Peculiar partial absent of sound

AviSynth script is as siple as this:

Code:

Source1 = AVISource("Source1.avi")
Source2 = AVISource("Source2.avi")
Source3 = AVISource("Source3.avi")

Part1   = Source1.Trim(1998, 4449)
Part2   = Source2.Trim( 334,    0)
Part3   = Source3.Trim(   0,  354)

Dissolve(Part1, Part2, Part3, 6)

Loop(2)     # Play it 2 times in a row

(So I joined small parts of 3 videos - with a small overlapping - and then doubled it.)

Playing back directly this .avs file don't cause any problem.
But after conversion to real video file (I tried different formats, e.g. MP4 AVC/AAC,
MP4 AVC/MP3, AVI XviD/AC3) the situation is other:

- Video is playing back OK
- Audio is playing back ALMOST OK - in the 2nd loop the Part1 is playing without sound.
 So the sequence of the sound is following: Part1, Part2, Part3, silence, Part2, Part3.

All source audios have the same format (listing from MediaInfo):

Code:

ID                                       : 1
Format                                   : MPEG Audio
Format version                           : Version 2.5
Format profile                           : Layer 3
Mode                                     : Joint stereo
Mode extension                           : Intensity Stereo + MS Stereo
Codec ID                                 : 55
Codec ID/Hint                            : MP3
Duration                                 : 3mn 18s   <---- different for every one
Bit rate mode                            : Constant
Bit rate                                 : 24.0 Kbps
Channel(s)                               : 2 channels
Sampling rate                            : 11.025 KHz
Compression mode                         : Lossy
Delay relative to video                  : 43ms  <--------------- here are some differences
Stream size                              : 578 KiB (1%)
Alignment                                : Split accross interleaves
Interleave, duration                     : 40 ms (1.00 video frame)
Interleave, preload duration             : 500 ms

If someone had similar problem or has any idea about understanding it or at least resolving it,
I gently ask for help.

(Windows 7 Ulitmate 64-bit, K-Lite Mega Codec Pack 12.1.6)

Which version of AviSynth? I think old version had some issues with loop function and audio.

You could also try:
newvid = Dissolve(Part1, Part2, Part3, 6)
newvid++newvid

Thank you for your instant answer.

My version is AviSynth+ 0.1 (r1576, x86) from 2014. jan. 02.

I tried your advice (++) but with the same (bad) result.

When I omitted the last command (Loop(2)), convert it and playback it with “Repeat forever”, the effect is the same – in second (and 3rd, 4th, …) playing Part1 is mute. I tried it with different media players.

Interesting enough:

When I joined only 1st and 3rd parts, i. e. when I omitted Part2 from the Dissolve command:

Code:

   .
   .
   .
Dissolve(Part1, Part3, 6)   # instead of Dissolve(Part1, Part2, Part3, 6)

Loop(2)                     # Play it 2 times in a row


the issue is still there – after converting to .mp4 (AVC/AAC) the sequence of sound in playback is
Part1, Part3, silence, Part3.

When I omitted Part3, the same problem:
Part1, Part2, silence, Part2.

But when I omitted Part1, all is OK:
Part2, Part3, Part2, Part3
- all with the sound!

And when I omitted both Part2 and Part3 (so only Part1 loops twice), the same problem:
Part1, silence!

So it seems as the sequence Part3, Part1 (from 1st to 2nd loop) is problematic - and so the sequences Part2, Part1 and Part1, Part1.

Part1 itself is played back always with the sound on the beginning of the whole video.

May something bad in the beginning of Part1 prevent playing the sound of it - but only after other (or the same) part?

Maybe something is wrong with the audio timestamps. That can cause problems. I don't know enough about AviSynth to know what exactly it does with those. Also, which program are you using for encoding?

LAV Audio Decoder has an option called "Auto A/V Sync Correction". Try disabling that. It is designed for handling small gaps, but can give unwanted results for big gaps in timestamps.

The Dissolve function also has an optional FPS parameter, which can be set to the audiorate to give better accuracy. Maybe that helps too.

Thank you for your tips, I tried both of them without success.
I use Any Video Converter Free 3.5.8 (versions 5.x process incorrectly dissolving of audios as lot of others programs).

Meanwhile I tried other things and I think I am now able to pinpoint the location of problem:
Something bad happens between frames 1218 and 1219 of the Source1.avi, a change from GOOD to BAD:

Code:

Part1   = Source1.Trim(a, 4449) # Works fine for a <= 1218, otherwise causes the "silence" problems


Unfortunately I neither know WHAT happens nor I have any idea how to go around it.

If you have some ideas, I will welcome them.

Very strange. I assume it does not happen if you use a different file as source1?

You can find AviSynth experts (and even developers) on this forum:
http://forum.doom9.org

You are right – it does not happen with any other file as Source1.

Thank you for your effort and for the link to doom9 forum. If I will know the reason (and maybe a solution, too) I will share it here.