By
trupro
Thu Aug 21, 2003 5:51 pm
Normalizing is a good way to go, but it only goes so far. If there are dynamic changes in the overall acapella, the noise level will be much more prevailent in the samples that are quieter. Normalizing will bring everything up in a sample, and that means abient noise as well. The result is that, instead of sounding even, the listener just hears that you turned the quiet ones up. If the original performance isn't too dynamic, normalizing is a good idea. But if not, here's what I do when I have the same situation: Run a compressor/Limiter plugin or process on the overall acapella before you cut it. A limiter is the best for the job, if you have it. Likewise, if you have outboard gear, record the acapella through the outboard Limiter into Soundforge first, then cut and normailze.
You want to limit it enough so everything comes up around the loudest points, but not som much that the noise level is boosted to an apparent level. Something like a 4:1 ratio, with a threshold low enough to cut injust touch lowest vocals in the accapella. If that squashes the louder parts too much, abviously lower the ratio a bit. Then, when you are done processing, you can normalize or mix the volumes just a bit to even them out some more. Then turn the cuts into individual files, save to disk, and put them in your sampler. Or, leave it uncut, and record the overall back into the ampler and cut there. No need to cut twice, right? Again, don't go crazy with the comp/limiter, but just enough to whip them into shape. Post again and let me know how this works for you.