true 24p vs fake 24p
Posted 31 December 2005 - 11:32 PM
Every camera is either native i or p scan right ? The HDV cameras try to appeal to the prestige that goes with the film look by promising a 24p look even if the camera shoots interlaced. Others like Panasonic make a native 720p camera.
I want to know why thepan/tilt strobing looks worse on all these cameras than it would shooting film at 24 fps. Even on the true 24p HD cameras I have seen at trade shows it looks worse. Sony has a new HD camera out, maybe the XD cam which I had a play with and I noticed the srobing in the viewfinder was a bit better than the HDV cameras but not much, certainly worse than what I was used to shooting 16mm at 24fps.
On the other hand this camera was being fed to a HD monitor and I noticed the strobing was not as bad on the monitor as it was in the viewfinder. So there must be a data display issue here. My question is can I trust the stobing that appears in the viewfinder firstly on true 24p secondly fake 24p. Which of the fake 24p systems is the best and how do they work or try to work ?
I would also appreciate a good explanation of true vs fake 24p and the editing issues.
Posted 31 December 2005 - 11:44 PM
Remember that you never see 24 fps film shown 24 times per second (except in the viewfinder, and that's very flickery). We see it flashed 48 times per second when projected, and on NTSC, we see it with a 3:2 pulldown at 60i.
So while 24P shown later on DVD or on film does have a slightly different feeling to motion blur, even though in theory at 24P with a 1/48th shutter, it should match 24 fps film with a 180 degree shutter, there maybe is a slightly "crisper" look to the motion... maybe it's a contrast / depth of field / edge enhancement issue, maybe it's the difference in an electronic shutter versus a mechanical one. But it's not radically different, if you look at a 24P movie on DVD like "Revenge of the Sith".
As far as "fake" 24P, it just comes down to sampling rate -- at 24P, you sample a moving object 24 times in one second, while at 60i, you sample it 60 times per second. So "fake" 24P systems have to first process motion sampled faster than 24, and then process it again to record it back to something like 60i.
In the case of the Sony HDV camera, the 24F process takes 60i capture, processes it to look more like 24P, then adds a pulldown to convert it back to 60i for recording at 1080/60i. Whereas at least with the Canon XL-H1, it changes the capture rate to 48i, and then converts that to 24P, and then adds a pulldown to get to 1080/60i. That's a little smoother (converting 48 to 24 instead of 60 to 24).
Edited by David Mullen, 31 December 2005 - 11:48 PM.