Boffins needed - technical challenge for impossible vocal parts

P

pure.fusion

New member
Hi all,

Good at engineering? I need your help.

In short, I want to cover a song in it's original key. Fundamental issue, I will not be able to sing in all the ranges need, both high and low.

In the past I have exported the backing, changed the sample rate, sung the part, then returned the samaple rates to acheive the goal but as you probably know this sounds quite fake for many reasons. I might add that changing to the next available sample rate in Coolpro is a massive jump in pitch.

What I need is to somehow drop the pitch by 2 or 3 semitones, sing the part then return to original pitch. The problem with this is that the pitch shifting removes all the quality and life out of the sound; just not usable. And before you ask, no I don't think I'm using quality pitch shifting software and I don't think I will be buying some any time soon. if only I could change the sample rate to the equivalent of 2 to 3 semitomes and not lose quality.

So, do you think I have any other options? Some other method perhaps?

Your thoughts?

Cheers,
FM
 
pure.fusion said:
What I need is to somehow drop the pitch by 2 or 3 semitones, sing the part then return to original pitch. The problem with this is that the pitch shifting removes all the quality and life out of the sound; just not usable.
Click to expand...
I'm a little confused here. You speak of 'pitchshifting'¬> is this a different thing to you than varispeeding ?
If it is, then I would suggest varispeeding if you can. On my DAW there is a pitchwheel like they have on the old cassette portastudios. I'm able to drop by 5 semitones and raise by 2 semitones. I often use this for treble and quad tracking backing vocals and sometimes on lead vocals, I'll drop by a semitone, 2 at the most because once you drop to 3 and more, it just sounds ridiculous. It does tend to kill the quality unless you're seeking to sound like a kid.
But if by pitchshifting you mean the same as varispeeding, ignore what I suggested.
 
Thanks for your guys.

Yeah, not so sure what varispeeding is. I suspect without looking it up that it's changing pitch and time together which is probably less destructive to the sound.

I need whatever it takes for me to sing a few semitones lower at the original speed, but modify the sound so that the final outcome is those few semitones higher to match the key of the recording; with the obvious problem being the change of character of the voice and quality.

I had a look at a few options on the weekend (I can't remember the names right now) but was surprised to find an adjustment for both pitch *and* character of the sound which could be my solution if the quality didn't deteriorate too much.

From what I read, WaveTunes looks like it gives the best quality.

Cheers.
 
In Cubase you can change the global key of the song and then return it to the original once you'r done.

What DAW are you using?

Cheers :)
 
Mo Facta said:
In Cubase you can change the global key of the song and then return it to the original once you'r done.

What DAW are you using?

Cheers :)
Click to expand...
I haven't wroked with Cubase but every pitch shifted vocal I've ever heard sounded fake to me. Does Cubase somehow avoid that?

And to the OP .... why does it have to end up in the same key as the original?
 
Lt. Bob said:
every pitch shifted vocal I've ever heard sounded fake to me
Click to expand...

Yup me too. That why I'm a bit excited at this Melodyne plugin that changes 'character of voice' to compensate for that. Still, I predict the quality will suffer too much to be usable but I'll know after I get some time to test it out.

Lt. Bob said:
Does Cubase somehow avoid that?
Click to expand...
No. Not that I know of, well at least not my old version anyway.

Lt. Bob said:
And to the OP .... why does it have to end up in the same key as the original?
Click to expand...
Is this one for me too (I'm not the op)? I'm a stickler for detail. Hopefully I will be fooling even the closest listener that it's the original song up to the point where the lyrics change. So it has to be in the same key, otherwise it's a dead giveaway.

FM
 
OK, that feature didn't become available til Cubase 4 or 5 and neither did Varispeed. In the newer versions of cubase you can select from a variety of algorithms that will best suit the application.

Time and pitch operations almost always have artifacts. For vocals it's best to use an algo that retains the formant but changes the pitch. This is indeed possible these days, but it will never sound like you're actually singing naturally because we can't accurately model the changes that occur in your vocal chords when you change key. Not only that, but there are a million other variables that I'm sure we're all very unaware of.

Anyhoo, best of luck.

Cheers :)
 
Right,

Just tried singing 2 semitones lower and changing pitch up with Melodyne plugin to suit the song.

I'm actually seriously impressed with the result from Melodyne and what they've achieved. But no, the quality changes too much and it just doesn't sound like me singing. You were right, it doesn't really cut it if the vocals are up front.

Certainly I might let one stay if it was one out of 5 harmony parts, so I'm glad I know now.

Cheers.
 
pure.fusion said:
Is this one for me too (I'm not the op)? I'm a stickler for detail. Hopefully I will be fooling even the closest listener that it's the original song up to the point where the lyrics change. So it has to be in the same key, otherwise it's a dead giveaway.

FM
Click to expand...
you are the OP (original poster).

I don't agree that a key change is a dead give away unless you're gonna have the two side by side. You're talking 2 semi-tones right now which is only a step. VERY few people are gonna notice that unless you're trying to fool someone with perfect pitch which very few people actually have. The VAST majority of listeners won't have a clue.
It is a conundrum since fake sounding voices are a bigger giveaway in my opinion.
Good luck.......
 
Lt. Bob said:
you are the OP (original poster).

I don't agree that a key change is a dead give away unless you're gonna have the two side by side. You're talking 2 semi-tones right now which is only a step. VERY few people are gonna notice that unless you're trying to fool someone with perfect pitch which very few people actually have. The VAST majority of listeners won't have a clue.
It is a conundrum since fake sounding voices are a bigger giveaway in my opinion.
Good luck.......
Click to expand...

Case in point Bob,
I did the Armistice 'mix this' a while back.
I raised the whole thing a semitone and sped it up and no one noticed.
Or at least, no one mentioned it.

There's certainly a limit to how far you can go though. 2 semitones is probably on that limit.
 
Steenamaroo said:
Cas in point Bob,
I did the Armistice 'mix this' a while back.
I raised the whole thing a semitone and sped it up and no one noticed.
Or at least, no one mentioned it.
Click to expand...
if anyone here had noticed they would have said so.

Look ...... I have the most egregious example of how clueless audiences are:
A buncha years ago some friends of mine that did a jazz duo (piano/guitar) were playing at a place in Baton Rouge called Poets. Two very good players.
It was primarily a jazz club and the audiences there were primarily jazz fans who came to all the jazz clubs and liked quality music .....

So there were several musicians there that night and we got to talking about this very subject during the break and my friends decided to do a test.
When they got up the first song they did was Misty and they each played the entire song in different keys 1/2 step apart!!!!
Every melody and chord and fill and even things they played in unison were ALL 1/2 step apart.

Now ...... that's as bad as it can be ..... you really can't get any more discordant than that. It was possibly the m,ost horrible sounding thing I've ever heard!
The musicians were in stitches but ...... not one single person in the audience noticed at all. No one so much as looked up or startled or anything. They simply didn't notice.

Another example of how clueless the general populace is are the many many vids you can find of people who clearly think and have been told that their singing is great but actually sound like a hog getting slaughtered.
 
The more the audience drinks, the better the players sound. That's why karaoke singers all think they're AI quality.
 
Lt. Bob said:
you are the OP (original poster).

I don't agree that a key change is a dead give away unless you're gonna have the two side by side. You're talking 2 semi-tones right now which is only a step. VERY few people are gonna notice that unless you're trying to fool someone with perfect pitch which very few people actually have. The VAST majority of listeners won't have a clue.
It is a conundrum since fake sounding voices are a bigger giveaway in my opinion.
Good luck.......
Click to expand...

Doh, I thought OP was the moderator for some reason.

I would notice if something played and it wasn't in the original key, it would just stand out like dog's balls. Anyway, lowering the key just moves the problem more to the other end and makes the lower parts harder.

Must be original pitch.
 
Lt. Bob said:
no it wouldn't .... and in most cases I doubt you'd notice unless you have perfect pitch.
Most people don't.
Click to expand...

Hey Bob, just out of curiosity, If you and I were in the middle of the forrest (ie, no music source around) and I asked you to whistle or sing me a 'E' note, could you?
 
You must log in or register to reply here.
Back
Top