Bug in avisynth

[wertherman]

If the video clips when doing an aligned splice are too long in avisynth 2.08 (probably applies to avisynth 2.5) then the audio disappears from the video (0 samples of audio is outputted in the video stream)

Length at which it breaks is not absolute and dependent on other factors as I originally encountered the problem while align splicing about 230 video clips together and it broke around the 13 minute, 31 second mark (rather than the 14 minute mark like the example code below).



# Code that has audio

BlankClip(length=1212322,width=720,height=200,fps=25, color=255)
audio = WAVSource("K:\avs\Monty Python - Lumberjack song.wav")

vid1 = audioDub(last, audio)

BlankClip(length=50000, width=720, height=200,fps=25, color=255)
audio = WAVSource("K:\avs\Monty Python - Lumberjack song.wav")

vid2 = audioDub(last, audio)

AlignedSplice(vid1, vid2)




# Code that doesn't

BlankClip(length=1212323,width=720,height=200,fps=25, color=255)
audio = WAVSource("K:\avs\Monty Python - Lumberjack song.wav")

vid1 = audioDub(last, audio)

BlankClip(length=50000, width=720, height=200,fps=25, color=255)
audio = WAVSource("K:\avs\Monty Python - Lumberjack song.wav")

vid2 = audioDub(last, audio)

AlignedSplice(vid1, vid2)


Only difference between two code segments is the length of the blank clip.

[WarpEnterprises]

2.08 only supports audio up to 6 (or was it 12?) hours. You have 1212322/25/3600=13.4 hours.
Try 2.52, it should work there.

[sh0dan]

Unfortunately it is only internally :(

As the VFW only supports up to 13h sound, there is no way it can deliver more than 13 hours per clip. 2.52 is able to handle longer sound internally (as compared to 2.0), but you need to trim the sound and put it together afterwards to get longer sound.

Edit: The 2.5.2 CVS binary (see sig) has a few fixes for long audio if you still should experience problems.

Sorry, but there isn't much we can do about that. :(

[wertherman]

Thanks for the help. Originally I was using 2.08 because there didn't seem to be a vobsub filter for 2.5 but having just checked gabest's site, there seems to be one now, so all is good.....

EDIT: Removed comment about processing audio because it was off topic and also because besweet appears to have a batch processing option using lst files which should work fine.