Just been playing around with a freely recorded project, using How to Use Tempo Warp in Cubase | Q&A with Greg Ondo - YouTube as a starting point.
I must start by saying this is a great feature. I just wanted to share my experiences as it has not been entirely plain sailing but after a few false starts I think I’ve got it where I want it.
The project is based on an acoustic guitar that sometimes leaves a weak beat 1 that the hitpoints don’t pick up so it wasn’t immediately obvious what to use as a starting point to analyse the tempo. I first hit on the idea of rendering a hi-hat track using the Tempo Track I’d created manually. While this works, it’s quite unnecessary because there’s an obscure command buried away that does the job for you: Render Audio Click between Locators (in Project > Signature Track).
You use this to Analyse Tempo but be careful - it switched all my tracks to Linear Timebase! I had to switch them back to Musical or it screwed up the next stage spectacularly. Fortunately, I’ve recently got into the Project Logical Editor which makes this an easy job. Here’s my preset but use at your own risk! Toggle Time Base.xml (1.1 KB). It goes in ‘[your Cubase version’s AppData folder]\Presets\Project Logical Editor’.
Something else to be careful of is how Set Definition from Tempo handles edited audio. Greg Ondo’s demo uses nice, neat, single-event audio tracks but my tracks were anything but. On the first attempt the SDT dialog reported some 80 conflicts which it announced it would bounce for me. The results were somewhat untidy, shall we say, and I strongly recommend you do this manually. To find out exactly which tracks need attention select all events on each, run SDT and see if any conflicts are reported. These are the tracks I bounced. I did it before I analysed tempo but I don’t think it matters if you leave it till now.
We are now back on Greg Ondo’s procedure and everything went smoothly from here on in. Incidentally, the PLE is also a good way to select all your audio files (Select all audio.xml (839 Bytes)), which you now need to do. Run SDT and Bob’s your uncle!
The only disappointment is that I’ve had to bounce my edits but overall I’m really chuffed with the results. I can set a fixed tempo or just speed the whole thing up, keeping the variations. Nice job, Steinberg. But don’t ask it to jump through too many hoops. And don’t ask me how to introduce new, un-SDT’ed audio in at this point. I’d love an answer to that if anyone’s got one.
As always, please feel free to comment, come up with a better way, tell me I’ve done it all wrong and should have sorted it all out much earlier in the project (I probably will next time!) but I hope somebody may find this useful and avoid the pitfalls before they get to them.