-
Double system sound Sync issue
Dear Cow Experts!
I have a 71 minute interview shot on an HPX-170 at 23.98 fps 1080PA. I shot with a camera mounted Røde recording to the camera for reference and I used a Tascam DR-40 for double-system audio. It was recorded at 48 kHz. Synced using audio/visual clapper. wav files were converted to 24bit AIFFs in Compressor prior to importing to FCP. (Im using 7.0.3 on 2007 MBP w/4GB ram under OS X 10.7.5)
When I sync the audio to the picture using the clapper, the audio is in sync well enough for 10-20 minutes. Around then is when the slippage becomes noticeable. By around 58 minutes its about 8 frames off. The audio lags behind the video.
I wonder if anyone might have a diagnostic suggestion. I suspected it was related to pullup/down and just by experimenting I found that speeding the audio to 100.01% helped a lot. That makes the offset less than a frame.
That’s not what I expected… I thought that the video, at 23.98 fps, was slower than real-time, implying that the realtime audio tracks should be slowed down to match. However, slowing the audio to 99.99% made it worse.
I have two questions –
1) Is there some pre-processing workflow I can apply to the double-system audio prior to import that will resolve the audio to the 23.98 video? What program(s) should I use?
2) Can anyone help me understand what’s going on here?Thanks
Lars -
16 Replies
-
Hi Lars,
Duplicate the sequence and try to sync the original Wav files.
rafael -
Didn’t work.
I tried two things: creating a new sequence, dragging the video into it, and clicked ‘yes’ for FCP to match sequence settings to clip settings. Then I dragged the .wav file into the timeline with ‘snap’ on, which allowed me to align the markers i had placed at the clap. Same result as with aiff – starts okay, but by the end off by many frames.
2nd thing was to create a sequence from scratch and add the .wav as above, but with similar lack of sucess.
What else can I try?
-
I also created a prores sequence at 24 (not 23.98) and added the audio and video to that. Audio still lags behind video.
-
FC has a very well known bug that puts audio out of synk on certain circumstances.
Have a look to this:
https://library.creativecow.net/articles/lyon_matt/
rafael -
I read Matt’s article/tutorial. Here’s what I’ve tried:
1) Set default sequence preset to 1920×1080 24p. Quit and relaunched FCP. Imported my video footage (DVCProHD 1080p30 23.98 fps, shot with HPX-170 at 1080PA24). Then I imported the AIFF (converted from .wav using Compressor). Synced to clapper, but same problem.
2) Set default sequence to DVCProHD 1080p24. Quit and relaunched FCP. Imported video footage, wav file. Synced to clapper, but same problem.
3) repeated above with default set to DVCProHD 1080p30. Quit & relaunched. Imported video and audio. WHen I put the video in the sequence, I did two ways, allowing FCP to match the sequence settings (making it 23.98) and not (forcing the 23.98 footage into a 29.97 sequence). Same problem, its out of sync noticeably an hour later. (It gets worse as time goes on but its perceptible around 20 minutes in).
4) I reset the default sequence preset back to Apple ProRes 1920×1080 24p, quit and relaunched, imported audio and video, forced the DVCProHD clip into the prores sequence (didn’t allow fcp to match settings), then synchronized audio at the clapper. Then I exported an XML of the sequence and opened it in TextEdit. The video clip and the audio clip are both NTSC: TRUE and TIMEBASE 24.
So now I’m really up it, since Matt’s fix seems not work. Have I missed something?
-
I have also tried to change the way the video clip is imported. So I opened Log&Transfer, and under preferences set P2-Plugin/P2 AVC-Intra to ProRes 422 (see grab:)
Unfortunately, it doesn’t seem to make a difference. When I reimport the clip, its compressor is still DVCProHD 1080p30, frame rate 23.98. In the Log&Transfer window, I see that the source format and target format are the same regardless of the settings in the preference pane:
So now I’m really stumped.
-
Most of the time when this occurs, it’s an issue with the original digitizing. Slow drives, full drives, or using the system while digitizing. Unfortunately once it’s captured wrong there’s no way to get the file to sync. You have to recapture . No redigitizing re-encoding will fix it.
Good luck
ToddTodd at UCSB
Television Production -
If I change the speed of the audio clip to %100.01, it stays in sync the whole way through. That suggests that it is a problem of 24 vs 23.98. Would mis-digitizing account for that?
[Todd Gillespie] “Slow drives, full drives, or using the system while digitizing.”
I’m using a CalDigit VR external RAID 0 over FW800 importing from a G-Tech mini 7200 RPM. There’s over a terabyte of free space on the CalDigit. I”m using a MBP 2.4 GHz Intel Core 2 Duo with 4 GB of RAM. I’m confident (but not 100% sure) that I left the system unattended during Log&Transfer). I can of course try it again, and I’l repost afterwards.
Interestingly, I installed a trial version of Premiere CS6, which will open the mxf files directly. Same problem!
-
As another test, I imported the audio into Audacity. There I sped it up by 0.01%. I also tried to change the sample rate to 48048Hz. I imported both back into FCP, and the first one worked very well, the second not. It was 4 seconds shorter than first (both were about 1hr 10min long). It didn’t matter if the sequence was prores 1080p24 or DVCProHD 1080p24
go figure.
-
So I downloaded SoundDevices WaveAgent app, and tried every combination of frame rate and sample rate and no luck. THE ONLY thing that works (so far) is to change the speed inside FCP to 100.01%.
Reply to this Discussion! Login or Sign Up