I am making a recording for work. I have a pre-recorded track of spoken word. It is supposed to be a "relaxation" cd but the person on the recording spoke too fast to be relaxing. How can I slow that track down and not the others? Thanks.
You could try time-stretching it, but it will probably sound metallic and artifact-y unless you use an expensive plug-in like Melodyne. I would recommend re-recording with a different voice talent, but that's just me.
You could also try just manually editing and adding space between his sentences and phrases. It will sound more natural since you aren't really changing the "speed" of the file. You are only changing the "pacing" of the presentation. Of course, depending on the source material, it may or may not work.
If you do that, try and find some room tone at either the front or back of the file and use it to fill in the gaps so that the inserted space sounds consistent with the rest of the recording.
Re-recording as sonnykeyes recommended is going to give you the best sounding recording.
I've only played around with the Time and Pitch Factory, but you can try the following:
1. Duplicate the track and copy the audio file onto it.
2. Mute the original track. You don't have to, but I would hide it. 'H'
3. Trim a section of the audio file, e.g. a complete phrase. I'm bar/beat oriented, so maybe a phrase that's about 2 bars long. Use the Marque tool to make your selection, then click Control-C to trim away audio on either side of this selection. The point is you want a small bit of audio to experiment with versus 60-90 minutes worth or whatever.
4. Double click on the remaining audio file to open it in the Sample editor.
5. In the sample editor, open Factory>Time and Pitch Machine. You'll see a Tempo selection with Original and Destination areas.
In the attachment you'll see that I changed the Destination Tempo from 120 to 100 bpm. Maybe you're working with SMPTE (hours: minutes: seconds: frames) so adjust the Destination setting according.
I'm also using isotope Radius. If you don't have that, you'll have to choose one of the algorithms supplied by Logic. Maybe someone else will chime in here, but you could try Universal or maybe Monophonic.
In the Pitch section at the bottom, leave Transposition and Harmonic Shift set at 0.
Make several copies of this trimmed bit of audio in step. Put each on on a separate track. Test out a Tempo change. Then go to the next track, try a different setting until to achieve a tempo and voice quality that is acceptable.
All good stuff above. You should also try Flex on the track. It would be easy to try different tempos with flex as it's non destructive, just slow the tempo down until you begin to hear artifacts. Also try the different algorhythms. Once you get a slowed-downness that sounds OK , do a bounce in place and then try lengthening the gaps as Eli suggests.
My gut feeling is that you'll still want to rerecord the voice, especially if there is no musical background to mask slight imperfections.