Re-recording Interview Audio

S

sabbott

Member
Hi there,

I expect a straight up 'no' in answer to my question, but just in case there's hope.....

I recently recorded video of someone saying a few lines to camera. I recorded the audio badly, but the pictures are very nice.

Bad weather is preventing a proper reshoot. So, is there any way I could just re-record the audio (properly this time!) and add it to my video in the edit? Or would syncing problems be unavoidable? It's only two short sentences I'd need to match.

Thanks for any advice!
 
As a general statement, yes, you can do this quite easily. The problems come in with the ability of the person to duplicate the lines. Some people have a lot of difficulty doing this.

I'm assuming you just have video editing software, so it is a bit more difficult, but it can be done.

Have them listen to each line or part of line several times. Then record them saying the line. If you can, play the line on one system and have then recorded on another. That is, have them say the line along with the playback. You'll find that they may have problems doing this. So break it down into smaller pieces if necessary - a half sentence or even just a few words.
You may have to splice bits of each recording in order to get the whole line in sync.
 
Last edited:
There are rather sophisticated software tools that will move syllables and phonemes around to match the picture.
But if you had one of those tools, you wouldn't be coming here asking that question.
So I'm afraid the answer is as you suspected.

There is also the issue of whether the subject would tolerate attempting to match the video with a dozen "takes"????
 
Damned SORCERY!

I did that kind of syc'ing MANUALLY for more years than I can count!

Grrrrrr.......grumble, grumble, zark, fark, a bargle......
 
Pretty impresive. Since it's just a short bit and it's a demo so who knows how well it works in everyday use... But still VERY impressive.
 
That is a big time saver. Granted, it was a professional VO talent saying the same line. But, even with actors that can't quite repeat themselves, as the guy said, this would get you far along in a very short period of time. Very cool.

So does ProTools have this now?
 
The Audition tool, although wonderful, would seem to rely on the original audio being 'clean' enough to match up with the replacement audio. So it may not work well if that original audio is horribly noisy?
 
that has traditionally been one of the problems with automated systems, you don't really want the end audio to sound like the starting audio (or you wouldn't be replacing it). And of course it only works if you keep your actors lines separated.
 
Having done numerous ADR sessions for film and TV over the past 36 years you will be amazed at how people can actually just copy what was originally said.

I usually just play people the line in small sections and get them to try and repeat it once they have heard it, several takes and you should get something that is pretty close.

Big advantage these days is that once you have the lines you can get the computer to match it further but if it is close enough anyway you may not even need this.

All this ADR software is great in theory but I generally find that it takes far too long and the talent can get very tired (impatient) waiting for it to do it's stuff.
 
Well you don't use the aligning software while recording, it's an editorial tool. I agree with how many times people do a decent job. I stopped using picture for ADR recording. Actors get obsessed with how they look and it just became really distracting. So I do the play the line say the line game and usually get a solid take in a try or two. Sometimes the ADR hardly needs editing. I also found that it usually goes smoother with someone with some music background. They match rhythm and pace usually very well.
 
You must log in or register to reply here.
Back
Top