-
29.97 Clip in a 23.98 Sequence… FCP Frame-Rate Converting?
Something strange just happened, which might effect my entire project’s workflow. I can’t quite explain it… in fact, I think it’s too good to be true… and hope that someone might be able to explain what’s happening.
I’m in the process of “HD-mastering” an old, feature film… originally shot on DV NTSC… for D5 / film-out. After exhaustive testing and terrible results, I found this method to work the best.
I take each individual DV clip and import it into Shake. I then uprez it to 1280×720 / deinterlace / antialias and export it as Uncompressed 8bit 29.97. I then take that clip and re-import it into Shake and frame-rate convert it, yielding an Uncompressed 8bit 23.98 clip. All the clips are then nested in a 23.98 Uncompressed FCP timeline, where I check them for smoothness and make sure they synch with the 23.98 audio track. It’s been painstakingly slow but this process has yielded very good results.
Now, the other day, I accidentally missed the second step and nested one of my Uncompressed 29.97 clips in the 23.98 time-line and, before catching my mistake, I exported the whole sequence. It took a few minutes to render (that’s when I knew something went wrong) and I realized my mistake. But, nonetheless, I watched that exported sequence and was absolutely shocked to find that the 29.97 clip synched and had the right in and out points (i.e. there was no slow-down effect, etc). It came out as if the frame-rate conversion had already been performed.
As a check, I went back into FCP and exported just that clip from the timeline and checked its values. 23.98, etc. just like the sequence / just as intended. My next check was to take that same clip and perform my regular work-flow and compare the two. I took that clip though my Shake frame-rate conversion cycle and compared the two clips. Identical size and parameters… and, to my shock, the FCP clip looked slightly better!!! With hours less rendering time!!!
Can anyone explain this? Did FCP perform a frame-rate conversion?
-
12 Replies
-
This is in FCP 6.0.1
-
https://www.apple.com/finalcutstudio/finalcutpro/support.html
This was the key feature to FCP 6.
It doesn’t always work though, so be careful.
-
I realize that FCP 6 is supposed to “play” everything in the same timeline / sequence, but, when I export, does it actually frame-rate convert? Does anyone have any experience with this?
If so, is this is a reliable work-flow? Meaning, if it “looks” right, am I to assume that it “is” right?
Thanks.
-
[m31films] “Does anyone have any experience with this?
If so, is this is a reliable work-flow? Meaning, if it “looks” right, am I to assume that it “is” right?
“What is the true framerate of your source footage? Was it actually shot at 30fps, or is it 24fps w/ pulldown added?
-
DV NTSC 29.97fps, 16:9… shot with a PD150, if I remember correctly.
-
But the source shouldn’t really matter, seeing as how FCP only sees what I get out of Shake after the uprezing / deinterlacing… that being:
Uncompressed 8bit, 4:2:2, Deinterlaced, 720p, 29.97fps QT
-
[m31films] “But the source shouldn’t really matter, seeing as how FCP only sees what I get out of Shake after the uprezing / deinterlacing… that being:”
The actual framerate of the source matters a great deal. If it was shot at 24fps, the process is simple. You’re not actually changing the framerate, you’re just removing the pulldown. Straightforward and no margin for error.
But converting actual 30fps to 24 is completely different. In this case you are changing the actual framerate. This is more like voodoo in the sense that the results can vary dramatically depending on what tools you use. I do not know how well FCP 6 does this, but this type of workflow is supported which likely explains your positive results. Haven’t tried it myself but I hope someone else who has chimes in. I have done it in FCP 5 and the results were not very good.
-
Exactly. Hence why I was doing all the frame-rate converting in Shake. It was the only method that yielded good results.
That’s why I find FCP’s converting so perplexing.
I can post sample clips if anyone is interested in inspecting / double-checking my results.
-
[m31films] “I can post sample clips if anyone is interested in inspecting / double-checking my results. “
Please? I’d love to check it out. I have gone the other way (put 24p clips in a 29.97 timeline with crap results) and I’d like to see this way.
Jeremy
-
JeremyG, here is a link to some before and after clips from my frame-rate converting in Shake — DV NTSC to 720p HD.
https://www.m31films.com/2007/09/25/sandcastles-hd-mastering-step-by-step-before-and-after-results/
In terms of the results I have been getting from FCP vs those from Shake, I’m in the process of re-rendering clips that would be sufficient for proper comparison.
Lastly, in terms of going from 24 to 29.97, all you need is a drop-down, no?
Reply to this Discussion! Login or Sign Up