I’ve done it a bunch of times with my Digitakt. I think it winds up sounding better than when I’ve tried to mix vocals in the DAW.
The trick is to split the acapella into chunks. So maybe 1 or 2 samples for verse 1, another sample for the chorus, and so on. I usually speed samples up by about 5.19 semitones before I record them anyway to save space. And that carried over when I’m working with an acapella. In this case, I would import them with Transfer since I don’t want to normalize all of the samples individually.
From there I put them on a track the same way I would any other sample. If works out really well. I find that it’s easier to make those slight adjustments I might need to make to get everything hitting on the beat just the way I want to. Obviously the rhymes will have been written to a different beat so if you change that up then the rapper’s flow is going to fall a little differently on it even if the tempo is matched perfectly.
Also, a lot of times it’s hard to get the BPM exactly right since a lot of people’s beats are like 98.3 BPM instead of just 98 or something like that.
If you’re going the phone route then I think that using something like Koala sampler would probably be best. If I were trying this then I would connect to the phone over USB. And I would have the headphone out from my phone connected to the line in on the Digitakt. I would set the Digitakt to just send MIDI and no audio over USB. That way I would be able to trigger the vocals and run them through the Digitakt’s FX. I could also use the microtiming to account for any latency.
That doesn’t work for me because Koala no longer works on my phone. But if you’ve got an iphone or an iPad then Koala probably runs fine.
So yeah, I hope that helps. Personally, I’ll be good.