off camera audio sync ... HELP

[David Nardini]

Ok, somebody please help me out or point me in the right direction ...

PLEASE NOTE : I'm no sound expert (clearly) and my 1st play with sound, so keep the advice 'basic' please ;-)

Recorded a long piece (with sound input on the R1) and this sound is in sync when I play back QT wrapper or rendered R3D material (55 minute speech) - lets call this audio-track-A.

Due to logistical problem (ie : could not pipe into the R1) also recorded a wireless mic output into a ZOOM digital recorded (recorded 48k, 16bit) - lets call this audio-track-B.

If I pull in audio-track-B into the FCP timeline (kept at 48k / 16bit), the audio track does not sync up with the visual (eg : may sync at the start but not at the end).

The FCP sequence properties tells me it's 48k ... but for some reason it will not match up.

Is there a utility which will 'match' (re-time) audio-track-B to the reference audio-track-A ???

Any advice welcome ... I'm sure I'm not the first and cannot believe there is no automated way to solve this ;-)
Best practice ? Links ? FCP utility (or can FCP do this !), etc, etc ...

Cheers ...

Note : R1 capture was at 24p / 3k and rendered out to ProRes HQ HD, should this impact things !

 

[Jesse Korosi]

Hey David,
I normally use Sync-N-Link for syncing sound in FCP with Red Footage. It works great, but it runs off of Time Code. So this may not work for you.

Do you know what bitrate/Khz "audio track 1" was recorded at?

 

[Blair S. Paulsen]

Did you check 24.00 vs 23.976 or 30.00 vs 29.97 frame rate mismatch?

Cheers - #19

 

[David Nardini]

Did you check 24.00 vs 23.976 or 30.00 vs 29.97 frame rate mismatch?

Cheers - #19

Hmmm ... I must be missing something ...

In my mind I have the idea that if both devices are recoding at 48K (ie at this rate per second) I had assumed that the FCP project frame rate has no impact and did not think of the ZOOM H2 recoding at an different frame rate. Is this what you are asking ?

Is the suggestion that once the R1 audio is on a FCP 24p sequence I need to adjust the ZOOM audio track by a factor (apple+J) to extend of shrink to make it match ?

Not that I did use a 5Dii but I'm starting to think that the following link contains some info that might help me (as the topic is the recording of sound with an equivalent device) : http://philipbloom.co.uk/2009/05/23/how-to-record-sound-with-the-canon-5dmk2/

EDIT : http://brucesharpe.blogspot.com/2009/06/dslr-dual-system-audio-999-solution.html

 

[David Nardini]

Hey David,

Do you know what bitrate/Khz "audio track 1" was recorded at?

audio-track-A = R1 (RedOne) track, default RedOne setting : 48KHz (24bit).

audio-track-B = The ZOOM H2 was recorded at 48KHz (16bit).

 

[Jesse Korosi]

If you are cutting in a 24P timeline and your audio is 23.98 that could be the issue? Which is what David meant I think

 

[Blair S. Paulsen]

48KHz is the sampling rate and is independent of the frame rate The amount of drift you describe could be attributed to a 24.00 vs 23.976 or 30.00 vs 29.97 frame rate mismatch, regardless of the sampling rate or bit count.

In the US it is almost always advisable to stick with the fractional versions, sometimes referred to as "drop frame" (23.976 and 29.97) rather than the full integer rates.

Cheers - #19

 

[Dan Hudgins]

Crystal

Crystal

Crystal recorders will not hold sync to +/- one frame longer than maybe 4 minutes at best, I have done tests using Pilot tone and you can see them walk off sync, computers will loose sync faster, maybe be out 1 frame in 30 seconds.

You can use the "resample" feature in your audio editing software. If you have pickup of the same audio in both WAV files do this,

1) Load both WAV as tracks in your audio editing software.

2) Adjust the "pull up" of the external audio so the peaks in the waveform match the head.

3) Scroll to the tail and look at the two wave forms, count how many samples the peaks are out of match, then look at the sample count, find the ratio of the error and use that as the resample factor.

4) Resample just the external WAV track to make it longer or shorter with the ratio and then look along the wave form to see that you are within 2000 samples all the way along, if not you will need to cut between words and do the resample in sections.