Obviously it doesn't involve the compressor. I do NOT want to raise the volume of anything that is at -15 db or lower. I DO want to raise the volume, a little of everything that is ABOVE -15 db. I thought it was done with compressor, but when I set the the threshold at -15, it doesn't seem to mean anything, because EVERYTHING gets louder. But there must be a way to keep all the talking (EXCEPT the S sounds, which are at about -15) up--a little bit
Copy link to clipboard
Use the Dynamics Processor:
The further over to the left you move that top segment, the louder it will get. It will probably overload, and almost certainly will sound dreadful, and you may have to experiment with it to get any sort of acceptable result at all. I have to add that I've never found the need to do anything like that at all, ever! But that is how you would do it.
It's very unlikely that you will achieve what you want this way. Have you tried using the DeEsser effect for the sibilance? And to make everything above -15dB louder, I'd use a compressor - and some make-up gain. This will mean that the peak values will decrease, and the overall level can be raised without crashing into 0dB. So everything that's important in the speech gets louder. It's also possible to downward expand everything below -15dB, which in general (if you do it gently) will sound rather better.
This is what downward expansion looks like, also using the Dynamics Processor:
Thanks, but I've tried with the compressor and no matter what I set the threshhold at, it pumps up EVERYTHING, no matter how low. I didn't think it was supposed to, but it does. HOWEVER, today I started messing with the spectral frequencies. Softer is darker, louder is brighter. I tried selecting one or the the other, making them louder and softer, and it actually seems to work. I' can't find a de-esser (including oeksound's) that works without messing up other stuff. But this weird method seems promising.