The Joys of Noisegate.
  • Equestria NarratorEquestria Narrator
    I would be curious to see what some of the others who are on the site, like Aggie, who work with this kind of thing more often think of Noisegates?   I've found that, after applying a round of compression, that a passing through a noisegate around -38 removes almost all the unwanted sounds.   Also depends on your gain.  I recently, again thanks to Aggie, learned that having a mid-range set gain and the compressing is WAY better than higher gain and having to deal with the artifacts that generates.  I don't even need to use the noise removal now, noise gating pretty gets me exactly what I want with no degradation of the audio.  The only caveat is you have to be careful as noisegate will cut out soft sounding things like F and W sounds if you're not careful.  
  • AanokAanok
    That's interesting, if this means you can get rid of noise removal I'll definitely have to try it out.
  • Equestria NarratorEquestria Narrator
    The last couple of chapters I've done have not need to use he 'noise removal' tool in Audition works well but need to be careful about those soft sounds and also when you WANT breathing noises or sighs as well, but have not really proved a problem. 
  • Equestria NarratorEquestria Narrator
    Well...I may be changing my tune here....had quite a bit of issue with it last night after I fiddled with my gain.  Again with the soft noise sounds, specifically F sounds tend to get cut off....I ended but going back to Auditions noise removal for that one small bit I was doing.
  • AanokAanok
    I tried fiddling with this the other day and I too had some issues with soft sounds being cut off and a few artifacts occasionally appearing during supposed moments of silence. I did manage to solve that, but at the cost of introducing some unpleasant distortion over transients (but this might just be me being a bad sound engineer).

    I think the best thing is still using noise removal, especially the one available on Audition's waveform editor, which is not adaptive but built on a noise sample, like in Audacity. I feel it produces less distortion than Audacity's algorithm, though, especially given that you can customise quite a number of parameters.
  • PrismicPrismic
    I start with audacity's noise removal with standard settings, which I sample with the quietest ambient noise I can find in my track. This cleans up most of it. I still like to use a bit of noise gate on top of that though, with a quarter second attack time to keep it focused on killing short spurts of noise vs quiet parts. So far I've liked the team up quite well.
  • Equestria NarratorEquestria Narrator
    Yeah, I'm still fiddling.  I was talking to an audio engineer and he likes to use EQ to isolate the frequency the background noise is showing and lower crank it down.  I've not tried that but when I work on my next chapter this week I'm going to give it a try.  Auditions noise removal is still pretty good tho, I'm just being picky. I -mostly- like the noisegate as it can also catch other noises that Noise Removal does not catch...I.E. Anything below the noise floor level.
  • Equestria NarratorEquestria Narrator
    An update here:  I still use the noisegate, but now I use the built in one on my ART VOICE Channel box.  Tuned just right it works a treat, although I still pass it through noise removal with a sample from what little ambient noise there is inside the waveforms themselves.  It really helps with keeping down the quiet incidental noises but can still be touchy.  I still find myself having re-record a line when certain soft F/W noises get nipped too much.
Add a Comment