Mitigate overdose of reverb

Yes, I’d just bought a hardware reverb unit way back when and as so often happens with new toy syndrome I got carried away on a song or two.

Is there a way to retrospectively counteract an excess of reverb on male vocals in a fixed mix that has other instrumentation going on?

No. That’s why we recommend straight recording as clear and perfect as possible and do all the effects in post, never touching the original tracks. Copy them to a thumb drive and lock them up somewhere.

If you had multi-track studio system, you could record both the straight and effects tracks side by side.

Koz

Unfortunately I’m left with what I have got koz, it was the eighties and I could only afford a four track portastudio. The cost to a home hobbyist of an 8 -track was prohibitive in the Uk at that time

I did record dry but subsequently had to do a fair bit of bouncing with effects applied. The lead vocal usually had its own track and reverb was introduced only during final mixdown to stereo cassette, but many of the masters have gone missing over the years.

I can’t play guitar any more because my wrists are shot where that is concerned, so for current music I tinker around low key with vsti synths etc on my laptop keyboard as I travel about. A hunt and peck approach to keyboard playing. Proper synth keyboards also get my wrists.

However I have quite a few tracks I never got around to finishing in the old days so I find it fun to see what I can do with them now and thanks to help here I’ve had a rewarding time.

Some of the higher end programs claim to be able to help with the convolver tools, but every time I try to pin someone down as to exactly how it all works, they hear their mum calling and have to leave. So unless someone corrects me, what you got is what you got.

I tell people I don’t play the piano, I play at the piano. Big difference.

Koz

I believe that is, at least in principle, possible to undo the reverb.

However, it will not be easy, particularly if it was one of the ones that used stretched coil springs as the delay element.

You’ll need much fancier software than Audacity.

Post an mp3 sample so we can get an idea of the extent of overdosed reverb.

I’m currently beefing up the tinny, bass ridden dynamics of '70’s and '80’s prog rock/jazz fusion CD’s I bought years ago and find I lose the reverb effect somewhat due to my incessant attempt EQ’ing back the loudness and brightness in the original that reverb removes. What I’ve found is portions of vocals/horns have their octaves split across the audio spectrum where quite a bit of the clarity is now shared between both high frequencies (around 3000-6000hz) and lower midrange (around 300-700hz).

There’s a setting and process by which Noise Removal will help get rid of echoes/reverb. I got it to work… once. It was such a surprise I wasn’t able to remember what I did or write down the settings. Koz

As I recall, it is basically:

  • creating the noise profile with the whole audio selected
  • Enter a reduction factor
  • set the sensivity to -20

This reduces all frequencies 20 dB lower than the original audio by the set reduction.
It is actually a pity that the sensivity doesn’t allow even lower settings, e.g. to reduce audio under -35 dB.
However, the tool isn’t not intended for this purpose after all.

This is me writing that down.
Koz

Thank you Robert (& Koz for mentioning it )

I’ll experiment with noise removal.

Hey! That’s not bad.

I’m also messing around now at balancing gain levels between original ‘reverby’ take and ‘de-noised’ duplicate and playing them back together

Original has a 10 second intro of assorted percussion and a couple of other interludes which don’t fare quite so well at the de-noised optimum for the rest of the track when the rhythm and/or vocals are on the go.

However, the tool isn’t not intended for this purpose after all.

No, but if you reduce the amplitude of the original work 10dB, take the profile, and then boost it back up, haven’t you done the same thing?

Koz

I should guess so.
However, the version for 2.1.0 is different and sensivity won’t be related in that fashion anymore.
In other words, pre- and post-amplification is the only way in future.