Happens to me fairly often. The main reason I bought C6 is to get the new and improved timestretching facility. The older one, in C5, worked okay but there were often artifacts, like “warbly” pianos and such. The new version seems to work perfectly.
Yep! That’s what I did with my - soon to be released 
- new song.
I’d been struggling with it for quite a while and finally found out that the groove and the tempo weren’t right…
I deleted everything except the midi track of the Rhodes, upped the tempo a few beats and started from there with a new groove.
Quite a job, but I’m much happier now and it’s nearly finished. 
This was a first for me though, usualy I’ve got a good idea of where I want to go with a song and play it through on my piano till I’m happy with the tempo and groove.
If this helps at all, you can check out a little test I did to verify what I’d already seen on youtube for time stretching in C6… So here is a little composition rattled off in about 15 minutes.
I created the tune, exported my stereo mix (some HalionSonic SE, GrooveAgentOne, and my less than great guitar playing), re-imported into a new project then messed with the tempo. It works very well. Hats off to Steinberg on this one!
It seems to be mainly the hihat’s going wonky with the timestretch, everything else sounds fine.
and hihats are easily replaced in midi anyway 
I’m old school that way myself. I’ve re-tracked a couple of songs that way instead of time-stretching it. I try to limit time stretching to smaller events.
If it’s wrong, it’s wrong.
Start from scratch, or you’ll never be satisfied.
I haven’t done too much with audio, but I just spent all weekend learning how to do this with MIDI. That warp tool is just awesome - I put a tempo marker at every bar, then went to the tempo track, and adjusted to flavor. Friggin’ awesome!
Often find that when I set what I think is the tempo and record it turns out to be wrong.
What I have found useful, is to jam the song for an evening on just the piano (or I guess acoustic guitar if you are an axeman)
I have a songwriting buddy who has the most gorgeous Bechstein grand, and when we write at his place I often just play that (while he sits on the sofa and shouts out ideas - LOL) and capture the whole lot on a little olympus recorder or my iPhone, and then when I actually start to work on a recording I’ll create the base tempo from the sketch. Usually much happier with the end result that way!
I have to adjust the tempo of songs every now and then (audio) I always found that the best result is to finish and mix song then stretch stereo file to new tempo using different time stretch software. (when a rerecord is impossible)
But maybe the new C6 time stretch is better overall on multi track stuff, haven’t tried it yet.
Another question or two Steve, how big a tempo change are you talking about? And is there more than one in the tune? Which version of Cubase you running?
I’m always afraid to start over because it never seems to be as good the second time
Wow, so true. And also, starting over seems to be associated with a time-sucking black hole … come crawling out of the bedroom studio, and everyone else is years older …
Dude, I can’t tell you how many times I’ve been through that exact same thing. Re-playing old parts… I first think… “the vibe that I once had just isn’t there!”, but when coming back to listen to it later, I see it IS there, and my playing is much better as well. it’s really hard to listen and judge objectively when being so attached to the “old” way.
… and most often it probably isn’t.
Ruining something becuase of a temporary low feeling of tempo is dangerous.
If the tempo always seems slow, then time stretch the mix.
In the old days we used to varispeed up the tape while mastering.
Thinking logically you could at least just export and then timestretch to see if another tempo is better. If this gives artifacts doesn’t really matter, you should be able to tell if another tempo is better anyway. Judging on how well the new tempo sounds you can decide on rerecording or trying to make the best of it with timestretch.
That’s what I’d do anyway, nothing worse than starting over and finding out it didn’t make it any better.
What’s weird is that I’ve gotten okay using Timestretch on a mix, but when I try to do it on individual audio parts the results aren’t usable. This is the main reason I bought C6, and one of these days I’m gonna see how it works on this
Different thought from me entirely. I’m just an acoustic guitar and vox guy. Nuffink else. I don’t do “tempos”. I just set up the mics and play. That’s not to say I don’t give a moment’s thought to tempo - of course I do. But it’s a moment and the recording get’s done.
These songs I record in this slovenly way are songs I play live and I’ve been aware for many years that I play them faster or slower “depending”… The songs don’t have a correct tempo. Recordings of them are snapshots. There may be median but, frankly, I don’t care.
Before the advent of recording, all composers could do was festoon their scores with adagio to presto markings, or if they were obsessed, then metronome numbers. Interpretation and re-interpretation was everything. We are in danger of setting music in concrete.
On the contrary, we’re talking about the freedom to make an adjustment IF (and only if) it will improve song, and it’s normally just a slight adjustment anyway
I reckon some of the best ‘feels’ I’ve achieved have occured in a few tunes that were actually recorded with no click, and therefore no fixed tempo. The songs tempo just followed it’s own natural ebb and flow. However I realise recording to a click makes life much easier when it comes to editing etc and when you’re working with both audio and MIDI/VSTi content.
I’m generally in the ‘start over’ camp if I later decide the tempo not right. And this does of course mean re-recording all the audio content. MIDI of course is not an issue. Personally I’ve never found this to be a particular problem:
Subsequent takes I can usually get sounding just as good if not better - it;s just about being consistent in your recording approach, and practiced in your playing. And if it does sound worse…do it again! …and again! …
I wonder too if you start out at a less-than-optimal tempo there’s a risk that you then end up playing each part and not getting the quite right feel… maybe.
I guess it depends on how much work you’ve already done too. Particularly, if it’s mostly audio content then the prospect of starting over may well be quite daunting. This would certainly be the case if live drums are involved!
I often have around 50% MIDI content in a pop/rock production so replaying a few guitars parts and re-singing isn’t THAT big a deal. In fact I did that recently with a clients song. He recorded his original guitars and vocals to a click at 142bpm. W A Y - T O O - S L O W. I sent him back to the mic with a bass and drum track (MIDI/VSTi) I’d created for the song at 150bpm - quite a big jump in tempo… but thankfully he completely agreed that it was much better at that tempo and dutifully re-did his guitars and vocals. 
Time-stretching would generally be a last resort I think…
I still cannot fathom how you could possibly get that far into a production and not realize that the tempo is FUBAR.
That is more than a temporary lapse of judgment.