Correct Frame Rate Conversions in Post, Monitor Playback & 3:2 Pull-Down

[Allan Kader]

Hey folks,

 

Just a few questions about Correct Frame Rate Conversions in Post, Monitor Playback & 3:2 Pull-Down during post that I hope some of you can help answer.

 

CORRECT FRAME RATE CONVERSION

• When working with footage with different frame rates (24, 23.976, 30fps etc), what's the best timeline settings? If I have the most 24fps footage, should I set my timeline to 24fps and import the other footage into that timeline?
• What happens to the 23.98fps footage when imported to a 29.97fps/29.976fps timeline? Does, e.g. Premiere Pro/Davinci Resolve make an automatic 3:2 pull-down?
• What happens to the 25fps footage when imported to a 23.98fps timeline?
• When does optical flow, frame blending and motion interpolation come into play when conforming from one fps to another fps?
• I know that in most NLE's you can "interpret" the footage and change the native frame rate to something else but when would I want to do that? Maybe some examples would help me understand a bit better.
• When does reverse 3:2 pull-down come into play in post-prouduction?
• What's the difference between 29.97 and 29.976, and when are they used?
• What's the difference between 23.976 and 23.98?
• When you have 24fps footage that you intend to show on television in NTSC (60Hz), do I need to make a 3:2 pull-down? Which then makes it into a i60 footage (which is interlaced for broadcast), correct?

PLAYBACK

 

• When playing back 24fps/25fps/23.98/23.976 fps footage on an external grading monitor, should I set my monitor to 50Hz to avoid judder? And for footage with a higher fps, should I set it to 60Hz?

 

Thanks :D

Edited January 24, 2018 by Allan Kader

[Perry Paolantonio]

I can answer some of these questions, others with more experience in Premiere will chime in too, I'm sure...

When does reverse 3:2 pull-down come into play in post-prouduction?

 

When you need to move 23.976 footage into a 29.97i format (note the i - it only works with interlaced footage because pulldown is field-based)

 

What's the difference between 29.97 and 29.976, and when are they used?

29.976 isn't a thing. Not sure where that number came from.

 

What's the difference between 23.976 and 23.98?

 

 

Same thing - 23.98 is shorthand for 23.976. But if you're working in an application that allows you to manually specify the frame rate, you must enter 23.976. If it's a pulldown menu and it's labeled 23.98, it's likely 23.976 under the hood - at least for professional applications.

 

When you have 24fps footage that you intend to show on television in NTSC (60Hz), do I need to make a 3:2 pull-down? Which then makes it into a i60 footage (which is interlaced for broadcast), correct?

 

The 24fps needs to be 23.976, then pulldown is added to make it 29.97i You can't go directly from 24 to 29.97i without first making the 24, 23.976.

 

When playing back 24fps/25fps/23.98/23.976 fps footage on an external grading monitor, should I set my monitor to 50Hz to avoid judder? And for footage with a higher fps, should I set it to 60Hz?

A grading monitor should adjust to the signal on the incoming SDI connection. If you're using a computer monitor via HDMI or something similar, you're not using a grading monitor. Via SDI, a grading monitor will typically resync to match the timeline setting automatically. Can't help you if you're using a computer monitor for grading - that's generally considered to be a bad idea because there's too much chance you're not getting an accurate signal.

Edited January 24, 2018 by Perry Paolantonio

[Allan Kader]

Thanks :)

[Aapo Lettinen]

if your different frame rate shots are mostly MOS (no camera or external recorder sync audio needed to use in perfect sync with the image) then you can do the close-to-your-base-fps format conversions by simply conforming the material to the different framerate from interpret footage settings WITHOUT ANY framerate pulldown conversion. this way you will maintain 100% picture quality on those shots and will only need to compromise on the shots which absolutely must use original location audio in perfect sync.

 

For example, if outputting 24.00fps you would ideally shoot 24.00fps on your main camera.

If having MOS shots on other cameras on 23.98 fps, 25fps you can just conform them all to 24 before editing and you will maintain all picture quality.

If having 29.976fps or 30.00fps MOS footage you can do the pulldown conversion IF the somewhat correct playback speed for the MOS action is important.

IF NOT, you can just conform these 29.976 and 30.00fps shots also directly to your base fps 24.00 before editing. they will be slightly slowed down but that may be OK for mos stuff anyway.

 

If needing original sync sound from location (for example someone talking to a gopro shooting 29.976 and you will need to use the camera audio track for the shot) then the framerate conversion would be needed because you can't change the playback speed ( =clip lenght)

[Aapo Lettinen]

I personally conform the fps in interpret footage ALWAYS when the camera/location audio is not absolutely needed. especially in nature docs where all the sounds will be done afterwards anyway and camera/loc audio is unusable 99.9% of the time.

The other thing is that nature stuff tends to be high speed most of the time and conformed anyway either in-camera (sensor fps higher than base fps on file metadata) or when making previews and offlines for editing (for example conforming in Resolve on import) or before editing on fcpx or premiere etc. when importing raw material...

 

technically it would be possible to do the framerate conversion the other way around than typically made (picture changes, audio stays intact) : one could conform the picture stream to different fps without framerate conversion AND then add speed+pitch change to the audio track just like done here in Europe when making 25fps tv versions from 24fps movies.

This would be actually quite handy if you need to use the camera audio track because most usable camera (sync) audio would probably be dialog and the sound designer would add sfx and foley and music etc. on the top afterwards so the speed change would go totally unnoticed and one would still maintain 100% picture quality compared to the original :)

[Allan Kader]

Thank you so much, aaop lettinen :D