people have it already somewhere on disk.

    The reason for a renamed copy is; to prevent conflicts with other applications that run FFmpeg simultaneously with Avanti. Because Avanti controls FFmpeg in exceptional cases by terminating the FFmpeg process by name, it would also terminate the other application.

    It could be omitted if the user itself would run a uniquely named FFmpeg version. Currently this is not possible but it could be added as an option.

    A practical example of the above mentioned is the case where FFmpeg hangs on e.g. a corrupted source. If you would try to restart the process a few times to find out what's wrong, there would be several FFmpeg's running (hanging) in memory without being noticed. With a unique named FFmpeg copy, Avanti can freely terminate any number of FFmpeg's from running if necessary.

    The reason that Avanti sees a UPX'd executable as a special build is because when you startup the program, it analyzes the data in the FFmpeg executable to make adaptions for different versions, e.g. the libfaac vs. aac adaption. An UPX'd executable is compressed so the program can't find the data it's looking for. There's no solution for that.

    BTW; USB 2.0 shouldn't be that slow in copying but it takes indeed twice the FFmpeg space for the time Avanti is running.
    Quote Quote  
  6. Member
    Join Date
    Mar 2010
    Location
    Italy
    Search PM
    I try to convert an mp4 file (AVC baseline@L1.3) but with -bf 2 and -sameq I'm not satisfied with what I've done.
    How I can improve the quality?
    Last edited by jkhox; 22nd Apr 2010 at 02:57.
    Quote Quote  
  7. That's a question about x264 encoding (I suppose). Sorry, I don't know the answer!

    I'm just a developer of a FFmpeg GUI (one of many). Perhaps try a dedicated x264 gui?

    Regards,
    Chris
    Quote Quote  
  8. Member KevinSartori's Avatar
    Join Date
    Feb 2009
    Location
    United States
    Search Comp PM
    Hey! I ran into a weird surround sound problem that I thought I should share. I have a PAL DVD that I wanted to convert to NTSC. There's a lot of music in the movie, so I also wanted to correct the pitch of the audio. I used DVD Shrink to create a single VOB file. I loaded it into Avanti and set the AV Frame Rate Conversion to P/N 23.976. I also added some SRT subtitles in the Subtitles Styler section to add hard subs to the image. It did the conversion fine, but when I went to play back the MPG file, my receiver told me the sound was 5.1, but I couldn't hear anything out of the rear surround speakers.

    In Media Player Classic HomeCinema, I disabled the internal AC3 decoder so that it would use AC3Filter instead. I played the original PAL DVD and I could see in AC3Filter that there were audio levels for all six channels. When I played the converted NTSC MPG file, there were only audio levels for front left, right and center.

    As you recommended, I stopped using codec packs and now only use the latest ffdshow tryouts, Haali Media Splitter and AC3Filter on Windows 7 x64. I double-checked the AC3Filter settings to make sure it wasn't set to only output two channel or whatever. It's set to AS IS (no change). As an experiment, I extracted the AC3 from the PAL VOB and then tried transcoding (recoding?) without the frame rate conversion and it converted and played back with all six channels. Then I tried transcoding the AC3 with the frame rate conversion and it again converted to what MediaInfo shows is a "six channel" AC3, but with actual audio only in the front left, right and center channels.

    When I transcode the audio without frame rate conversion, it feeds directly to FFmpeg. If I transcode the audio with frame rate conversion, it runs through Avisynth first, so I suspect the problem is with Avisynth. I have the latest Avisynth 2.5.8 installed. As a way around this problem, I dug out BeSweet and converted the AC3 that way and then muxed it into an MPG container with the converted video using Avanti, but BeSweet is pretty slow and it would be nice if I could do the whole process with Avanti. Any ideas why Avisynth is blanking out the surround channels when I try to convert PAL to NTSC? Have you ever noticed this yourself?

    As a side note, following the first conversion I did, when I first played the NTSC MPG, I was a little baffled that the frame rate was correct but the sound was still pitched. When I hit the Insert AUTO Script button in the Avisynth Script Editor, I noticed TimeStretch( pitch = 104.270938), which seems to correct the lower pitch back to the pitch of the original PAL audio. Removing the line fixed the problem, as I wanted the lower (original film) pitch of the slower 23.976 frame rate. It might be more intuitive if there was a Correct Pitch Adjustment checkbox or some such in the AV Frame Rate Conversion section.
    Last edited by KevinSartori; 4th May 2010 at 12:58. Reason: typo
    Quote Quote  
  9. Originally Posted by KevinSartori View Post
    When I transcode the audio without frame rate conversion, it feeds directly to FFmpeg. If I transcode the audio with frame rate conversion, it runs through Avisynth first, so I suspect the problem is with Avisynth.
    It souds like it's downmixed somewhere in the chain but it looks like your FFdshow settings are right. I'm currently not able to check it out in detail (moving to another house), but can you try to convert the audio separately (video disabled). That uses other Avisynth commands and I'm curious if that differs.

    As a side note, following the first conversion I did, when I first played the NTSC MPG, I was a little baffled that the frame rate was correct but the sound was still pitched. When I hit the Insert AUTO Script button in the Avisynth Script Editor, I noticed TimeStretch( pitch = 104.270938), which seems to correct the lower pitch back to the pitch of the original PAL audio. Removing the line fixed the problem, as I wanted the lower (original film) pitch of the slower 23.976 frame rate. It might be more intuitive if there was a Correct Pitch Adjustment checkbox or some such in the AV Frame Rate Conversion section.
    Yeah, some dvd's are pitch corrected, others not. I've noticed this problem a while ago and considered a checkbox for it as you metioned. It's not a big change but I'm awefully short of room at the gui. I'll still try to find some because it's an important addition.
    Quote Quote  
  10. Member KevinSartori's Avatar
    Join Date
    Feb 2009
    Location
    United States
    Search Comp PM
    When I try to do just an AC3 file, Avanti tries to start up Avisynth and I get an error:

    Input #0, incorrect Avisynth duration ... must be 01:41:04.52 [ 101.12 min ]

    MediaInfo shows the duration of the AC3 to be 1h 36mn, so the 1h 41mn duration would be the correct length after conversion from PAL 25 to NTSC 23.976, but Avisynth is having some kind of problem.
    Quote Quote  
  11. It's not a Avisynth error but a error of the FFmpeg Avisynth interface. It has no effect on the conversion.

    Code:
    Input #0, incorrect Avisynth duration ... must be 00:09:44.40 [9.74 min.]
    
    Input #0, avs, from 'G:/avanti_path/excerpt_final.avs':
      Duration: 00:00:00.00, start: 0.000000, bitrate: -2147483 kb/s
        Stream #0.0: Audio: pcm_s16le, 48000 Hz, 2 channels, s16, 1536 kb/s
    As you can see at the FFmpeg input data, FFmpeg shows "Duration: 00:00:00.00". In that case Avanti doesn't use the FFmpeg info but get it elsewhere and reports it actually more as a note.

    MediaInfo shows the duration of the AC3 to be 1h 36mn, so the 1h 41mn duration would be the correct length after conversion from PAL 25 to NTSC 23.976
    It's normal that Avisynth reports the converted duration.
    Quote Quote  
  12. Member KevinSartori's Avatar
    Join Date
    Feb 2009
    Location
    United States
    Search Comp PM
    I got the error, but then the conversion didn't happen. Should I test anything else?
    Quote Quote  
  13. Originally Posted by KevinSartori View Post
    I got the error, but then the conversion didn't happen. Should I test anything else?
    Do you mean; ... you loaded a ac3 at source 1, then disabled video, set AV frame rate conversion to P/N 23.976 and didn't you get a converted ac3 at all? Or do you mean you still get a three channel result? Please post some more details.
    Quote Quote  
  14. Member KevinSartori's Avatar
    Join Date
    Feb 2009
    Location
    United States
    Search Comp PM
    Hey! Sorry I wasn't more clear! No, it didn't convert at all. It converts fine with all five channels if I don't try to convert PAL to NTSC, but if I set AV Frame Rate Conversion to P/N 23.976, then I get that incorrect Avisynth duration error and it quits.
    Quote Quote  
  15. Don't understand why it quits? I don't work much with six channel audio but have a few ac3 test clips that I use for testing. These work fine.

    Code:
    16:25:56 - Process started using FFmpeg version SVN-r22988 [ffmpeg22988].
    ======================================================================================
    
    Input #0, incorrect Avisynth duration ... must be 00:00:35.43 [0.59 min.]
    
    Input #0, avs, from 'G:/avanti_path/Broadway-5.1-48khz-448kbit.avs':
      Duration: 00:00:00.00, start: 0.000000, bitrate: -2147483 kb/s
        Stream #0.0: Audio: pcm_s16le, 48000 Hz, 6 channels, s16, 4608 kb/s
    ---------------------------------------------------------------------------------
    Output #0, ac3, to 'F:/destination_path/test_final.ac3':
        Stream #0.0: Audio: ac3, 48000 Hz, 5.1, s16, 448 kb/s
    ---------------------------------------------------------------------------------
    Stream mapping:
      Stream #0.0 -> #0.0
    
    ======================================================================================
                                             +--------------------+
                      +----------------+     |                    |
       Source [1] --> | Avisynth AUTO  | --> |       FFmpeg       | --> Destination [1]
                      +----------------+     |                    |
                                             +--------------------+
    ======================================================================================
    
    16:26:00 - Free space on destination disk at start 29,161 Mb.
    
    FFmpeg messages/warnings:
    
    [avs @ 018be0c0]Estimating duration from bitrate, this may be inaccurate
    
    [ac3 @ 019a0210]No channel layout specified.
      ... The encoder will guess the layout, but it might be incorrect.
    
    Metadata:
      ... encoder         : Lavf52.61.0
    
    16:26:10 - Process finished after 00:00:10 hms with messages/warnings.
    
    ... Console processing <Ctrl><S> possibly shows more details.
    FFdshow (volume) shows all six channels.

    I have the audio settings set to... ac3, bitrate 448 kbps, sample freq 48000 and channels to 6.
    Quote Quote  
  16. Member KevinSartori's Avatar
    Join Date
    Feb 2009
    Location
    United States
    Search Comp PM
    Are you also trying to convert from PAL to NTSC? That's where I have a problem. I wonder if my problem is with the FFmpeg build I'm using. I'm still on 0.98. I'm at work now, so I can't check the exact build, but I can later.
    Quote Quote  
  17. Originally Posted by KevinSartori View Post
    Are you also trying to convert from PAL to NTSC? That's where I have a problem.
    Yes, "P/N 23.976". You can see at the FFmpeg input that it runs a Avisynth script, not the ac3 directly.
    Code:
    Input #0, incorrect Avisynth duration ... must be 00:00:35.43 [0.59 min.]
    
    Input #0, avs, from 'G:/avanti_path/Broadway-5.1-48khz-448kbit.avs':
      Duration: 00:00:00.00, start: 0.000000, bitrate: -2147483 kb/s
        Stream #0.0: Audio: pcm_s16le, 48000 Hz, 6 channels, s16, 4608 kb/s
    The source properties show a duration of 00:00:33.98 and the resulting clip 00:00:35.43 as shown above.
    Quote Quote  
  18. Notes @ Avanti 0.4.7 release ...

    http://avanti.arrozcru.org/upgrade.htm

    This version has to do for a while because I'm preparing a move to another house (for medical reasons). I have to dismantle my computer setups and the telephone company needs "four weeks" (of paperwork) to connect me at the new house. I hope the new things in Avanti 0.4.7 won't give to much trouble because I can only reply to posts for still a couple of days.

    And then this ...

    http://ffmpeg.arrozcru.org/autobuilds/blog/2010/05/07/libavfilter-is-in/

    From FFmpeg version SVN-r23056 up the well known crop and pad commands don't work anymore (afaik). A working version of "libavfilter" is now included and you need to replace e.g. -padtop and -padbottom with something like; -vfilters pad=640:480:0:0 (width, height, x, y).

    This is very interesting but with this FFmpeg change the Avanti crop/scale/pad option will fail, when you select FFmpeg to do the job (Avisynth still works fine). It needs a lot of research to keep Avanti compatible with both old and new FFmpeg versions and as you may understand, it will take some time (even before I can start with it).

    Chris K
    Quote Quote  
  19. Member
    Join Date
    Jul 2008
    Location
    United States
    Search Comp PM
    I think the extension for libtheora is wrong in the codec database manager. Currently its ogm but according to http://wiki.xiph.org/index.php/MIME_Types_and_File_Extensions ogm is not listed. I think it should be ogv. ogm files still play, I just thought you should know. Thanks!
    Quote Quote  
  20. It's a bit of chaos with media file extensions. It seems that .ogm is unofficially adopted at some places. Perhaps with the thought that .ogv would stand for a raw Theora stream? MPCHC has it on its formats list as "Ogg Media file; .ogm, .ogv".

    "xiph.org" is the official source so you're right, it should be .ogv exclusively.

    But with my FFdshow / MPCHC setup, playback fails when I use the .ogv extension while the .ogm extension does fine? MediaInfo however reports the .ogv file as valid theora/vorbis in ogg container.

    Chris
    Quote Quote  
  21. Member ricardouk's Avatar
    Join Date
    Mar 2005
    Location
    Portugal
    Search Comp PM
    Hi Chris!

    Hope everything is going well with the "moving/changing house", from what you said i realize you don't have much time nowadays but is there any chance of including webm/vp8 support on Avanti?

    http://www.webmproject.org/

    I've been searching the net for ffmpeg builds with vp8/webm support but the only one i found so far doesnt have libvorbis, the default vorbis encoder in ffmpeg from what i googled is not very good and we cant pick audio bitrate values only quality settings unlike libvorbis that has superior quality and we can pick a audio bitrate.

    Do you know where i can get a ffmpeg build with vp8/webm and libvorbis? Are you planning on adding support for it on Avanti?

    http://www.webmproject.org/tools/


    http://code.google.com/p/webm/downloads/list


    Thanks
    I love it when a plan comes together!
    Ricardo Santos
    Quote Quote  
  22. Originally Posted by ricardouk View Post
    Hope everything is going well with the "moving/changing house", from what you said i realize you don't have much time nowadays
    I'm still at the stage of packing stuff.

    but is there any chance of including webm/vp8 support on Avanti?
    Where it concerns FFmpeg support for encoding/decoding vp8, I think it's only a matter of time. Avanti don't need no special adaptions for this except including a few new templates.

    http://multimedia.cx/eggs/vp8-and-ffmpeg/

    You could ask Ramiro Polla if he's willing to include libvpx but it might be a bit too early.

    http://ffmpeg.arrozcru.org/forum/

    I'm not involved in FFmpeg development and my main concern for Avanti is to keep it compatible with the most recent FFmpeg revisions. Currently this is the addition of a revised libavfilter that takes over the old crop and pad commands. This is still incomplete because its scale filter doesn't support the scale algorithms yet. But I have to use it anyway because the old pad commands don't work anymore.

    Chris
    Quote Quote  
  23. Member ricardouk's Avatar
    Join Date
    Mar 2005
    Location
    Portugal
    Search Comp PM
    Im a bit confused now lol....

    From what i read there are 2 ways to encode to webm with ffmpeg, i wasnt aware of the seconde one:

    1- ffmpeg build compiled with vp8
    http://micksam7.com/blog/index.php/?p=743 (windows builds here)

    2 - normal ffmpeg version through an external libvpx file
    http://code.google.com/p/webm/downloads/detail?name=vpx-vp8-debug-src-x86-win32mt-vs8-...0.zip&can=2&q=

    Originally Posted by Chris K View Post
    I'm not involved in FFmpeg development
    i knew that, i was hoping that you knew someone (trip45, paolo etc) that might have a working ffmpeg build with libvorbis and vp8 but didnt share it.

    can you tell me how can i use ffmpeg in the seconde example, or point me in the right direction? I see a lot of chat on the web but i dont see any info for "regular/casual" users like me.

    Thanks for any info you can give
    I love it when a plan comes together!
    Ricardo Santos
    Quote Quote  
  24. The second link gives the source code and a win32 build of "libvpx" (the dll's). It has to be added to a folder with a shared FFmpeg build that is made aware of the library during compilation. Afaik, it can't be used with just an arbitrary FFmpeg build.

    I've played around a bit with the FFmpeg executable on the first link but I currently can only run Avanti as user on a borrowed laptop (to have web access).

    Avanti expects a FFmpeg version that accept to set a video codec. I would expect "-vcodec vp8" to work with that executable but it doesn't. It is a special build that is only capable to encode webm vp8. I can't change Avanti to support all kind of experimental builds made by individuals.

    Just be patient and wait for the FFmpeg gurus.

    EDIT:

    Found it! The codec name is "libvpx_vp8". Here's a template for Avanti to try ... UNI_WEBM_VP8.7z

    Only for FFmpeg builds with "--enable-libvpx-vp8" in the configuration. Encoding is extremely slow and the DirectShow webm decoder I downloaded crashed on the file. MediaInfo sees vp8 and vorbis in it but says its a mkv container?
    Last edited by Chris K; 26th May 2010 at 16:42. Reason: addition
    Quote Quote  
  25. Member ricardouk's Avatar
    Join Date
    Mar 2005
    Location
    Portugal
    Search Comp PM
    i used the template with the ffmpeg build from the previous link with vp8 and it displays the following message:

    FFmpeg messages/warnings:
    Unknown encoder 'libvpx_vp8'

    at the moment i can convert a video with the ffmpeg build with vp8 using a script/bat but the vorbis quality isnt hat good, i need to use libvorbis but that build hasnt libvorbis on it, so i converted the audio to ogg with LameXp so at the moment i have:

    1 webm video with vp8 and vorbis (bad audio quality) and one ogg/vorbis audio track, can ffmpeg discard the vorbis audio in first file and mux the audio i converted with lamexp? Kind of remuxing the video with different audio track?

    Thanks
    Last edited by ricardouk; 27th May 2010 at 04:54.
    I love it when a plan comes together!
    Ricardo Santos
    Quote Quote  
  26. Originally Posted by ricardouk View Post
    i used the template with the ffmpeg build from the previous link with vp8 and it displays the following message:

    FFmpeg messages/warnings:
    Unknown encoder 'libvpx_vp8'
    I used the older build "r2". The latest "r3" crashed on my system. It reads "built on May 20 2010 13:11:25 with gcc 4.2.1-sjlj (mingw32-2)"

    This is the output on the Avanti log window;
    Code:
    Output #0, webm, to 'F:/destination_path/test_final2.webm':
        Stream #0.0: Video: libvpx_vp8, yuv420p, 720x576 [PAR 16:15 DAR 4:3], q=2-31,
          450 kb/s, 1k tbn, 25 tbc
        Stream #0.1: Audio: vorbis, 22050 Hz, 2 channels, s16, 64 kb/s
    at the moment i can convert a video with the ffmpeg build with vp8 using a script/bat but the vorbis quality isnt hat good, i need to use libvorbis but that build hasnt libvorbis on it, so i converted the audio to ogg with LameXp so at the moment i have:
    1 webm video with vp8 and vorbis (bad audio quality) and one ogg/vorbis audio track, can ffmpeg discard the vorbis audio in first file and mux the audio i converted with lamexp? Kind of remuxing the video with different audio track?
    You could try it with Avanti. Video (with unwanted audio) on source 1 and separate audio on source 2. Then set codec fields to "Copy Audio" and "Copy Video" and see what happens. You need to do it with the special FFmpeg build because a regular build cannot decode webm vp8.
    Quote Quote  
  27. Member ricardouk's Avatar
    Join Date
    Mar 2005
    Location
    Portugal
    Search Comp PM
    It works, i have to overwritte the container box with webm:

    http://img227.imageshack.us/img227/3215/avantig.jpg

    one side effect... audio only plays for 4 seconds and then disappears, guess i'll have to wait for more complete "official" ffmpeg builds.

    Thanks for your help Chris, will wait for Avanti to support it...meanwhile back to scripts.
    I love it when a plan comes together!
    Ricardo Santos
    Quote Quote  
  28. Hi Ricardo,
    I was successfull doing encoding the audio and video seperately:
    You will need the following:
    ffmpeg with libvpx support
    oggenc from www.rarewares.org
    mkclean: http://www.matroska.org/downloads/mkclean.html
    mkvtoolnix: http://www.bunkus.org/videotools/mkvtoolnix/
    I
    did the following with a test video called Wildlife.wmv:
    ffmpeg -i "c:\Users\Public\Videos\Sample Videos\Wildlife.wmv" -acodec pcm_s16le -vn test.wav
    oggenc2.exe -b 128 test.wav
    ffmpeg -i "c:\users\Public\Videos\Sample Videos\Wildlife.wmv" -vcodec libvpx -an testvid.mkv
    Launch mkvmerge GUI v3.4.0(included in MKVtoolnix)
    Add both test.ogg and testvid.mkv to the Input files list
    Specify output file name for example: finalresult.mkv
    Press 'Strart muxing'
    mkclean.exe --doctype 4 finalresult.mkv finalresult.webm
    I only have tested on VideoLAN classic player version 1.1.0rc
    Quote Quote  
  29. Member ricardouk's Avatar
    Join Date
    Mar 2005
    Location
    Portugal
    Search Comp PM
    Hi Leon!

    Are you using the ffmpeg build from Micksam7 blog?

    My method is more or less like yours

    1- demux audio from source
    2- encode video to webm with ffmpeg
    3-load webm file onto mkvtoolnix build with webm support, deselecting the audio audio track (minimize mkvtoolnix)
    4-convert audio from step1 with lamexp to ogg
    5-load the ogg track to mkvtoolnix
    6- mux the video as webm

    heres the mkvtoolnix version with webm support:
    http://www.bunkus.org/videotools/mkvtoolnix/webm/mkvtoolnix-unicode-3.4.0-build2010051...-244-setup.exe

    will try your method

    where did you get your build from?
    In your example i dont see any reference to the ogg output in the command line (typo?)

    Thanks for the info

    PS: Welcome to videohelp, guess you seen my comments elsewhere about this
    I love it when a plan comes together!
    Ricardo Santos
    Quote Quote  
  30. Appologies if my reply is all in one line. I am typing from my phone. Thanks for the welcome. Yes, I am using the ffmpeg build from the Micksam7 blog. The encoding speed speems really slow. The ffmpeg from Miro Video converter seemed a bit faster, but I might be wrong because I used a different PC and different content. I don't think I made a typo, oggenc automatically creates an output file with .ogg extension based on the input file name. Also Im not 100% sure if I created the WAV file correctly because of all the 16 bit pcm choices ffmpeg had. I chose one randomly. Does anyone know if there are differences between them? it seems to sound/work correctly
    Quote Quote  



Similar Threads