southpaw3473
I have requested it for a long time as well. For me, in the old analog tape days, I used it to bring acoustic instruments in tune with other tracks. Bagpipes, bombards, fixed tone flutes and many other instruments are very often not in concert pitch. By altering the tape speed you can bring those into tune.

 
There's an inelegant workaround (but it does work) for that problem, as long as the instrument isn't horribly out of tune.
 

1. Record the part from the out of tune instrument, accepting the fact that it will be annoying to hear it sound out of tune as you record.
2. Slip-edit the clip you just recorded to a specific number of measures and bounce to clip (actually you don't have to do this, but it makes life easier later on).
3. Bring the file into your sampler of choice.
4. Adjust the pitch bend until the pitch is correct, then render.
5. The audio will now be the correct pitch, but its duration will be either shorter than or longer than the original number of measures. Ctrl-click the clip edge and drag to the original number of measures in step 2. (I'm always amazed at how few Sonar users know about this extremely simple technique for stretching duration.)
6. Bounce the clip to itself to get out of preview mode and utilize iZotope's high-quality offline rendering algorithm.

 
Done! There are also some situations where the loop construction window can do the job. Maybe I'll describe that later.
 
I'm still not quite understanding what people actually want to accomplish. Tape vari-speed affects all tracks of an entire recording, and the process changes both pitch and duration - you couldn't change one without changing the other. Is this what you want?
 
I used vari-speed for years until I entered the digital domain. As a result, I have a goodie bag of workarounds that let me do 99% of what I used to do with vari-speed (although I don't burn out capstan motors any more). If you can tell me specific things you want to do, I may have a workaround.
 
(The technique described above is similar to something I used to do with an Emulator II and tape to fix out of tune vocals - I'd sample the out of tune section, pitch correct it manually, then punch it in over the old vocal.)

...wicked
Two words Craig: THE. BEATLES.
 
I did a slowed down vocal and just dumped the mix into reaper, sped it up, sang to it, then returned it to normal and bounced it back out and brought it back into SONAR. Sounded amazing.

I do that all the time in Sonar. For a slowed-down vocal:
 
1. Premix your tracks to a stereo mix. Mute all the other tracks.
2. Use Process > Transpose > Audio to transpose the premix up a semitone (or two or whatever).
3. Sing along with the premix to create a vocal clip.
4. When done, use Process > Transpose > Audio to transpose the vocal down a semitone (or whatever you transposed up by in step 2).
5. Bounce the transposed clip to itself to use the high-quality transposition algorithm.
6. Delete the premix and unmute the tracks.
 
Now your vocal is pitched down and it also doesn't have any duration issues.
 
Reverse the transposition process to have vocals that are transposed up and sound brighter. The audio quality will be great thanks to iZotope's pitch transposition algorithm.

CakeAlexS
Sounds like a nice idea. But I want the meat and potatoes features first...!

That used to be part of the meat and potatoes of analog recording.  I agree with what Craig said, though, and I'm sure that the gap between realtime and offline processing of "tape speed" will narrow soon.

I should also add...the reason why I say I do this "all the time" is because I often do my own backing vocals. By recording as described above, with one vocal one brighter and one darker, it sounds like there are two different singers.

Aha! Found an example! Check out my cover version of "Poison" at https://www.youtube.com/watch?v=8Oy_I7JX3xg
 
At about 40 seconds the chorus kicks in with the "transposed" vocals. I only did pitched up, though.

Lynn

CakeAlexS
Sounds like a nice idea. But I want the meat and potatoes features first...!

That used to be part of the meat and potatoes of analog recording.  I agree with what Craig said, though, and I'm sure that the gap between realtime and offline processing of "tape speed" will narrow soon.

Not really. It was razor blades, chalk and tape and a tape machine. Oh and a desk with a patch bay! Wasn't an everyday use feature. Thrash metal bands loved it though.... (Talking 80s 90s here just from personal experience).

Still its a nice feature to have and now it's 2014.

Hi Craig Anderton,
 
To answer your question: Sometimes, I used vari speed purposely to get the "chipmunk effect", not to avoid it. I'd put a couple of background voices in, singing in harmony "answering" the lead voice, for a comic and/or childlike effect. It was quite effective and above all, it was fun !
 
I'd love to have it back. Maybe it's already my second childhood talking or something!
 
LJ
 

skinnybones lampshade
Hi Craig Anderton,
 
To answer your question: Sometimes, I used vari speed purposely to get the "chipmunk effect", not to avoid it. I'd put a couple of background voices in, singing in harmony "answering" the lead voice, for a comic and/or childlike effect. It was quite effective and above all, it was fun !
 
I'd love to have it back. Maybe it's already my second childhood talking or something!
 
LJ

I think this will let you have it back - I'm not at my music computer, but give it a try and let me know if it works
 
Let's suppose you want the chipmunks to be an octave above the original music. It's testing the limits of stretching, but the following should do the job.
 
1. Create a premix of your tracks. Mute the other tracks.
2. Transpose the premix down 12 semitones with Process > Transpose. The premix will now sound like crap.
3. Ctrl-click the end of the premix and drag to 200%. It will now be twice as long and sound like even worse crap, but it should be good enough for you to be able to sing along with it. Record your vocal.
4. Don't transpose the vocal up +12! Instead, bring the vocal clip file into your sampler of choice.
5. Transpose up an octave (this will cut the length in half and transpose up an octave), render, and drag the rendered clip to the right location on the timeline. You can transpose by changing the root key, leaving pitch bend all the way up, holding down a key an octave higher, etc.
 
The fidelity of the pitch-up voice will be pretty much identical to what you got with tape. Happy Chipmunks!
 

Wow, Craig Anderton, thanks a bunch! I'd have never come up with that, especially the part about bringing the vocal clip into a sampler instead of transposing it up an octave just like that.
 
I do have one question: Is there a certain sampler that you have in mind that this would go especially simply/quickly with?
 
I can't wait to try it. Thanks again  :)
LJ

Anderton
 
 
I'm still not quite understanding what people actually want to accomplish. Tape vari-speed affects all tracks of an entire recording, and the process changes both pitch and duration - you couldn't change one without changing the other. Is this what you want?

Hi Craig,
Thanks for the idea for the workaround. I truly believe if someone needed to know how to make Sonar brew coffee you would find a way to do it (which would tie in with the phantom Sonar coffee cup).
 
The way we used to use tape speed was to record the backing tracks first, making sure they were in pitch. Then, if the offending instrument was sharp, increase the tape speed slightly during tracking so the original tracks were in tune. Record and return the speed to normal. It was very effective and a common technique.
 
Again, a way to do this would be to change the sample rate by degrees I would think. Am I wrong about that?