Problems with 24p vs 24pA & desperate for help...

theiverson

New member
I want to shoot a project in 24p, and I'd prefer to do it in 24pA to obtain motion that's more or less identical to film. I'm not planning on going to film, though.
But as I understand it (or thought I understood it)...24p normal uses the 2:3:2:3 pulldown cadence to give the illusion of 24p, but it records at 29.97 fps in compliance with NTSC standards. 24pA uses 2:3:3:2, and it's intended for extracting true progressive footage, in the right timeline, on the right NLE. I have Final Cut HD, with Compressor. The former has a capture and timeline setting for 'true' 24p editing and output, and the latter lets you choose between NTSC output and a particular field dominance--or 24p output, which renders the idea of field dominance irrelevant.
Anyway, I'm shooting some test footage (on blank Sony/wet tapes) in both modes. I'm putting the 24pA footage through the pulldown removal process, capturing the 24p footage as NTSC-DV, and going through the requisite editing and compression techniques; I want to see how the footage looks on a DVD before I proceed; the final project is going out to DVD and possibly being projected via DLP. OIS is off, and I'm keeping on track with the fact that you have to move much more gradually in 24p than you would in 60i or 30p.
The 24p footage looks great on DVD (played on a 27-inch set). It was compressed fairly heavily (I'd say around 3.9-4.1 mbps); I couldn't see any compression artifacts on the TV screen, and they were very faint when being played back on the computer screen. And the movement was perfect, with the right amount of blur on moving images. Of course, the computer screen introduced some comb artifacting, but not much.
This is all in almost direct opposite to the captured and compressed 24pA image. It looks terrible, for lack of a better word. The video has an extreme case of strobing--any movement at all in the image makes it look like a streaming Web video that's had a sudden loss of bandwidth. There aren't any comb artifacts or interlacing errors--it's true progressive I'm looking at, but it's not fluid at all. There are compression artifacts everywhere you look--pixels, moving black blocks in the darker areas, and in one of the shots--a slow pan with an oak tree in the background--the leaves themselves seem to be comprised of one or two pixels. It looks incredibly low-res, and it had the exact same compression setting as the 24p standard footage, if not a little higher (probably around 4.1-4.4 mbps).  I've checked the settings on both scene presets, and they're identical.
The first or second time this happened, I just assumed that I was doing something wrong. So I started researching all of the possible settings within the program, started combing through the boards here and at 2-pop, and I haven't found anything. Has this happened to anyone else? I mean, all isn't lost--the 24p standard footage looks fine, and I guess what I want to know is how important going to raw 24p is if you're not going to film, and the closest you're going to get to it is a public showing at which a DLP/LCD projector is going to blow up the image on a massive scale. I'd rather have the image slightly soft/blurred than plagued with compression artifacts or have the kind of motion it has now.
As kind of a P.S...if this is something that's covered in Barry's book, that might be another way out of it--I've been too long in getting it, and I'm gonna order it and the DVD later today.
 
Re: Problems with 24p vs 24pA & desperate for

Re: Problems with 24p vs 24pA & desperate for

Well, first, you’re falling prey to a fundamental misunderstanding of the difference between 24pN and 24pA.

There’s no “illusion” involved in 24pN. It’s true 24p. The *only* difference between 24pN and 24pA is how it’s recorded to tape. Both are recorded to tape in 60i; 24pA merely breaks down and records the frames in a slightly different arrangement so as to make it easier for NLEs to extract the original, true 24p frames. Once you get them on a 24p timeline, there’s no difference between them. They’re both 24p, and any NLE which can remove the 2:3:3:2 pulldown of 24pA can remove the 2:3 pulldown of 24pN.

So, when you render out from a 24p timeline, there shouldn’t be any difference between the two. I have to think you rendered them differently.

If you were to play 24pA and 24pN footage straight from the tape on an NTSC TV, you *would* see “strobing” with 24pA footage, because of the different cadence – ONE of the 24p frames is spread over three fields in the cadence and it results in an ever-so-slight pause in the picture, which is noticeable but not fatal.

Now, there are differences besides pulldown cadence between the F5 and F6 scene files, at default. I don’t think they should be causing the issues you’re describing, if they’re that pronounced, but a difference between the detail and matrix settings could account for SOME distortions, stair-stepping, etc.. The F5 setting is geared more toward display on an NTSC TV, whereas F6 has the filmmaker more in mind. Both are workable for either, of course, but you should keep it in the back of your mind when you’re shooting and handling the footage later.

The tapes have NOTHING to do with your problem.
 
Back
Top