Bug: Attenuate can't work with pre-ambience

SpectraLayers
1

Hi @Robin_Lobel ,

I’ve noticed that, as great as it is, for Attenuate to work it needs to have an noise floor/ ambience reference after the target of attenuation, otherwise it won’t work. This makes it impossible to use on anything at the end of clips or right before wanted sounds that are not at noise floor/ambience levels.

Please modify it so it can also take into consideration ambiences before the events as well.

Thanks!

2

In the manual it says Attenuate module is working with levels relative to the noise levels surrounding the selection. I think it is just how it behaves

3

Yes I know, I made the post exactly because it is not working like that for me. It needs ambience after the material to de attenuated, otherwise nothing happens.

I’ll fix my last phrase.

4

It’s worth noting spectral synthesis for insertion of interpolated audio from surrounding content requires:

  1. A time domain buffer sufficient to accommodate the FFT window size
  2. Additional time domain room to crossfade with existing audio when saving the result

For the above reasons, spectral processes will encounter limitations at (or very close to) the start and finish of the WAV file.

There are two methods to deal with this:

  1. Reduce the FFT window size. This shortens the required time domain buffer. The downside of this is reduced accuracy for fixing low frequency problems.

  2. Insert silence audio into the WAV file to extend the start or end of the file (depending on where the problem is located). This eliminates the restrictions on spectral processes by providing sufficient time for the FFT window and crossfade.