There has to be a solution to matching old vocals to a new version of the same song, I just don’t know how.
In 1992 a former band of mine recorded one of my songs after I left the band. The vocal performance is perfect but the rest of the recording is ehhh…and the bass player totally ruined what I wrote. It was recorded tight but without a click track. Lets say the tempo is 125 BPM (as far as I can tell).
The singer passed away in 2022.
In 2023, I re recorded my song the way it SHOULD have sounded and cut it to click. None of the singers I have brought in have that special spark that was done originally. My new version is at 118 bpm.
I have been able to successfully pull the OG lead vocals out through a stem separator.
So, how would you go about mapping the OG vocal to the newer slower version? Remember, the OG version was cut without click.
What would be the step by step instructions, down to the painful detail, you would do to match the old vocal to the new version?
A HUGE thanks in advance for anyone willing to help me though this technical problem solving. There has to be a way to do it, I just don’t know how.
Would I pull all the original stems out first, map the tempo and import that audio file with tempo embedded into new version?
I will also have to contend with the fact that the stems are at 44.1 and the new version is at 48khz.
Please make your reply as nerdy and break it down step by step. No detail is ‘too detailed’. My brain thrives on the most specific of steps.
If the old mix is 44.1k then sample rate convert it to 48k first. I assume this is why you’re saying stems are 44.1k.
Use timewarp tool on the old version to map it’s tempo, set definition from tempo to write that tempo to file then you can delete the tempo map and set to 118.
Then do the stem separation and drop the vocal into the new one.
The “detail” that matters is the accuracy of your tempo map which is why I suggest doing it manually with warp tool rather than using the tempo detection functions.
If you need more detail look for tutorials on making old live audio play to fixed tempo, that’s the main part of what you’re doing here, after that dropping in the stem it should just fit.
Okay, wait. Do i render the stereo file at 118 and then go to my external (online) stem separator? Or are you suggesting i do this a different way. If the latter, kindly give me a few detailed steps.
I’ve only ever used extracted stuff within the same mix to rebalance old tracks a little so not spent much time looking for best quality as any bleed fits in but I used Ultimate Vocal Remover and just searched online for some suggested settings to try and it worked out fine for my use.
TBH I think most will get you a reasonably workable vocal…it’s presumably the thing they’ve been trained on the most….I haven’t tried the Cubase built in yet but hear it’s not as good as Spectralyers Pro which seems to be highly rated
