Colourspace and lossless codec for capturing before AviSynth

[FlimsyFeet]

I'm using VirtualDub with a BT878 based PCI card.

I can think of two options for lossless AVI capture:
1. YUY2 compressed with huffyuv
2. YV12 compressed with Lagarith

I've always used 1 in the past, becasue someone once told me that YUY2 is what the chipset uses internally. Also YUY2 has more chroma detail (4:2:2 instead of 4:2:0 used by YV12).

But most avisynth filters these days work on YV12 video, which means I have to ConvertToYV12 in my script. Fine for Xvid encodes, but CCE needs YUY2. Can I use ConvertBackToYUY2, or is that only for RGB conversions?

Would it be preferable to use option 2 and not have to convert at the beginning of the script? The capture files would be smaller. Was there an issue with using YV12 if the source is interlaced?

[IanB]

... YUY2 has more chroma detail (4:2:2 instead of 4:2:0 used by YV12).The "more chroma detail" is not the issue here. The crucial concept is each line has it's own chroma information, and this really matters when it comes to processing interlaced video. It means you can simply avoid all issues of interlaced YV12 until you are ready to deal with them.

:search: for Interlaced YV12 in the Avisynth Users forum and you will find a plethora of issues. Issues that are best avoided.

:edit: Opps wrong forum

[FlimsyFeet]

So - generally avoid YV12 for captures, unless your source is progressive (i.e. PAL movie) and you want small capture files?

[IanB]

Even then, you still must be carefull and know exactly what is going on. E.g. Your source may well be progressive but the capture driver still might use interlaced chroma placement...

[Poutnik]

I can think of two options for lossless AVI capture:
1. YUY2 compressed with huffyuv
2. YV12 compressed with Lagarith
..........


Why not to use option
3. YUY2 compressed with Lagarith ?

It supports it.