Audio encoding to a certain frame, separate from video?

[Letricheur]

Because of the long time it takes to encode on my machine, I am splitting the encoding of a DVD in CCE by using the trim function in avisynth. Because of the reported CCE audio encoding problems I am only encoding the video, leaving the audio to another program. This means I have to multiplex the correct length of audio with the encoded video segment.

Is there a program that will allow me to encode the correct length of audio, separate from the video? Another thought I had was to split the DVD by making several DVD2AVI files, which will give me the correct length of audio for each segment. The problem is that DVD2AVI does not seem to allow encoding by frame number, only by visual guess-work.

[MaTTeR]

I can't try it right now with BeSweet because of some strange VB problems I'm having.

However, I've used HeadAC3he in the past to do exactly what you describe(encode by frame segs). Hope I'm understanding your question right:)

[Letricheur]

I think you understood the problem OK. I want to be able to encode just the audio that accompanies, say, video frames 0-63297. Thanks for the tip, I'll try HeadAC3he.

[Acaila]

Be carefull what frame values you input in HeadAC3he then. Because video-frames are not equal in length to audio-frames.
Instead find out the duration in mm:ss:ms of the video and enter that value for the audio to encode in HeadAC3he and it should work just fine.

[Letricheur]

Have discovered 3 solutions for my problem of piecemeal video encoding in CCE and muxing the separately encoded audio.

1. The simplest is to load the relevant *.d2v-based avisynth file into TMPGEnc and set it up to encode just the audio stream to *.m2p. The trim function in avisynth that I used to split the video encoding chore works to ensure that just the right amount of audio is encoded. The *.m2p from this and the *.m2v from CCE can then be easily remuxed in bbMPEG. There seem to be no sync problems.

2. I use VBR encoding in CCE but for the first pass (needed to creat the *.vaf file) I use CBR. With this pass, I encode the audio, creating a *.mpa file. When doing the next VBR multipass, I do not encode the audio. I then mux the audio from the first CBR pass and the video from the subsequent VBR passes.

3. (SLOW) I load the *.avs file into VirtualDub, set the video to direct stream copy and the audio to full processing. I run the preview then save the *.wav file. This must then be converted to *.mp2 (e.g. in TMPEGEnc.

Method 1 is fastest and works well. Again, I only split the encoding into several pieces because my (laptop) computer is slow (CCE RT 0.31 or so) and I can't afford to let it run continuously all the time.