FCP 5.1.2 108024Pa import - works, but TC not right

Verko

Member
(edit) Barkis gave the answer to this issue in post #10 on this thread. Message to self: RTFM!

I just updated to FCP 5.1.2, using software update on a G5 Quad. Everything worked fine, and the new import window is great.

One issue, however, is that the imported clips are still using 30 frame timecode. Each frame of timecode is tied to a frame of video, but each second goes up to frame 29, when frame 23 should be the last before going to frame 0 of the next second. I am also getting some discrepancies with stated duration versus actual duration. This is with the duration shown in the clip browser, as well as in the viewer.

The upshot of all this is that currently, timecode is not useful for playing with the clips - for example, if I figure I want to go ten seconds ahead, so I just type in ten second later timecode, well, it will be more than ten seconds later. Not good. Durations are not exact either.

When I drop a clip onto a 23.98 timeline, they play normally on the timelline without rendering and everything is fine.

The clip viewer in the P2 import window shows 30 frame time code, of course, because the original is 24 frames living within 30 frames with the advanced pulldown. That's all fine. When you step through, there will be 4 frames normally, then the fifth frame is the same as the fourth (the viewer doesn't show fields).

Is anyone else experiencing this? This is a nasty gotcha if you're not expecting it.

But otherwise, the new P2 import window rocks.


Verko
 
Last edited:
I just had a chance to review what FCP is doing without being under the gun. It seems that every frame of the converted files with the advanced pulldown removed has the same timecode as the frames before removing pulldown. In other words, the new, real 24 frame file has timecode that goes, for example, frame 0 1 2 4 5 6 7 9 10 etc.

Kinda screwy; I guess I can understand their logic: just leave the same timecode frame for frame. But for actually using the clips, who would want non-sequential time code in their files?

The stated duration difference I noted was not a 20% difference as one might expect if the new files had full sequential 30 frame time code; now I know why. By looking at the "media end" timecode of an eight second file, for example, the new P2 import method produces 12:26:57:19, while importing without removing pulldown THEN reverse telecine in Cinema Tools has a media end time of 12:26:57:15.

When I tell the viewer to go "plus 30 frames" on these new non-sequential files, it goes forward one second, even though one second should be 24 frames, not thirty. Going forward +24 only goes forward approximately 19 frames, depending on the pulldown cadence. Going forward +1.00 moves forward one real second.

Since the pulldown sequence will remove the same numbered frames every second, this isn't as bad an issue; the main discrepancy comes with sub-second movements. This problem would affect most when chopping something up precisely in small increments (under one second).

Verko
 
I've been checking, I don't see any wrong TC when I import my footge. It's got the TC according to how my camera and FS-100 were set up. I tried to duplicate this problem and can't seem to do it.
 
BenB said:
I've been checking, I don't see any wrong TC when I import my footge. It's got the TC according to how my camera and FS-100 were set up. I tried to duplicate this problem and can't seem to do it.

Ben,

Thanks for your post. I am shooting to P2 cards, 1080 24pA. If I understand correctly, you are shooting the same to FS-100? Does the FS-100 remove advanced pulldown during record? If so, that's great. For me, I'm inputting 24pA through FCP P2 import, and it's just throwing away the pulldown frames (as it should) and leaving the same timecode frame for frame (which it SHOULD NOT!).

Is anyone else running into this? Not that it should matter, but I'm running the G5 Quad.

Dan, are you telling me I should be screaming, or everyone else (because nobody besides Ben has responded)? Just curious. I'm not a screamer, hehe; I just move forward, and I'm trying to analyze the problem and see what other people are experiencing. I've contacted Apple, of course. To me this IS a major issue; the fact I am talking about it at all is my version of screaming. Perhaps some (expletive deleted) dork at Apple thought "hey, let's keep the TC relationship between original frames and converted frames intact" or something ridiculous like that. In any case, I generally appreciate Apple's product line; I used to own an Avid DS Nitris, which was really the cat's meow. I owned it since 1998, but I got tired of spending $70K every 2 years or less to stay current (hardware and software). More than two years out of date and it's not fun. It took me about a year and half to switch over completely and wean myself off the Nitris. It's taken every plug-in I can find and AE, Combustion, and Maya to really replace the Nitris. And Motion is really fast, if not precise. I've done a full DI from Super 16 to uncompressed 1080 24P with FCP and AE, which worked out really well. The Synthetic Aperture Color Finesse CC was absolutely vital, and it gave me the ability to do anything I could do with a da Vinci or Pogle.

Dan, I noticed your signature line; can the P2 log software remove 24pA pulldown when converting to Quicktime (of course with correct 24 frame TC)? I looked on the website and didn't see any mention of that capability. Forgive me if I missed it. For now, I am inputing with the FCP P2 import as before, without removing pulldown, then batch reverse telecine-ing in Cinema Tools. One more step.

Thanks for the response,

Verko
 
I thought this is what 5.1.2 corrected. But you're still going into Cinema Tools to batch reverse because of some perceived time-code error relative to your 24pA clips? That's about all I could retain from the aforementioned "analysis." Odd.
 
Justin Reade said:
I thought this is what 5.1.2 corrected. But you're still going into Cinema Tools to batch reverse because of some perceived time-code error relative to your 24pA clips? That's about all I could retain from the aforementioned "analysis." Odd.

Not perceived. Maybe it's a small point to the casual user. For editing, accurate TC is important. I'd rather take the extra step one time and have sequential TC every time I use the footage in the future, than have non-sequential almost arbitrary TC. I thought I'd mention the issue so people who care wouldn't get surprised by it.

Thanks for your input,

Verko
 
I meant perceived in its literal sense, in that you "PERCEIVED" this error. I didn't mean to suggest that it was some figment of your imagination. I'm just surprised the update didn't FULLY correct this issue. I thought I was done with the days of batch reversing. :(
 
Justin Reade said:
I meant perceived in its literal sense, in that you "PERCEIVED" this error. I didn't mean to suggest that it was some figment of your imagination. I'm just surprised the update didn't FULLY correct this issue. I thought I was done with the days of batch reversing. :(

Thanks for clarifying, Justin. I didn't think I was crazy, but ya never know!

After upgrading to 5.1.2, I was so impressed with the new interface (because the previous version was so lacking), and then so disappointed to see this issue. It's almost a step backward. But it works out for now, because I can rename clips and so forth on input, batch them all to one folder, then let Cinema Tools reverse telecine in the background.

I'm sure Apple will deal with this issue soon. Right Mr. Jobs?

And by the way, am I the only one experiencing this issue? Anybody?

Thanks,

Verko
 
Clip time vs. Source Time

Clip time vs. Source Time

HI Verko,
have you tried changing your project settings to clip time instead of source time? Go to the project settings menu item ( think its under the Edit menu) and change over to Clip time, that should show you your 24fps timecode it for the whole project. You can also change this in each timecode window individually if you control click on the timecode field. Let me know if that gives you the correct timecode display.

barkis
 
Thanks Barkis!

Thanks Barkis!

Barkis said:
HI Verko,
have you tried changing your project settings to clip time instead of source time? Go to the project settings menu item ( think its under the Edit menu) and change over to Clip time, that should show you your 24fps timecode it for the whole project. You can also change this in each timecode window individually if you control click on the timecode field. Let me know if that gives you the correct timecode display.

barkis

Gooooooooooooooooooooooooood call! I should have RTFM. Funny that I haven't run across this problem before; it must just affect what happens on import from P2, because I have been working with mixed 24, 30, and 60 frame clips in the same project for years, and FCP has never "interpreted" the timecode like this before. I've input from the DVX 24pA without issue for years (as far as I can remember; I haven't input anything from the DVX in at least 6 months).

Your explanation is backed up by page 443 in the User Manual, and they mention that source time is for referring back to the original tapes (card data in this case). It never occurred to me that it would show "source" time on converted clips; I assumed that it would show the proper "clip" time without having to be explicitly told to. Now I understand their thought process. Personally I think it should default to clip time, but at least now I know!

I feel pleasantly stupid, pleasant because I found out what I needed to know, and stupid because the answer was in the manual and I didn't do the searching I should have!

Thanks again, Barkis! Works flawlessly!

Humbly,

Verko
 
Back
Top