-
Why AE wants to remove a pulldown in this situation?
Hi,
I’m not sure this is right according to what I’ve understand…
I have shot footage with the Panasonic HDX100 in 24p mode (not advanced), edited the material in Premiere Pro 2, then got into AE to import the scene.
After many hours of effects, I’ve realized that AE has listed my footage as being 23.976 and has put a pulldown removal on the footage…
As far as I know, the camera makes the transformation within the camera and is set at 29.97 when captured.
Am I mistaken here? Do I need to tell AE to reinterpret all the footage with the option listed as “use the frame rate from file : 29.97”. If so, why did AE decided to do this?
I’ve done lots dynamic masking with keyframes and was wondering why there were many duplicated exact frames, resulting in me having to double the kf every time a second dupe would arise.
Oooooh, I hope this is not problematic….
Thanks for your clarification,
Pat
-
10 Replies
-
There’s a tip on the “Remove 3:2 or 24Pa pulldown from video” section of Help that might address this issue.
“One of the formats used by Panasonic P2 cameras is 24p over 60p using pulldown, but it’s not a regular pulldown cadence, and After Effects doesn’t automatically identify it correctly.
Use one of the 24pA cadences, even though this format doesn’t really use 24pA. It should all work out anyway.
The cadence appears to be 2:2:2:4.”
I got my information for the tip from a bug report, so I can assure you that all the right folks know about this issue.
———————————————————————————————————
Todd Kopriva, Adobe Systems Incorporated
putting the ‘T’ back in ‘RTFM’ : After Effects Help on the Web
——————————————————————————————————— -
Humm…. I’m not sure I understand…
I don’t have 24pA material, but strictly 24p, so as far as I know, my footage is 29,97 since it’s converted inside the cam already, meaning that I would believe it should be also interpreted as typical interlaced footage in AE, no?
My renders look *alright*, but I’m perplexed by a somewhat blurriness to the ensemble of shots… Could my concern lead to that result?
I have no idea what is the 24pA cadences and how to use that number. I use After Effects 7, and I have not seen this listed in the interpreting footage menu…
I thank you for your help, but I’m still confused!
Pat
-
The footage may have been captured (input into the computer) at 23.976. We use a Firestore drive, and the new firmware for the drive allows you to capture 24P right away to Quicktime, with no pulldown. It’s eerily easy. So you may actually be working with 23.976 footage.
If I recall, if you take progressive footage and separate fields, it gets soft.
We don’t need to always trust AE’s interpretation. Have you tried reinterpreting with no field sep, no pulldown removal, opt-clicking on the footage in the Project window, then stepping through the footage with the R arrow key? If the footage moves forward one frame with each click and you see no interlacing, it’s progressive and correctly interpreted.
Let us know what happens.
-
doesn’t the hdx100 shoot to tape or p2?
if you shot to tape, then i’m pretty sure the footage had to be recorded with a pulldown to bring it to the standard ntsc frame rate 29.97. you would then remove the pulldown at capture or prior to editing to work with it in its native 23.976 frame rate.
i’m not sure this applies to the hdx100, but with the hvx200, i know that even if you shoot to p2 there are some options for 24p… there’s 24p, which records 1 duplicate frame every 4 progressive frames, making it 29.97 until you remove the duplicate frame. and then there is 24pN (the ‘n’ stands for native) which only records the 24 progressive frames, no extra duplicate frame…. i also remember an issue recording to a firestore that only allowed the 24p mode (with the extra frame), but from what steve said, it may be resolved now.
Kevin Camp
Senior Designer
KCPQ, KMYQ & KRCW -
Oops. I was talking about the HVX. Shows how much I pay attention.
-
Arrrgh, my big mistake here, I don’t know why I wrote this, but I’m indeed working with the AG-DVX100 camera and with DV tapes (not HDX100)…. Maybe this will clear things up a little…
Sorry to you all for this! Thanks Dave for noticing it.
I’ve made a test though re-rendering a scene with the footage without the pulldown, and it looked pretty much the same aside from some jagged edges here and there, and from what Steve says, I believe I haven’t previously encountered interlacing in my shots, so it might be ok after all. (I can’t verify at the present moment)
Now does my problem make more sense or everything is fine?
Pat
-
if you shot 24p to dv tape, then it was recorded with a pulldown. you would have liked to have removed the pulldown prior to editing, but it sounds like you have already edited…
at this point you may be able to just re-interpret the footage in ae (select footage, then file>interpret footage>main) and choose not to separate fields, then change your comp(s) to 29.97fps. you will, of course, see interlacing in 2 out 5 frames and this may cause problems with you masking (and other things), or it may not… you’ll have to see.
if this doesn’t work, you’re in for a lot of reworking… once you’ve edited with a pulldown, there is no easy way to remove the pulldown. the pulldown will need to be removed for each shot and then reassembled.
Kevin Camp
Senior Designer
KCPQ, KMYQ & KRCW -
You guys are right… I realized I did edit in PP2 in a regular DV NTSC 29.97 project.
BUT since AE has actually decided on its own to re-interpret the footage as I now understand should, i.e. from 23.976 separating the fields, w/ removed pulldown (in my 29.97 comps), resulting in my first concern about duplicate frames, I wonder if this “late interpretation” (in AE instead of PP2) winds up affecting the image quality in my final renders. I’m asking this because, my renders actually look fine on TV/DVD aside from sometimes having a suspicious softer feel… Or could that just actually be the typical DVD mpg compression degrading the image (which I put to the max BTW at CBR 9000)
Could this be the compromise of not having dealt with a true 24p project in PP2 in the first place?
Thanks ALOT to both of you on this matter,
Pat
-
OK:
I’ve tried re-interpreting the footage in AE7 back to the “use frame rate from file 29,97)” and automatically I can see a difference in the still image, as it gets a blurrier look.Here are 2 other things I’ve tested:
1-Made a 24p project with about 6 shots in a row in PP2 with my AGDVX100 footage. The footage had a remove pulldown made automatically by the program. So all should have played slick in 24… Well strangely enough, it was ultra jerky, had a major field-like motion problem. I checked the project’s playback settings and there were 2 choices : (1) Repeat frame ABBCD and Interlace Frame 2:3:3:2. I tried both and the jerky image situation still goes on.
Nonethless since this was just a test, I didn’t care about this “bug” and imported that project into AE7. It detected my sequence as being 23.936 and removed the pulldown as well, separating the fields too (lower first).
I made an AVI, let the defaults there (Use comp’s rate, field render off, pulldown off)
2-Made a STANDARD DV NTSC 29,97 project with about 6 shots in a row in PP2 with my AGDVX100 footage. Basically what I’ve done in my previous full project.
I imported the sequence into a new AE7 project. Again removed the pulldown, and kept it flowing in my 29,97 comp. Still made the duplicate frames but no jagged edges on the 2 frames every other 3. So an in all same thing as my concern.
I exported a AVI out of it let the defaults there (Use comp’s rate, field render off, pulldown off).
NOT : None of these comps had spfx or color correction added in AE7. I just wanted to compare the footage in its raw form.
Made 2 DVDs out of this test. I used TMPGenc 3.0 encoding/authoring. Both detected the footage as one being progressive and the other interlaced. And yes I have Progressive Scan DVD Player.
THE RESULT: BOTH look exactly the same, no quality difference. Unless my eye is not trained enough, but there’s definitely nothing drastically dramatic if anything. So I guess I’ll be quite alright with what I’ve done so far, which is a huge relief.
Although, I must admit that my fxed scenes do still look a little softer. Maybe it has to do with me using Cinelook and Film Damage plugin filters (with NO grain or defocus options enabled I must say) and a copious amount of color correcting filters….
I thank you guys once more for having stuck with me on this one, I have learned anyhow!
Maybe what kept me instinctively going for the Standard DV NTSC project in PP2, is the fact that I have a vague memory of reading somewhere to not use the 24p project feature, cause it was actually only valid for 24pA in PP2… Could that have anything to do with the jittery playback motion I earlier mentioned? To be on the safe side next time, I would rather deal with 24 straight progressive frames and not have the double frames pulldown process going on especially when doing rotoscoping and extensive masking effects. So any clue on resolving this one would be great!
Pat
-
Alright, alright! I’ve read most of it, and it is safely kept as a bookmark for future uses. ; )
Thanks!
Pat
Reply to this Discussion! Login or Sign Up