Melodyne Video

[TheRealRoach]

The singers voice I'm working on is pretty raspy and I'm experiencing on average 2 notes that need to be re-defined per phrase. The video recommends checking the "rough" box for singers like this and then dragging the note up or down into it's proper place. Is there a way I could re-detect the whole performance for a rough singer to begin with?

Possibly. You'll have to check the manual on this one though. I've never really had problems with Mel's detections and thus no need to redefine the detection parameters. Sorry.

As for the bass, I've been selecting the whole performance and using the automatic pitch correct function (you know the one from the edit menu with a slider from 0-100%). I guess I'll have to give the manual pitch correcting a try to see if it changes the transients timing.

This would be the automatic/auto-tuning method i mentioned.

But just to be clear, is manual pitch correcting when you click the pitch correct button on the toolbar (which indicates the proper pitch of the note in blue) and then double clicking on that note to move it to it's proper place OR is it when you select the note, hold ALT, and drag the note into place?

Both of these scenarios would be what I call "manual." Although when you double-click you are note-snapping (i.e. melodyne puts the blob where it THINKS it will be best suited), and when you hold alt you are breaking the note-snap grid in order to adjust the pitch in fine incriments.

A note on blobs: they're f-in useless. They only represent an average of the pitch within the blob. You should always have the option turned on which shows the squiggly line that represent the actual pitch. If I had my way, I would have an option to disable the blobs from appearing all together.

In the end you have to use your ears every bit of the way. I usually correct one bar at a time and have that bar cycling on a loop-mode. It usually takes about an hour for one vocal track for the typical pop song.

[JB872]

Originally Posted by TheRealRoach

Possibly. You'll have to check the manual on this one though. I've never really had problems with Mel's detections and thus no need to redefine the detection parameters. Sorry.



This would be the automatic/auto-tuning method i mentioned.



Both of these scenarios would be what I call "manual." Although when you double-click you are note-snapping (i.e. melodyne puts the blob where it THINKS it will be best suited), and when you hold alt you are breaking the note-snap grid in order to adjust the pitch in fine incriments.

A note on blobs: they're f-in useless. They only represent an average of the pitch within the blob. You should always have the option turned on which shows the squiggly line that represent the actual pitch. If I had my way, I would have an option to disable the blobs from appearing all together.

In the end you have to use your ears every bit of the way. I usually correct one bar at a time and have that bar cycling on a loop-mode. It usually takes about an hour for one vocal track for the typical pop song.

I gave up on using Melodyne with this particular vocal (to much guess work involved). I'll definitely try what you mentioned for the bass next time. thx for your help.

[TheRealRoach]

I gave up on using Melodyne with this particular vocal (to much guess work involved). I'll definitely try what you mentioned for the bass next time. thx for your help.

Cool. Any chance at actually getting the vocalist back in to take a stab at the take again? Taking time to work through the part with the vocalist really pays off, both for the quality of recording, and the skill of the vocalist.

[jonnyb8390]

I'm having difficulty using the Spot to Pro Tools function. I'm a manual guy, but I have to admit - the melodyne manual only got a brief thumbing through. I've been very effective in using the programs pitch correction features, but every time I try a new method of importing/exporting my files in hopes of the "spot" function working, I still have no luck. (I know, I'll most likely have to actually open up the manual again, and it's probably something really simple too!)
I've actually reverted to the classic style of using Melodyne with Pro Tools - consolidating a selection and creating a new audio file within Pro Tools that begins at a fixed point, then importing back to that same point when finished with Melodyne. Something strange happened though the last time I tried it - when I looked at points to confirm accurate location (lip smacks, etc...), they were off by 123 samples (@96kHz), and 345 samples in another instance. I realize that at that sample rate I'm really splitting hairs, but I'm wondering if maybe you could shed some light on any of these issues.

Thanks Much

Jon