Do I need a 50fps progressive video to get fluid DVD?

[pgod]

Hello,

I've already read very often that deinterlacing doesn't make sense if I only grab my DV tape content and put it on a DVD to watch it on my TV.

I guess this remains true also if I do some scene cutting or color correction, as this doesn't work more bad with fields than it would with frames.

My question is: when DOES deinterlacing make sense (let aside "I want to watch it on my PC")? Say I feel like Spielberg and want to throw CGI stuff from 3DMax or sort of into my movie. Do I render the CGI stuff into fields too? Or better leave that progressive and deinterlace my movie too? How about Keying, or cutting out moving persons (just an example!) from my movie? Interlacing lines would make that work horrible then, no?

And if in these (or other) cases deinterlacing makes sense, then one more question on how to do it right: After long researching and trying, I think the deinterlacing described on http://www.100fps.com/ is clearly the best, but it gives me a 50 fps movie (I live in the PAL world btw.). Now PAL (on TV) is 25fps, but fields, so when I deinterlace my movie as described into 50 progressive fps, render it back to 25 progressive fps (because I don't want a double-sized file), edit it (see: Spielberg), and put it onto DVD (where it is again interlaced?), then do I end up seeing only half the frames (half the fluid movement) on TV as if I watch my DV tape onto TV?

I know this is described badly (I am not english), so I try to say it another way too: Do I loose fps when not working with the 50fps file (which I would only use when i want to use slomos)?
Because I can imagine that when the DVD encoder renders the video back into fields, it can take frame 1/50 per second for Fields 1,3,5,7,9... and frame 2/50 for fields 2,4,6,8..., while when the source is 25fps progressive, it can only take frame 1/25 for even AND uneven fields (so it would be faked interlacing after all), resulting in only half-fluid movie like the original on my DV tape.

Anyone understands my babbling?

Thanks!

--
Manuel

[DJRumpy]

The only time I ever de-interlace, is when my source is telecined. When a video is telecined to 29.976, it is interlaced as a result. Many broadcast shows are telecined, as well as VHS movies. It makes sense to reverse this process, to restore the video back to the original 23.976 frames per second. Doing so has two benefits. Reduce bitrate requirements, and reduced filesize.

There is no benefit to attempting to de-interlace a true interlaced source, when the video is to be viewed on a television only. If you plan to view it on a PC, the interlacing lines will show, which irritates some, hence the urge to de-interlace. These lines are not normally visible on a televsion.

A television is not capable of displaying both fields at the same time. It displays one at a time (i.e. 1,3,5,7, etc on the first pass, then 2,4,6,8,etc on the second pass). A PC monitor does display both fields at the same time, which makes the interlacing visible. The method of de-interlacing your talking about essentially Bob De-interlacing. It takes each field (either odd, or even) and doubles it, turning an interlaced image, into a progressive one. Another way to describe it, is to clone 1,2,3,5,7,etc to 2,4,6,8,10,etc. You end up with two duplicate frames, one for each field, of the same material, and hence, double your frame rate. You can then eliminate 1 of these duplicates. It's highly unlikely that your DVD player would handle an MPEG that was 50 frames per second.

The answer to all of your questions are here: http://www.lukesvideo.com/interlacing.html