Quick and dirty 60i > 24p method

[Dave Valencic]

Make an After Effects timeline, set the framerate at 23.98. Drop your 60i (29.97fps) project onto the timeline, render at 23.98.

That's it.

As long as it was shot at 1/60th sec, it looks perfect. I'll post a sample once I get something edited.

Still working on a good way of getting the footage back out to tape. FCPHD handles 23.98 DV quite well in a 24p timeline and will output back to tape fine.

[Frank Ladner]

Dave:

In testing, I have found that using a slower shutter speed, like 1/60 or 1/30 (thus getting more motion blur) helps with a straight render to 24 fps. However, I have only tested with footage captured in Frame Mode (using a GL2), being careful not to pan around too much, so I am interested in how 60i looks. Seems like having all that inbetween information would reduce the stuttering effect.

Also, I wonder if you precomp that and render out the 23.976 w/ 3:2 pulldown back to 29.97...how would that look? Guess I'd just have to try, eh?

Maybe setting comp 1's framerate to 59.97, separating the frames, and doing a deinterlace (giving 60p), precomping that and then in comp 2, changing the framerate to 23.976...would that work better than using 60i? I'm just thinking out loud.



Thanks!

,Frank

[Dave Valencic]

i've tried that 60i>60p>back to 24p, and i found AE's built-in interoplation much better and alot faster.

From what i've tried w/ some XL1S 30p footage, its next to impossible to get GOOD looking 30p>24p. When you have double the information with the 60i it helps alot without a very big decrease in resolution.

Although I converted an interview shot at 60i to 24p, and there were alot of diagonal lines in the background and they came out a tad chunky on the 24p version probably due to the deinterlacing.

Since I have the ability to work in a true 24p timeline in FCPHD, I've never tried adding in the 3:2 pulldown. In theory, it should work fine and look pretty much the same.

[Frank Ladner]

Oh, ok. I've never used FCPHD, but I see what you're saying. That makes sense.

[Dave Valencic]

Today i'm going to try to shoot some DVX100A video both in 1/48sec and also in 1/60sec and see if the 24p @ 1/60 looks like the 60i>24p. I'm thinking that it will look identical to the REAL 24p shot at 1/60 shutter.

Some video (especially one of a concert with alot of movement in it) looked to me like it didn't quite have enough motion blur (when converted from 60i>24p) so that makes me think that the difference between 1/48sec and 1/60sec shutter, although subtle, makes the difference. Because the movement of the 60i>24p looks to be dead on to the real stuff.

Just for fun I'll try to go ahead and get some SDX900 24p video in the next few days. In the past when I worked with SDX900 and DVX100A footage side by side, the 24p on the SDX900 seemed a little different in the movement than the DVX100A (both in 24p mode, not 24pA)

[Frank Ladner]

Dave: I'd be really interested in seeing that!

Thanks!