Sound doesn't seem to sync with waveform

andrew00

Well-known member
Hey,

So this is strange to me.

I'm editing some audio of this band, one source is a room mic and the other a desk feed. The room mic is 48khz wav and the desk feed 192kps mp3.

When I sync the two audio sources up by the waveforms they sound out of sync. However when I place the mp3 track one frame before the room mic track, this is all in FCP btw, they sound in sync.

This creates problems because I'm editing with one mp3 source, which itself is prone to drift/lost frames.

So with other songs from other bands using the same combo (from the same night) this was a relatively simple, if boring, process of watching the waveforms across the songs and taking out a frame when the mp3 source jumped, moving it forwards one frame and we're back in sync.

However with the above it's complicating and confusing me!
 
Hey,

So this is strange to me.

I'm editing some audio of this band, one source is a room mic and the other a desk feed. The room mic is 48khz wav and the desk feed 192kps mp3.

When I sync the two audio sources up by the waveforms they sound out of sync. However when I place the mp3 track one frame before the room mic track, this is all in FCP btw, they sound in sync.

This creates problems because I'm editing with one mp3 source, which itself is prone to drift/lost frames.

So with other songs from other bands using the same combo (from the same night) this was a relatively simple, if boring, process of watching the waveforms across the songs and taking out a frame when the mp3 source jumped, moving it forwards one frame and we're back in sync.

However with the above it's complicating and confusing me!

Is there some reason for using mp3 at any point until the final output???

Mp3 is a lossy compression technique, and as such various methods are used to compress the data... and so, it would be recommended to avoid any such compression when you are processing the sound in any way. You may have been just lucky before, in that the material was "compressed" with less artifacts, at least visible to you in this type of processing...

Also, as a note, since the typical quanitization of editing in a NLE is on the 'frame' level, editing audio in such an environment will prove problematic. Editing audio in a application that has editable resolution down to the level of the digitized audio is better...
 
Aye it wasn't my choice - I just got what I got given as it were. It actually sounds pretty good quality wise, but yes the errors are rubbish.

So am I right in what you're saying - that essentially because FCP is a frame based editor, the wave forms don't necessarily have to line up, for things to be nicely in sync I should, in future, edit the music track in an audio editor - logic, cubase etc, then reimport it into FCP as a finished thing?
 
No what he is saying is that because FCP is frame based your never going to get the sound in sync.
BUT that is not totally correct. You can edit at sub-frame level in FCP but it's a PITA. And with some luck you might get things in sync. Part of the problem your having is that your not working with apples and apples. The room mic doesn't have the same sound as the direct so aligning them by waveform is wanky at best. The reason it sounds better ahead of the room mic is that the room mic will have lots of "room" on it and that (reverb) naturally comes after the direct sound. Probably your a 1/2 frame off or so and what your noticing is that rev BEFORE the direct sounds wrong.

So first dump the MP3 and slap whoever recorded to MP3. Transcode that to a wav file so it is at least playing at a predictable speed. Then sync it by ear. If you switch to STP you will have an easier time editing sound and easily be able to move at sub-frame increments. If I were doing post sound I would tell the picture editor to drop them in where it sound OK and not worry about it too much. Once the edit is done I would have them give me an OMF with 4 sec handles. That should be enough to allow me to really sync the two files.
 
Back
Top