How to frameserve (dvd2avi / VirtualDub / Avisynth / VFAPI)

[manono]

or do I have to use ConverToYV12() at the end of my .avs scripts with video sources that are not already in the YV12 colourspace?

I already told you:

ConvertToYUY2(Interlaced=True)

if interlaced and for CCE. If progressive leave off the Interlaced=True. All MPEG-2 encoders output YV12, but CCE accepts YUY2. HCEnc accepts YV12. TMPGEnc accepts RGB24. I think (but am not positive) that the rest accept YUY2. If you don't feed them what they want, they'll convert it themselves, and I'd prefer to have AviSynth do the conversion. Except for HCEnc which will refuse to accept anything not already YV12.

Opening the .avs in VDub for further filtering not only leads to slower encoding, but also to inferior results. It's going to convert anything you feed it into the RGB colorspace. The fewer colorspace conversions the better.

[meldavid]

I was actually asking about ConvertToYV12() in regards to encoding any source file to DivX or Xvid using VirtualDub. I'm not sure if I should use AVISynth to do colour conversion or just let VirtualDub and the DivX/Xvid encoder handle it. I suddenly changed topic so you thought I was still asking about my DV Cam capture and frameserving to CCE.

...

And the question I was asking about CCE was regarding the Luminance Levels setting (0-255 or 16-235) when frameserving from VirtualDub. But I think I tried an encode last night and CCE didn't accept the *.vdr file anyway.
So I'll have to frameserve to CCE using AVISynth only.

[manono]

I also need to know if the Xvid/DivX codec or VirtualDub will take care of the colour conversion process automatically or do I have to use ConverToYV12() at the end of my .avs scripts with video sources that are not already in the YV12 colourspace?

I see now. Since that was the first mention of encoding for AVI, I ignored it. I thought you were still asking about frameserving to CCE. Same principle. If you use the VDub filters, it's going to convert to RGB32 (and it might be a good idea for you to do the conversion in the script). If you don't, and your source is a DVD, you need to add nothing as no colorspace conversion will take place. It stays YV12 all the way if you use Fast Recompress (not Full Processing). That's one of the advantages - one of the big ones - of doing all the filtering within AviSynth - no colorspace conversion at all. Here's the AviSynth YV12 FAQ:

http://avisynth.org/mediawiki/YV12_FAQ

In CCE I keep the luminance levels at 0-255. I encode from DVDs, and that's what they all use. I don't know anything about DV camcorders. Maybe someone else does.