What format *will* work with the VSE?

My camera records in MPG format with AC3 surround sound audio, and Blender can’t use that audio (neither can almost anything else, even when I have the codec installed), though the video works fine.

After more than a year of on-and-off effort to solve this problem, I finally found that I could import the video into Windows Movie Maker and render it as plain old .avi. Now the audio works fine but Blender says “either it’s not a video or FFMPEG is not compiled in.” The video plays fine in anything else; why not in Blender? There’s some kind of curse keeping me from ever using the hours of footage and days of painstaking work that went into recording a movie with my friend last year!

Here is a sample clip straight from my camera: http://uploader.polorix.net//files/874/M2U01015.MPG
You will need AC3 Filter to read the audio.
Although as I type this, I stumbled upon a perfect solution that would require the video and audio in separate files, if you can convert that file so that both audio and video work with only file in Blender’s Sequence Editor (and tell me how), I would be eternally grateful! In my next film which uses your solution, I’ll even put you under Special Thanks/Technical Support in the credits.

Blender’s FFMpeg “hooks” seem to change on every new release.

Files that would load in ver 2.44, no longer load in 2.45 or SVN.

I’ll take a look at your file and see if I can load it.

Are you on 'WinDoze or Linux ? … what version of Blender are you using ?


AFAIK the VSE will split the video & audio into separate tracks when successfully imported, so as long as you have good synch points in your footage, why not import them as separate external tracks? Plus there is an as-yet-unresolved bug that can introduce some bad audio glitches when importing combined audio/video from some formats.

If you have a means of producing separate audio and video tracks, maybe try bringing them in separately and re-synching? I’ve done this using Virtual Dub to produce a “silent” avi and a .wav file, and it seems to work OK, though I haven’t tried it with dialog footage (possible lip-synch issues).

I’m using this SVN build (Windows; WinDOWS!:evilgrin:): http://www.graphicall.org/builds/builds/showbuild.php?action=show&id=560

The same thing happens in regular 2.45.

@chipmasque: I could do that, but there are a few hundred video files for our movie, and even though it wouldn’t be that hard to sync them if I use only Blender for editing, it would be nice to not have to. Especially considering it’ll be hard enough to create a decent fanfilm out of this rabble of footage and sub-par acting. :ba:

when I click the speaker, I’m getting the error “Resampling with input channels greater than 2 unsupported.” followed by an unceremonious exit.

2.45.14 wants nothing to do with it.

I used the Total Video Converter program to convert your file to Mpeg4Avi, then was able to load it into Blender 2.44, and it loads and plays.

Couldn’t load your original file or the MSMpeg4 file into 2.45 or SVN.

… And btw the Mpeg4 file is 1.3meg compared to your 10meg file ! … Quality appears to be the same.

I’ll experiment further.

Btw I’m using Win XP with .NET ver v2.0.50727.

I think it was around or after ver 2.44 that the Windows build compiler was changed to use the .NET libraries ??? … and this might be a contributing factor to the FFMpeg problem.


@papasmurf: That’s because it’s surround sound (more than 2 channels, which would be stereo). You’ll have to convert it to a format Blender can use.

@Mike_S: Sorry, you posted while I was writing. I guess it’s a problem with the current FFMPEG, then. Thanks for the help.