Acapellas4U.co.uk

Hi Folks,

So, I've lined up the MIDI timeline to match exactly the original track. There are a few small, but important variations. Now, what I want to do is take that audio, and stretch it so that it lines up to the MIDI beat, but with the MIDI beat exactly steady. What is the simplest way to do this?

Thanks in advance!

-Joe

You really should try and give more info with questions like this, but, as usual, i'll just try and guess what you're talking about...

Are you trying to match an acapella with Sonar, so it matches your midi beat? Ok, i'll assume that's what you mean, and if i'm wrong, then you've just wasted 20 minutes of my life lol.

First off, if i understand you right, you've already matched Sonar's tempo with the acapella as best you can, so that's half the battle won.

Now there is a couple of routes you can take. If you would like to change the tempo of the track, and keep the vocals in time, then the easiest way, in my opinion, is to slice the vocal track into 1-bar sections (make sure snap to grid is enabled, click the vocals at the start of each bar, then push 'S'. That will slice on the beat).
Once you have sliced at every bar (or every 4 bars if you're feeling lazy), you then select all the slices together and push 'CTRL+L'. This will enable 'Clip Looping' on all selected vocal slices, and you are then free to change the tempo to whatever you want, and the vocal slices will remain in time.
If you're hoping to make a really drastic tempo change, then i would advise double clicking each vocal clip, and changing the slicing resolution from 8 to 16 or 32, otherwise the vocals might not sound right. Changing the resolution is the longest and most annoying part in this process, but if you spend 15 minutes doing that, you'll know that the vocal track will be ready to match any song at any tempo you like!

Another way to sync to tempo is actually a whole lot simpler but a lot less flexible. You can just hold CTRL, click the right-hand edge of the unsliced vocal track, and drag it left or right to match the tempo you're track is at. The main problem with this method, is that you need to bounce the clip to a new clip (resample), otherwise you are hearing Sonar's online stretch which doesn't sound good or even accurate. When you bounce to clip, Sonar uses the offline stretch, which is much more accurate. Though, be aware that there may be a few timing errors if you are stretching an unsliced 4 minute acapella. Also, this method is destructive, so you need to have a tempo in mind before you stretch. So there's a couple of reasons why i prefer the first method.

Ok, well all that was really only relevant if you were hoping to change the tempo, so i threw it out there just in case.
To your issue then...
So you have your vocal track matched to the tempo that you want, but there are a few places where the vocal seems a little out of time, if i've got this right. Well the best way to deal with this is to use Audiosnap. Normally i would have followed method 1 and already sliced into bars, so i could just turn off looping on the problem clips and enable audiosnap, but if you haven't sliced anything yet, then you should do that first. So if the out-of-time vocals are in bar 47, then slice the track at bar 45 and bar 49, so you have a 4-bar section. Now right-click the problem sample and select 'Audiosnap Enable'.
Now you will see lots of orange bars on the sample, most of them landing on transients. These bars are what you need to move if you want to sync out-of-time vocals. There is a lot you can do with these little bars, including quantizing the audio to match the song tempo, but i'm not entirely sure what you're problem is, and i'm not gonna write a 2-page essay here, describing what can be done, so i'll just say this:
Clicking the round section at the top of an audiosnap marker lets you move the marker to a new location without stretching any audio.
Clicking the middle (stem) of the marker allows you to move the marker and drag the surrounding audio with it.
So basically, if a word is starting a little bit before the beat, when it should actually be on the beat, make sure that word has a marker over it, and drag the marker to where the word should be. This will only stretch the audio from where the previous marker starts, so it doesn't stretch the whole sample. That's the beauty of audiosnap.

Hopefully that's enough to fix your problem!

Thanks for the knowledgeable and helpful reply! Audiosnap has proven less a snap than I'd hoped, but, I think with your new info, I can make it work.

to use bpm identificator

Some times I use the split tool to move the vocal where I need it if it is a bit of, takes a lot of time but works well