After using this tool for over a decade, I've realized that its UI is setup in reverse from what you would
expect, and visual representation of the effect is as well.
Here are some screen grabs of a recent session cleaning a mono VO track.
I have custom settings, and use an envelope to be more agressive outside voice range, protecing the voice.
This image shows settings that have been giving me good results (counter-intuative to the UI y Axis negative DB scale).
Once processed you can see in the waveform on the left as well as the frequencey analisis, that the baseline noise has been reduced. (Red is the unprocessed, Yellow is processed in the Frequencey Analysis tool )
I was setting up a new system with my custom settings, and started second guessing myself, so I made the envoelpe based on the Y axis oreintation as well as the spectral frequency NR preview... But the results are the opposite of what you would expect : and envelope that protects the non voice range, and produces no meaningful noise reduction.
Red Freq Analisis is unprocessed, and magenta is post NR.
Here is an A/B output with Audio so you can use your ears instead of scopes 🙂
Its an odd one, but I thought I'd post it.
Best,
David
AdChoices