how do dv-cameras interlace?

[((( atom )))]

short question: how do dv-cameras interlace? do they capture a full frame at once and interlace it or do they capture 2 fields one after another? this seems quite interessting to me in terms of how would i deinterlace the video afterwards.

(just had a discussion with someone that told me, i actually couldn't deinterlace dv material. well i did, it looks good, so now i want to know the techniques behind it..)

[bb]

They capture two fields one after another, with a time difference of 1/50s (PAL) or 1/60s (NTSC). That's what makes it so hard to get progressive frames out of it.

bb

[avih]

@atom, it what bb said was not the case, then we could hardly call it interlaced... hmmm... progressive, yes, it would be progressive scan ;)

[sync]

For a detailed description see this:
http://www.100fps.com

[((( atom )))]

read threough 100fps.com again (did it ages ago) and decided not to deinterlace the material, since i want to output to dvd for tv-viewing anyway.

..but i don't like it!

[ulfschack]

What I would be interested in knowing is that if the same CCD pixels are used for both odd and even fields.

I guess that it would be a waste of space/pixels not to reuse, but again maybe using certain pixels for even/odd respectively allows for faster shutter speeds taking "ramp down period" or something to that effect into account ... I don't know enough to make a half decent guess, but the reason I ask is the big difference of the number of pixels in various DV cams.

I've seen ranges from 320 to 1300+ thousands of pixels. I suspect the higher resolutions are for stills, but for lets say 540 kpx , who would want stills of that resolution? What exactly are they used for? Maybe better interpolation of colors? My thinking tells me that 288x720 (< 320 000) should be more than enough for interlaced DV.

Any thoughts?

cheers