When placing 23.976p footage in a 29.976p timeline, 3:2 cadence is applied to the clip but that cadence is calculated based on the source clips frame 0 and not the sequence's frame 0.
The result is if you took the same clip in a comp and placed it multiple times but at different starting points, the cadence would shift based on the source clips starting point. This means if you want clean cadence you need to 'edit on cadence' which is beyond tedious.
Other software would always print cadence based on the sequence frame 0, so regardless of where the footage is positioned in time, the same cadence pattern is applied across time to all clips in the same cadence pattern.
Is there a way to get AE to behave like this?
The other odd thing abotu AE is it applies cadence as AABCD, so the first frame of every clip is doubled. There are many different cadence patterns but this one is not usually encountered, ABCDD is more standard. Is there any override to this?
When motion picture film was transferred to video, the frame rate was kept at 29.97 for the video, the film chain (the thing that advances the film, was slowed from24 to 23.976, and a 3:2 pulldown was used to interlace the footage.
When you cut the video, the cadence depended on where the cut was made. Sometimes the frame was A on the first field and B on the second. Other times it was A on the first and A on the second field. Sometimes B, B, sometimes BA. It all depended on when you made the cut.
AA does not mean the first frame is doubled; it means that the first field and the second field come from the first frame of the original footage. Both fields are from an identical time. If you go to another frame, you will get one field from one frame and the other field from the next one. Any motion blur in a frame with fields from different times will be slightly exaggerated, but most people cannot see it unless you pause the footage and examine the motion blur very carefully.
There is no such thing as ABCDD. I don't know where you got that. There's frame A, then Frame B, and you need a combination of fields from the two frames before you get to the next A frame. That is why some frames have both fields from one frame, but every third frame has one field from one frame and the next field from the next frame. That's how interlacing works.
After Effects does not leave holes for the scan lines. You can see the individual fields in interlaced footage if you double the frame rate of the comp. 3:2 Pull Down is only used with interlaced footage. No modern cameras shoot interlaced. Broadcasters still interlace the signal, but almost all content now has identical fields.
I know how this stuff works. I have been in the business since 1970 and shot 16mm and 34mm motion picture films for major networks and some of the biggest advertising agencies in the country for the first 20 years of my career. In about 1975, a friend of mine named Clive Tobin, was a master technician and inventor of the best telecine systems (film to tape transfers). He built a film camera motor that ran at 29.97 fps, would sync to a Nagra recorder, and I bought one so we would not have to add 3:2 pulldown to footage shot for commercials. I got a lot of work because my film to tape tranferes looked better than anyone else's. Pretty soon, a lot of the majors were shooting at 29.97 instead of 24.
There are very few modern cameras that will shoot interlaced, so the 3:2 pulldown/interlaced question is mute. Dropping 23.976 footage in a 29.97 comp will give you some blended frames, but it would be an extremely rare situation if you needed to interlace the footage to make it look better.
Thanks Rick for taking the time to respond. However I think you mis-understood part of my question or the info I provided as nothing about my workflow or my question has to do with interlaced media, nothing has been telecine'd. These are camera originals, which happen to be shot on a RED.
In AE and most all NLE's, cadence is added in a 3:2 pattern to progressive material, regardless if the frame rates are "pull-down" frame rates (23.976/29.976) or whole value frame rates (24/30).
When I am describing a frame pattern of AABCD, I am describing the 3:2 cadence progressively, where the A frame is the :2 and the BCD frame is the 3: of that cadence pattern.
Again with the caveat that we're talking about non-telecine, purely progressive workflows, most NLE software will repeat the fourth frame to create the fifth, an example would be ABCDDABCDD, etc...
AE repeats the first frame as in AABCDAABCD, etc...
AE is not adding frame blending by default, which is a good thing, it's just adding clean progressive cadence.
However, it bizarely (IMHO) adds the cadence to the source clips frame 0 once added to a comp vs. adding the cadence from the comp level down the way most NLEs do (Premier, Resolve, Avid, etc...).
This means if you slip shots around in a comp, you have to worry about cutting on cadence which is horrendous.
A solution is to stay native to the 23.98/24 footage and nest the final comp into 29.97/30 but It would be better if AE detected the frame rate difference (which it clearly does) but always add that 3:2 based on the comp starting frame, not the source clip starting frame.
Perhaps there is a script that solves this... thought I'd flag this to the community and see what others have found.