Does anyone have any tips on how to REMOVE unwanted reverb from a sound recording? I’m trying to clean up some sound in a video edit. Part of the video is a recording of a live performance of a song in a church. The performance was captured by the in-built mic on the camera, several feet away from the performer. Hence it is very ‘boomy’.
I’ve tried EQ and a compressor and even applying different reverb settings, but with not much success. You’d think that someone would have invented a tool for this.
BTW I’n doing it in Spark LE Plus from http://www.tcelectronics.com. It uses VST, AU and other plug-ins and is great. I can go over to Cubase if you’ve suggestions in that dept.
Answer: (From Cameron Haigh)
The ambient reverb of a room will almost always have much less high frequency sound than what you get coming directly to the mic from the sound source. I think EQ is your best option.
My suggestion on how to get the best results from your eq is this (using parametric eq – cubase has this, I don’t know about spark):
- turn up the gain
- sweep the eq around the bass to mid frequencies until your “boomyness” becomes more prominent. You have now found the frequencies that you want to remove.
- turn down the gain so that you are now cutting back those frequencies instead of boosting them (don’t do too much to start with, say -3db or -6db – give it more if you need).
- Experiment with the “width” (what is the correct term for this control? I don’t know!) knob to find the best setting (I’m talking about the control which adjusts how narrow or broad the range of frequencies affected will be.)
Now for some constructive eq:
- turn up the gain on another parametric eq
- sweep around mid to high frequencies until the sound has more clarity
- adjust the range/width/whatever-it’s-called!
- play with the gain again to see just how much gives you the best results
At this point your sound may become very thin & lacking in depth due to the prominence of higher frequencies over lower. Whether or not this is a problem depends on what instrument was being recorded. Was it a solo voice? Solo voice with instrumental accompaniment?
Final suggestions for making the most of the sound are:
- Can you add (ie.record) new musical accompaniment which will mask or replace the low-mid frequencies which you have removed?
- Perhaps you could create a new sound file of the performance with the pitch shifted down one octave. If mixed subtly with the equalised version of the original it might – repeat, MIGHT – add some depth without sounding awful.
Here’s a stab in the dark – I’ve never actually done this before. It relies on phase cancellation (just like the karaoke button on some consumer stereos). Your audio needs to have been recorded with a stereo mic – I presume it was:
- take your original left and original right channels.
- create a copy of the left channel and invert the phase of this file.
- mix your inverted left channel with your original right channel into a new mono file. This should sound more or less like the original sound source was removed and you are left with the “boomy” reverb because whatever sounds arrived equally at your left and right microphones have cancelled each other out. We’ll refer to this new mono file as “reverb channel”.
- Now mix the reverb channel with the original (NOT phase inverted) left channel into a new mono file we’ll call “final left”
- Invert the phase of the original right channel and mix it with the reverb channel – call this “final right”
- listen to your final left and final right files. Try mixing them together in stereo as well as mono. The theory is that these files contain more of the direct sound and less of the boomy reverb.
Do they sound any better than your original sound? Probably not, but I thought I’d throw this idea in just in case it was helpful. Keep in mind what your final output source is going to be. Be very wary of phase cancellation if you mix it in stereo and your final output will be mono (eg most consumer TVs) – always check it in mono to be sure it still sounds ok.