Certainly no one can see the difference between film run at 24fps or 25fps but that's not what this is about rather it's about how it gets processed in post. And I have to thank Grazie and his Glow problem for the spark that lit the fire in my brain here.
25p pretty well always ends up in video as 25PsF. One can very simply merge the fields to get back the original frame.
24p ends up as 24PsF with two possible variants defining how the pulldown is applied. A simple field merge will not produce the original frames for every frame, to get them back some intelligence is needed and only for 24pA is it 100% accurate.
So what?
Well it doesn't mean much until you start applying FXs. As Grazies problem highlights merging interlaced fields to add FXs can get things screwed up introducing flicker. Yes, interlaced video is a PIA.
But before the progressive scan mob get all uppity consider this. As i noted above 25p will always field merge to produce a correct frame, the fields will truly have no temporal separation. One for the PAL mob I guess.
Now for 24p if we do a IVTC so we're apply FXs to the original frames then again all is sweet. But what if we didn't. We've now got an interesting set of problems. For certain frames in the sequences with an FX that relies upon merging fields we're inevitably going to introduce flicker or judder on moving objects!
Why does any of this matter. We'll I've read a number of posts (mostly elsewhere and I suspect they're not Vegas users) from people saying they're abandoning 24p as it's just too jerky, has too much flicker. Now I found that odd, properly shot telecined film always looks fine to me and some of these posts were from guys that had plenty of film experience.
So my gut feeling is there's nothing wrong with their 24p footage, it's how it's being processed in post. Now I don't know much about how any other NLEs apart from Vegas process 24p. I know that with care (removing pulldown) in Vegas one could avoid these problems. Certainly food for thought on a cold Saturday morning.
And then again I could be totally way off beam....
Bob.
25p pretty well always ends up in video as 25PsF. One can very simply merge the fields to get back the original frame.
24p ends up as 24PsF with two possible variants defining how the pulldown is applied. A simple field merge will not produce the original frames for every frame, to get them back some intelligence is needed and only for 24pA is it 100% accurate.
So what?
Well it doesn't mean much until you start applying FXs. As Grazies problem highlights merging interlaced fields to add FXs can get things screwed up introducing flicker. Yes, interlaced video is a PIA.
But before the progressive scan mob get all uppity consider this. As i noted above 25p will always field merge to produce a correct frame, the fields will truly have no temporal separation. One for the PAL mob I guess.
Now for 24p if we do a IVTC so we're apply FXs to the original frames then again all is sweet. But what if we didn't. We've now got an interesting set of problems. For certain frames in the sequences with an FX that relies upon merging fields we're inevitably going to introduce flicker or judder on moving objects!
Why does any of this matter. We'll I've read a number of posts (mostly elsewhere and I suspect they're not Vegas users) from people saying they're abandoning 24p as it's just too jerky, has too much flicker. Now I found that odd, properly shot telecined film always looks fine to me and some of these posts were from guys that had plenty of film experience.
So my gut feeling is there's nothing wrong with their 24p footage, it's how it's being processed in post. Now I don't know much about how any other NLEs apart from Vegas process 24p. I know that with care (removing pulldown) in Vegas one could avoid these problems. Certainly food for thought on a cold Saturday morning.
And then again I could be totally way off beam....
Bob.