Stretching an audio clip

Hi, I’ve recorded someone singing “oooooohhh”. Except he didn’t sing it for long enough. I cannot get him back into the studio.

So, I want to extend the oooooohh somehow. I’ve tried paulstretch, but it sounds funny. I’ve tried slowing the tempo, but it sounds choppy.

So I duplicated the clip and appended it to the original. Did this a few times, to make it the length I want. That seems to work, except between the clips, there’s an noticeable “click”. Like “ooooooh” click “ooooooh” click “oooooh”.

I’ve tried to remove the clicks by using a crossfade, but that doesn’t help.

I’m out of ideas. Any ideas out there? Thank you!

Brian

Stack the Oooos one above the other and use the Time Shift Tool (two sideways black arrows) to push them sideways more and more as you go down. They will all play at once unless you prevent them with the SOLO or MUTE buttons. Now use the envelope tool (two white arrows and bent blue line) to fade gently between them.

http://manual.audacityteam.org/man/envelope_tool.html

Koz

Are you suggesting to use the envelope tool to fade out one clip while the other one fades in? I tried exactly that, but not with the envelope tool. I used the fade out and fade in, and it sounded wobbly, like i could hear the decrease and increase in volume at the fade points. What is the benefit of using the envelope tool?

What is the benefit of using the envelope tool?

Fade In and Fade Out are fixed. Their unchanging fade characteristics don’t always meet the needs of the production and there is even Effect > Studio Fade Out which is a different fade process designed to work around shortcomings of the existing Fade Out.

The advantage of the envelope tool is its free-form nature. You need it slightly louder, move the curve at that time up slightly. Famously, you can adjust volume word by word if you need that. There’s nothing automatic about it, though, so you could be there for a while. If the performer was expressive while they were presenting the Oooo sound, this patch job could be very difficult.

And yes, it’s perfectly correct that all the methods of manipulation that take the sound apart and put it back together again generally sound funny. Effect > Change… tools generally fall into this category, particularly because most people want them to do things they don’t do. They won’t do gender or age changes for one example. “Make my voice sound like” postings almost always fail.

Have you constructed the limit scenario? When do you give up and try to get the performer back—or use a different performer? That’s one of the more painful jobs of the Producer.

Koz