Altering voices (speach)

Hi, first post here.

I’m totally new to recording and mixing audio, yet I find myself needing to record multiple voice trax. Problem is I have one voice artist who needs to be 20 people.

The Talent is great, gets right into character, but under all his tricks, eventually you start hearing the same voice.

Any tricks of the trade. Is there something I can do that will help make I sound like we?

Thanks in advance

Not really. You can use pitch shifting, some of the pitch shifters have the ability to change the voice’s “gender,” and you can play tricks with eq, panning, delays, and have the singer try to sing in different voices. Are you going for a large chorus effect? You’d also need to think about the acoustic space for all of those singers. But it still won’t sound like a lot of different people.

No way to make necessity a virtue?

The old way was the opposite of pitch shifting. Change the speed of the tape and force the singer to sing in a different range at a different tempo. The pitch and timing comes out right when the tape is played at the correct speed but the natural tone is slightly different. This can be done using a wave editor. Do a mix and shift it up or down in pitch in a wave editor, but don’t compensate for time. Let it play slower or faster. That what you want. Keep tabs on the EXACT amount of shifting. Import that changed wave into n-Tracks and record some voices to that. Open the vocal tracks in the wave editor and shift them the exact same amount in the opposite direction, putting them back in the same key as the original, again not compensating for time or temp. You want the vocals to speed up or slow down to match the original. Import the changed vocals back into the original song. As long as there are no resampling errors everything should just line up. Resampling of this type does no touch the data, it simply changes the sample rate in the header. It acts just like speeding up or slowing down a tape deck.

Pitch shifting, after the fact on parts sung in the desired pitch already, probably won’t give you want you want unless you are shifting them into harmonies. Then you still have the same voice with whatever artifacts that are added when the pitch shifting tries to keep the tempo the same.

Thanks alot. It’s all dialogue, but it’s supposed to be a bunch of different people, mostly they will be talking alone not at the same time.

Different “characters” in a play if you will.

is there a good wave editor that runs on windows or linux (prefferably)

I say get a second voice talent in. maybe even just use yourself on a couple. There is no way you can record 20 voices by one person and have them sound all completely different. you could maybe do 10 or 15 if you pushed it, but not 20. At least that is my opinon. Please feel free to prove me wrong. I am always a fan of spreading out voice talent to get a varied sound. I like doing drama stuff, though. I should post the skit I’m putting on my CD and get some comments. hmm…

edit: nevermind… :)

Quote (TomS @ Jan. 26 2005,19:41)
edit: nevermind... :)

ok.... ???

Yes, when attempting to produce any form of art in which voice is a component, I would say that it is advisable to acquire voice talent as a first step towards completion. Indeed, if I may, I’d suggest that it is sine qua non.

On the other hand, there is that speech synthesis software. Sooner or later we’ll be able to do away with the “talent,” methinks.

edit: here it is:

Well the “talent” although I would probably substitute an other word, is me. Where I live there aren’t to many people who either excell at public speaking, or 2 grade arithmatic, infact you might say I am the only person who could do it for thousands of miles. The project has tapped every sent including but not limited to software and machine to render the 2D and 3D imagry. What I am left with is a few mikes from long ago and alot of hardware that plugs into motherboards. The person who was to originally do all this and had the skills and equiptment died during a bad storm recently so I have what I have and I am going finish this #### thing even if it kills me :)

All suggestions are a huge help, any links or literature, anything at all. Thank you.

Listen to Mel Blanc. Once you know it’s him doing all those voices you can easily tell, but the toons characters are all very different because of little accent changes, timing of delivery, and some other timings I can’t really put my finger on. But the tone of the voice is very similar on all of them, Rich Little is the same way, though it’s harder to tell with him sometimes. I think for you purposes digging into these kinds of artists, and really listening to how they do what they do, not by watching them because mannerisms goes a long way into being convincing, will go a really long way. Of course, all the good ones have very generic non-descript natural voices already. It would be difficult for someone with a naturally very distinct voice to loose that distinction by just changing accents.

Of course, that brings up an interesting point. What type of thing is this? I mean, what type of voices are needed? Mel Blanc (WB cartoon guy, right?) did very exaggerative voices. That helps a lot. But if you want more serious stuff, true variety is more important. If it is very animated type exaggerative voices you want, then I’m sure you can do it yourself. If not, then there are probobly more people around than you think. I’m gonna post my skit. I got this friend of mine to do the girl part. She is by no means an actress and does not fall into the category of public speakers either. But I had her do it, and I have so many people ask me if the skit is real. It isn’t, but some good editing and getting her in the right mood went a long way to getting it sound pretty real. So I’m sure you can find people. Or maybe you could find an artist collab online and send them the script, they send you an audio file of their lines. There are more options than you think.


You’ve hit the nail on the head. It’s all pretty serious stuff, alot about pain and anguish, but told matter-of-factly. I’ve been working on “character inflection” and things of this nature, but there simply isn’t enough range in the material.

Fortunately many of the characters are scattered randomly through the piece. It’s in the sections really where I have 3 or three individuals involved in direct dicsourse I get into trouble.

Rich little I am not haha :angry:

Have you ever gone to/seen a GOOD one-person play???

An engaging speaker with stage smarts and a great talent for storytelling can get away with doing an engaging housefull of voices. The main word in that last sentence was: talent. Without it you’ll just come off sounding like one person reading a phone book. I think the last good example I saw was Patrick Stewart doing “A Christmas Carol” solo. There also seems to be a small core group of people working the animation industry who supply the bulk of voices (RIP Mell Blanc; he was a national treasure…)

Not a lot of people around who are used to doing that sort of thing these days. Back in the days of radio plays and serials there were a lot of people who had developed those kind of dialog muscles, but the age of dumbed-down TV saw them all go away. What a shame…

You might try contacting a local college or high-school theater group. You’d probably find a LOT of people who would dearly love to work on your project just to get a cover credit for their resume!