Help me! Duplicate vocal issue sending me insane!!!

[S79]

Something that's been driving me crazy far, far, far too long now

This is for a vocal house track I'm working on in Cubase 8.5 32bit.

I have a vocal chorus part/wav which I've duplicated "almost" exactly twice so that it repeats, so nothing to complicated:

- Chorus line 1
- Chorus line 2
- Chorus line 1
- Chorus line 2

The bit that I changed is is I've spliced a small section from the end of chorus line 1 to replace/swap the end of chorus line 2, so...

- Chorus line 1
- Chorus line 2 (now spliced with the end (8 beats) from chorus line 1)
- Chorus line 1
- Chorus line 2

I've attached an image that hopefully explains this better? (N.b. The top line is just a vocal double, lower line is the main)

Vocal Issue2.PNG

This is a long complicated way to ask why this small(?) change (in volume/proximity/something?) would cause the two sets of (almost identical) choruses to sound totally different. I.e. the second chorus set is louder/boomier. As it's bass heavy dance track it is very very noticeable. This is despite multiple dynamic eq's, multiple compressors, sidechain eq ducking the vocal against the kicks and snares, matching automation etc.


Now the spliced/copied end part from part one isn't quite as loud as the as the original ending from chorus line 2, so I boosted it with clip gain on the audio, which definitely help even out the two choruses, but the difference is still noticeable. So, my questions...

- Should I just be high passing more to begin with?
- Multi band compression ?
- Is it how it's hitting the compressors?
- I'm not commpressing enough?
- Is it a phase issue?
- Is it a change in proximity effect?
- More sidechain ducking?


I think my core question is... the difference in level/volume is barely discernible, so why is it impacting the sound so much? When they are almost exactly the same? Even if the difference is not small, why is the compression not levelling it out etc?

Also I wonder if it's a(nother) weird cubase thing, i.e. changing the sound/some kind of drift as time advances?

[enroe]

I think my core question is... the difference in level/volume is barely discernible, so why is it impacting the sound so much? When they are almost exactly the same? Even if the difference is not small, why is the compression not levelling it out etc?

Also I wonder if it's a(nother) weird cubase thing, i.e. changing the sound/some kind of drift as time advances?

Mmh, without listening this vocal part I can only speculate. What comes
into mind is: If you playback two nearly identical parts you'll always
have phase-phenomenons. This means that the amplitudes amplify and
then even completely cancel out elsewhere. That sounds pretty sick.
Maybe that is your issue.

[lfm]

Come to think of a mistake I did once.

There was a setting in daw "select automation with clip" or similar.
And cutting and pasting I got automation over to a place I did not expect.

And using default copy/paste is sometimes unclear which options are checked.
There is usually a "paste special" or similar where a dialog has a lot of checkboxes what to paste or not.

I would make an extra track putting the copied section beside the original position and do solo exclusive between those and hear if it became different.

Very small differences in how close vocalist are to mike at different takes can demand some work. I had such a case with a hired vocalist, and had to increase reverb send as she was getting closer to mike and similar. There was more room ambience in another take.

Just a couple of things....

[thecontrolcentre]

Very small differences in how close vocalist are to mike at different takes can demand some work.

Good point. Vocalists are not likely to sing something exactly the same each time ...

[tehlord]

Sounds like a phase issue.

[S79]

Many thanks for all the replies, much appreciated I will try to respond to each point.

Also I should point out my title might have been inaccurate, what I actually meant was 2 x consecutive duplicate parts, not the extra track on the top that I'm using as a vocal double/thickener...however by the sounds of your replies this vocal double track might be a factor!

Mmh, without listening this vocal part I can only speculate. What comes
into mind is: If you playback two nearly identical parts you'll always
have phase-phenomenons. This means that the amplitudes amplify and
then even completely cancel out elsewhere. That sounds pretty sick.
Maybe that is your issue.

Unfortunately can't share the track/clip at the moment as it's an unreleased track.Were you referring to the vocal double in the top half of the arrangement? Or referring to two consecutive identical parts? Are you saying the double/backing vocal track could be an issue?

[S79]

Come to think of a mistake I did once.

There was a setting in daw "select automation with clip" or similar.
And cutting and pasting I got automation over to a place I did not expect.

I've got the setting for keep automation in place but thanks for suggestion.

I would make an extra track putting the copied section beside the original position and do solo exclusive between those and hear if it became different.

Apologies I'm not following, can you expand on this please?

[S79]

Very small differences in how close vocalist are to mike at different takes can demand some work.

Good point. Vocalists are not likely to sing something exactly the same each time ...

Yes, I guess inconsistencies are part and parcel of vocals.

My issue is what am doing wrong , i.e. Am i not eq'ing enough, not high passing enough, not compressing enough etc to level it out the small differences?

[S79]

Sounds like a phase issue.

Again, please can you expand on this.

Should I run the wav sample through Izotopes RX phase correction/alginment tool (for example?)

[S79]

I am thinking remove the extra vocal doubler/thickener track on the top anyway and use a perhaps a different technique to thicken the sound of the vocal (parallel comp?).

Also I tested with vocal double/thickener track removed AND with an exact duplicate, now the difference is barely there, so a big improvement! But sadly i need to chop/replace the end of that vocal bar :/

Vocal Issue3.PNG

[lfm]

I would make an extra track putting the copied section beside the original position and do solo exclusive between those and hear if it became different.

Apologies I'm not following, can you expand on this please?

Copy on same position on timeline, just two different tracks
- and listen if sounding different as a copy exclusive solo between them

If something funny happends in the copy/paste procedure itself.
- then knowing it was not to copy/paste process making it different
- was there a clip effect that did not come over with copy?

But when two takes are a bit different it's heard as you insert after another take part. It's enough with tiny bit closer to mike and will deminish room ambience and increase volume.

Good vocalists also retract from mike going stronger in voice so many parameters.

You then you might need to automate both volume and some reverb send as I mentioned before to make it an entity.

Just some things I would try troubleshooting....