I have to do some ghost singing for an artist (female).
 
Last time I had to do this was years ago. Back then I just slowed down the multitrack until it was in my register, sang whilst emulating the artist's slowed-down voice, then put the speed back to normal.
 
If all I have are recorded audio tracks (no split clips yet, no midi VSTi, just straight 7-track identical-length audio from start to finish), is there an easy way to do this in Sonar?

1. Bounce the tracks to a premix that you can sing against.
2. Refer to the tip in Week 20 of "Friday's Tip of the Week" about "SONAR's Fine Tuning Pitch Transposition Window."
3. Change the premix clip speed as desired, e.g., -40 cents. Sing along with the premix. When you're done, follow the same tip to speed up your vocal by +40 cents.

 
This is true "tape varispeed" where the time shortens when you transpose up and lengthens when you transpose down. If you want to preserve the clip duration, instead of using the pitch transposition window select the clip and choose Process > Transpose. Note this works only in semitone increments. So for example you'd transpose the premix down 1 semitone, sing along with it, then transpose your vocal up 1 semitone.

I concur, but I do it by creating a sample of the entire bounce, importing into KONTAKT and play one note to trigger the mix and use the pitch wheel value to set the desired slow down or speed up then bounce this new speed audio and do the same in reverse to import back into original project.  This method gives you more control of the speed.  It is easy to create the sound of the tape constantly getting faster (as on the YES ALBUM at the end of "Yours is No Disgrace") or suddenly getting slower and coming to what appears a complete stop as heard on UltraVox "All Stood Still" (you can set the pitch of the wheel in KONTAKT) to have wild swings in tempo to do these larger swings (up to 2 octaves plus or minus programmed into the pitch wheel movement). In addition to singing, you can use this method to play an instrument to the vari sped audio mix (this was done in a few sections of Genesis' Lamb tracks (the piano intro to Carpet Crawlers was recorded at half speed 7.5 ips then just played by at 15 ips.  All this did was raise the piano part an octave and made it sound twice as fast as it was performed but also sped up everything about each note (attack, sustain and release of each piano note was double speed when mixed giving a unique sound not possible on the original RMI piano naturally), reverse was done with the low voices in the songs "The Grand Parade" and "In the Cage".
 
Either way, this isn't as easy as the old vari speed tape deck method. It is a lot of administrative work. And I wish that there was a  simple speed knob in SONAR to do this intuitively.

I don't think the OP is interested in tape stop effects, but in fixed xpose up/xpose down. But note that you don't have to use Kontakt for "tape stop" effects. Rapture or Rapture Pro can do +/- 48 semitone transpositions, which is enough to do a decent job.

@Craig: A bit fiddly, finding the right pitch, but one gets the feel for it after a few tries. I am working again.
Sir, I don't know where you find the time for all stuff you do, but as always, thank you very much.
 
@jimkleban: Good idea. Following your lead I tried with Cakewalk's own Dropzone and it does work.
It would need additional administration putting the chosen slowdown back into a track to sing to, but it's a technique worth keeping in mind.
 
In the old days we'd put the singer's track phrase-for-phrase onto notes in the Akai sampler.
The vocal track was then re-recorded to the backing track whilst triggering the vocalist's phrases on the keyboard and correcting pitch with the pitch wheel. In this way the song was re-arranged and phrasing and pitch was corrected at the same time. A mistake (and there were many) meant stopping and restarting from a drop-in point.
 
...and that's something I never, ever want to have to do again. The width of dead zone on the DX7 pitch wheel is burnt into my amygdala...
 
I once tried setting up the triggering of the Akai on the Roland MC500, but it was too much additional work and I never bothered again.

I know this question, regarding a simple Tape Varispeed function often comes up on this forum.  Do some or most other DAWS have this?  It must be a difficult to implement, otherwise I can't see why Cakewalk would not consider adding it.  I can certainly see where it would be a valuable addition. 

Is the ONLY reason you want to do this to bring down the pitch to your register temproarily while you track?
 
If you do not need the track to slow down as well (which to me would make things less accurate and possibly introduce more artifacts than what I am about to suggest)...
 
Just do what Craig said kind of sort of BUT without the timing stuff.
 
So...
 
1) Do a Bounce to Track (to a new track) of all the necessary tracks in the project for you to sing the part accurately. The less tracks included the better the quality will be when you transpose down.
 
2) Solo the newly bounced track/mixdown. This is your backer track you will sing over top of. BTW before you bounce make a mix that will benefit you most while recording your vox... kind of like a headphone mix.
 
3) (Optional) Make a copy of the clip in the bounced track and put it in a new lane in the track. This is so you retain a copy of the original bounce in case you need to try a different transpose method. Mute the original clip so you can only hear the copy in the bounced track.
 
4) There are multiple ways to transpose a clip in Sonar these days. There is the DSP Transpose function (in the Process Menu), there is Melodyne, there is V-Vocal (if you have it installed) and all the synths/samplers. Try the Process > Transpose option first. It's probably the easiest and works well.
 
5) Transpose down by the amount required. Use an exact transpose value in semitones. Not Cents and make a note of how many semeitones you went down by (so you can transpose up easily later).
 
6) Create, solo (leaving the bounced track soloed as well as your backer track) and arm a new track. Record the track. There is no timing changes so the part does not have to be sung slower or anything. Real time speed (which is the whole point of this post as opposed to the other time warp methods)
 
7) Once the part is recorded and edited you can use any of the pitch tools mentioned earlier to transpose your vocal track up to the project pitch. Unsolo and mute the Bounced "backer track" and unsolo the new vocal track for the transposition process.
 
8) For the transposition I would say Melodyne is probably the best option because the quality is good and you can also correct any off pitch notes in the process. If you use Melodyne be sure to use Bounce to Clips after the pitch change. If it sounds weird try going into the Inspector (press I), open the "Clips" tab, Click the "Audiosnap" sub menu and with the clip selected change the "Offline Render" to something else (like if it was Radius mix "Bass" switch to Radius Mix "Vocal" or whatever works best). If you use DSP Transpose you don't have to worry about it because it bounces itself.
 
tl;dr
 
Bounce a mix of the project.
 
Transpose that clip down using Process > Transpose or Melodyne by specific interval
 
Record vox to downtuned mix track
 
Transpose vocal clip up to original project material and mute/delete the downtuned back track.
 
 
If you want REAL varispeed (where everything slows down too)... use Reaper. It's got a Varispeed fader that's ridiculously simple to use.
 
Cheers.

BTW... Guitar Rig has transpose and speed options in it's "Pre" Tape Deck tool. Anyone who owned X1 (which came with GR4) has this accessible.
 
I wouldn't use it for this (not even sure how to set up the project for it) but it does varispeed type stuff. You import a wave into it and fiddle with the dials.
 
Not great quality. Good for learning/writing wicked fast solos though.

@Beepster: Thanks for that. Melodyne does, however, collect artifacts if shifting more than a few tones. Maybe I'm doing it wrong...?
 
Slowing down is preferable because articulating the details of someone else's voice is not easy. Or maybe I'm just used to doing it that way.
 
I may even pull my Tascam 244 from the cellar. I recently was able to buy a full set of rubber wheels and bands for it from Tascam (after all these years!) so it's in A-1 condition and I still have a box of 10 of the most superb TDK AR-X cassette tapes for which I optimized the bias settings.

jpetersen
@Craig: A bit fiddly, finding the right pitch, but one gets the feel for it after a few tries. I am working again.
Sir, I don't know where you find the time for all stuff you do, but as always, thank you very much.

One major benefit I forgot to mention - zero artifacts, because you're not trying to preserve duration.