shooting film at 24fps, PAL telecine question

SNOOPYLEMANS

New member
dear mr mullen,

i am an avid reader of your generous posts here and on the cinematography list, thank you for sharing so much of what you know. i am a young producer working in china, and such knowledge has been very helpful and encouraging to me.

here's the situation:
we are shooting 35mm film in china, and our editor is from the US. we've shot at 24fps, and our telecine is coming back on DV PAL at 25fps, our editor is working in Final Cut Pro, and we will be going back out to film at 24fps.

the question is regarding the telecine method: they are adding a frame every 24 seconds. our editor is skeptical about how this method will affect our film out.

is this a preferred or acceptable telecine method for our end use, which is film? will there issues on synching sound, working this way? should we consider having them redo the telecine?

thank you in advance for your help.
sincerely,
ays
 
You're not going to use a PAL transfer as the basis for a transfer to film, are you? Seems like a waste of quality if you spent the money to shoot on film. Unless this is a documentary with a lot of PAL video clips mixed into the footage.

Or are you talking about conforming the negative using the EDL from a PAL edit? I don't see a problem with that if you can correspond the time code to the keycode.

Anyway, if you are really talking about transferring a film-to-PAL telecine transfer to film using a film recorder -- with the hit in resolution that would cause -- it was probably a 1:1 transfer, i.e. the film was just sped-up to 25 fps when transferred, so each frame of film now is represented by one frame of video. So a transfer to film would also be 1:1, but the transfer would be projected at 24 fps. So there is no real problem with the speed of the footage. The only problem is the sound, which if cut in PAL, will have to be speed and pitch-adjusted to convert to 24 fps.

Most people, if they really wanted to do a digital intermediate, would re-transfer the film to HD or 2K for a transfer back to film. The PAL transfer would just be used for the offline cut to create an EDL.
 
Yes, that is correct: we are using the PAL edit list to conform the negative. I was trying to say that film is our final output medium, and not video.

The question, (I think), is that the telecine house here in China may not be running the 24fps at 25fps (making it 4% faster) but rather is adding 1 frame to every 24 frames... Does that make sense? The understanding I have is that this is not as common a method for PAL telecine as the 4% change. And I am concerned about how we should adjust our FCP post workflow to accommodate it?

Complicating matters is a language and possibly an experience barrier (besides my own ;-) -- that the lab and the telecine house do primarily commercials and music videos; we cannot rely heavily on their experience with film. I just want to ask the right questions...

Thanks very much.

ays
 
On this topic, I am very confused by the NTSC / PAL barrier.

I just finished shooting in Europe, but the film will be posted entirely in the US. We shot 24 fps instead of 25 and had the lab telecine to NTSC instead of PAL to be able to use NTSC editing facilities.

The PAL system just seems to be so superior, because a film shot at 25 fps can be edited 25 fps and displayed on a TV at 25 fps.

With NTSC, we are shooting 24, and when we go to NTSC we create fake frames to accomodate 29.97 for TV display. So in NTSC there is a 4:5 ratio between 23.976 and 29.97, meaning one in every 5 frames is some type of composite false frame... am I right?

I guess my question is... with us stuck on the NTSC system, what is even the point of shooting 24 fps except if there will be a projected film release? If the film is, say, a low budget short shot on 16 and will exist only in the digital world after telecine, it seems like a lot of quality is being sacraficed by NTSC.
 
I'm gona take a crack at this BUT

when you make your telecine transfer from film to NTSC.... you'll most probably have a 2-3-2 pull down.. which means that the frame rate is 24 frames pers second.. giving you that cinematic feel in Digital, but its compensating for those extra frames that NTSC 29.97 frames usualy play...

... so yea you should have much problems.. it just looks more cinematic... as in 24 frames..

its the same way the Canon GL2 works... but thats where I'm pulling my general conceptions of the transfer
 
The reason to still shoot 24 fps in 16mm even if the release will be on video is that:

(1) most movies are shot at 24 fps and given a 3:2 pulldown to convert to 60i NTSC, so we are used to that look in NTSC countries;

(2) while shooting at 30 fps would in theory improve temporal artifacts and create a smoother look for a transfer to NTSC, it is not a good frame rate should a PAL transfer also be necessary;

(3) most movies released for home video go to both NTSC and PAL markets

So stick to 24 fps, if not, then 25 fps, but avoid 30 fps.

Now some people also think that 30 fps transferred to NTSC with no pulldown looks TOO smooth and therefore video-ish, but I don't think that is real problem -- 30 fps film or 30P video still looks more film-like than 60i video capture. It becomes a matter of taste whether 24 or 30 looks "better". Certainly PAL viewers don't rely on a pulldown to make something look film-like. But the real problem with 30 fps / 30P is the poor conversion it makes to anything other than 60i NTSC or 60i or 60P HDTV.
 
The 24P "Advanced" format in some NTSC cameras does repeat every fifth frame I think to convert 24P to 60i, but this is done just to make it easier to extract the excess frame and edit in true 24P in post.

Standard pulldown instead has a pattern that causes every third video frame to contain one field from one original film frame but the next field to be from the next film frame, making the pulldown smoother than 24P Advanced (because there isn't a repeated frame, just a repeated field) but harder to extract.
 

Network Sponsors

Back
Top