Huffyuv capture questions

[sanlyn]

Quote:

Originally Posted by jt_retro

I guess a 240p picture is just 480i, except that the 2 fields inside each frame are from the same instant in time (and also the vsync circuit is triggured at different times compared with a real 480i signal). Does that sound about right?

Nope. 480i consists of two images, each stored in the frame as 240 lines in height. The 240-line image you get from each field is expanded into full-height 480p frames by deinterlacing.

If you want 240-line fields from 480i frames, use SeparateFields() in Avisynth. Or use BobDoubler/AlternateFields in Virtualdub. A 720x480i video will be split into 720x240p half-height fields.

I don't know that frames smaller than 240 pixel height accept interlacing in any standard delivery format. 320x240i interlaced would be one really oddball video format with 4:3 320x120 fields.

[sanlyn]

double post.

[msgohan]

Quote:

Originally Posted by jt_retro

Under what scenario would a codec decompressor give a series of fields, instead of woven frames? I'd would have imagined that the de-interlacing stage would be after decompression, thus would require woven frames for deinterlacing to work?

Deinterlacing doesn't require woven frames. In this case, the deinterlacer just needs to know that it's received separated fields and then it adds a line in between each existing line by some method of interpolation.

Quote:

I guess a 240p picture is just 480i, except that the 2 fields inside each frame are from the same instant in time

No. What you're describing is 480p30, not 240p60.

480p30 sent as 480i has 2 fields from the same time, alternating spatially 1/60 of a second.
240p60 has 2 frames from different times, occupying the same spatial position, refreshed with a new complete image 1/60 s. (This way the image doesn't twitter up-and-down the way it would if the frames were drawn as fields.)

[jt_retro]

Sorry folks for the confusion regarding my 480i vs 240p statement - I mustn't have had my coffee that day