[LAU] WhyPhil? A webshow about Linux Audio (and then some)

Fons Adriaensen fons at linuxaudio.org
Fri Jun 3 18:01:35 UTC 2016

On Fri, Jun 03, 2016 at 10:43:11AM +0100, Yassin Philip wrote:

> Hmm no, the voice is mixed (by pulse) with the music ; I think
> existing episodes will stay "as is" :( but I want to get it right
> for the next ones. I'm going to try input => compressor =>deesser =>
> reverb ; but about that " lowpass shelve filter with a cutoff around
> 150 - 200 Hz" business, is there a way to insert such a processing
> inside my realtime effect chain (Calf Only, ATM) ?

According to what my ears tell me, the problems with the presenter's
voice in these videos are just the result of using effects where you
shouldn't. Just forget about compression, deessing, reverb etc. None
of that is needed. We've had perfect voice presentation (e.g. on radio)
decades before any of those effects even existed.   

Just make sure you have a more or less correct level, eventually use
a highpass or LF shelf filter to correct the mic response if you're
talking close to it. If you're a perfectionist use a peak limiter,
ensuring it doesn't do anything most of the time. All the rest is



A world of exhaustive, reliable metadata would be an utopia.
It's also a pipe-dream, founded on self-delusion, nerd hubris
and hysterically inflated market opportunities. (Cory Doctorow)

More information about the Linux-audio-user mailing list