Spectral editing selection and cutting/pasting

Hello,

I wish to use spectral editing to cut all frequencies out from 200hz down to essentially zero. I want this so I can then use this as essentially another set of tracks for “mixing” vs using eq, a filter, etc. as this will insure zero changes in tone, etc…just wholesale level changes in that spectrum vs the rest, without any shaping, etc.

However, I cannot seem to select the set of frequencies properly…I seemed to have been able to highlight the portion I wanted once last night, but then could not cut that out to save in another file.

What is the correct method to do all of the above? I will probably need a little hand-holding.

Much thanks in advance!

“Spectral Editing” is simply a different interface for filtering.

If you want to reduce frequencies below 200 Hz, you need a “High Pass” filter. If you want to retain frequencies below 200 Hz and reduce the rest, then it’s a “Low Pass” filter that you need.

To achieve a very steep cut-off and minimise filter delay, use the Equalization effect (“Draw” mode).

I understand that. I can use another program I have and even filter at 64 db/octave. While that is very sharp, it’s not a straight cut, and when mixing with the rest (that is high passed at the same point and rate), there’s still an issue.

Is there a way of cutting out a range in spectral mode? When I have worked on projects in the studio, using CubeTec, etc., we have been able to go in and remove bits spectrally as needed vs filtering. Right now I’m preparing something for proposal, and there’s no budget for studio time for doing this (or using the multiband in CubeTec to bring up/down a range equally, with the compression off).

I should add that I was able to select the frequencies again (I had forgotten to re-enable spectral selection when I re-started), so right now I have just the frequency range I want highlighted, but how do I cut just these out, so I can then paste that into a new file, save that, and then save the remaining un-cut frequencies in another file?

I understand that. I can use another program I have and even filter at 64 db/octave. While that is very sharp, it’s not a straight cut, and when mixing with the rest (that is high passed at the same point and rate), there’s still an issue.

It’s impossible to make a “straight cut” (in infinitely-sharp filter) in software or hardware.

If you want a file with only low frequencies a low-pass filter is the answer. If you want a file with only high frequencies a high-pass filter is the answer. Spectral editing is not the answer.

BTW - the cutoff frequency is defined as the -3dB point, no matter how sharp the filter.

If you make a high-passed copy and a low-passed copy they will generally NOT re-combine “nicely”. Filters tend to introduce phase shifts (and possibly delays). You might have better luck if you can find linear-phase filters.

I tried a quick-and-dirty test, high-passed and low-passed at 200Hz and 12dB per octave (using a white noise sample). When I mixed the two I got a big (rather wide) notch centered slightly-above 200Hz.

The Equalization effect is linear-phase.

Exactly…using even the 64th order Butterworth filters in Cool Edit, there’s still an issue with “mixing” them.

Thank you all. So there’s no way of cutting out what I need spectrally in Audacity then?

YES. With the Equalization effect.