Workflow for 29.97 NTSC -> 23.98 -> 25.00 PAL

[Bouke Vahl]

Not sure what happens, but a .1 % difference corresponds with 30 vs 29.97
Thus, the 23.976 file has the same duration as the NTSC version.
Speeded up by 4% is apparantly going wrong.
(there is a lot of complains about Apple not getting these things right)

I would suggest to export the sound from the original to Wave, and speed that up. (This is a small process where only the sample rate is altered, thus a pitch shift, but for some people good. Lots of hollywood productions go to PAL this way)

To create a speeded up .Wav, use WaveAgent or QTchange, or any other audio package you have laying around.

Bouke

https://www.videotoolshed.com/
smart tools for video pros

[Neil Sadwelkar]

After step 2, where you have a 23.976 fps movie with sound that’s perfectly in sync, you need to place video and audio together in FCP and export as a QuickTime movie self-contained. There won’t be any loss if the seq settings are matched to the clip settings, ProRes in your case.

This 23.976 fps QuickTime with sound embedded can be opened in CinemaTools and conformed to 25 fps. The sound will be proportionally speeded up to 4.01% along with the picture so it will all be in sync.

———————————–
Neil Sadwelkar
neilsadwelkar.blogspot.com
twitter: fcpguru
FCP Editor, Edit systems consultant
Mumbai India

[Michael Pfost]

Thanks for both of your responses.

As you advised:

1. I went ahead and placed the footage on the FCP timeline and exported as a self-contained QT movie. Would this also work as just a reference file?

2. I then used CinemaTools to conform the file to 25fps.

3. Then I used Compressor to change the file to PAL 720×576.

At the end, the episode was in-sync and there were no blended-frames. YIPPEE! Now I just have to do this 120 more times – LOL.

I wonder why the Compressor workflow doesn’t work properly? I was hoping to use Compressor with time-remapping to do the same thing and reduce the pitch of the audio. I guess it sounded too good to be true.

Anyone have a good workflow for do a large # of files? I already have each episode in a ProRes file 23.98 fps.

[Neil Sadwelkar]

There’s a section in the Compressor presets that does adjust speed while compressing.
So you can set up your first job to do the 29.97 > 23.976 reverse telecine and chain that to a 23.976 > 25 with 720×486 > 720×576 convert.

But 120 times? You need a preset that does both that you can apply to all. I could make one for you if you could mail me a small, even 2 sec ProRes NTSC 29.97 file.

My address is neilsadwelkar at me dot com.

———————————–
Neil Sadwelkar
neilsadwelkar.blogspot.com
twitter: fcpguru
FCP Editor, Edit systems consultant
Mumbai India

[Michael Pfost]

Here’s the problem… I did use Compressor to do exactly that, but the audio would always be out of sync by .1% in the end (final PAL version).

I tried going to 23.98 and 23.976 — same issue.

I’m starting to wonder if ProRes is causing the issue or perhaps my audio codec (tried Apple Lossless and Linear PCM).

I’ve read countless forums with people having the same issue, but always having to speed up the audio in the end.

It sounds like Compressor needs to apply “Audio Pull-Down” or “Audio Pull-Up.” I’m not sure – it has driven me crazy.

[Neil Sadwelkar]

I created a NTSC ProRes movie 10 sec long. It has a colour bar, some moving coloured squares, a running NTSC TC burn-in, and white flashes accompanied with 1 frame beeps. And a music track as well.

I exported that into an NTSC ProRes movie. Ran it through Compressor with custom settings for the 23.98>25 speedup as well as the Reverse telecine and 720×486 > 720×576 blowup.

The resulting PAL movie is 9s 14fr long (expected) and the audio beeps are in sync with the flashes. A couple of beeps look like they are a fraction of a frame off (from the waveform) but that’s maybe because of the 29.97>23.98 pull down. But overall, video-audio sync is maintained.

I can upload or mail both movies to you to see,

———————————–
Neil Sadwelkar
neilsadwelkar.blogspot.com
twitter: fcpguru
FCP Editor, Edit systems consultant
Mumbai India

[Michael Pfost]

I believe you 🙂 The problem is use this same workflow for a 52 minute program. The audio starts to drift off, eventually being about 4 seconds off in the end. This has been my problem.

[John Knapich]

What is involved in step 1/reverse telecine?
Are you just taking a pro res setting and changing the frame rate?
I ask because I have a similar situation and was wondering if the process effects the sound quality in terms of pitch etc.

FCP 6.06, OS 10.5.8 2x3GHZ Quad-Core Intel Xeon, Kona LHe, Dulce Duo-eSATA 8 Drive, 4TB Raid.

John Knapich
Creative Director/Partner
Assembypix.tv

[Michael Pfost]

It is quite easy to “Reverse Telecine” a 29.97 clip with Compressor.

Just turn Frame Controls “On” and choose “Reverse Telecine” under the Deinterlace option. Make sure the “Set Duration to:” is 100%.

You’ll then have your film at 23.98 fps if it was telecined correctly.

I use ProRes 4:2:2 files for the whole workflow. I’ve also found that it is actually faster to not use multiple clusters when processing. (I takes me about 20 minutes of rendering for a 60 minute file and I’m using an 8-Core 2.26 GHz MacPro).

[Neil Sadwelkar]

Actually if the compressor+ cinema tools process works you can do both in a batch. Batch compress in compressor to handle the reverse telecine. Then batch conform the converted clips in Cinema Tools. CT does more than one movie in one pass.

I’ve yet to find time to do the test I’d done, on a long piece.

———————————–
Neil Sadwelkar
neilsadwelkar.blogspot.com
twitter: fcpguru
FCP Editor, Edit systems consultant
Mumbai India