Correcting "cavernous" sound?

[Cellotron]

A couple things that have worked for me dealing with overly ambient tracks is to reduce the Side channel gain, and sometimes to boost in the mids around 500 - 900Hz range.

Best regards,
Steve Berson

[wado1942]

You can use notch filters to attenuate room modes, which adds clarity. There's nothing you can do to get rid of reverb, though. There's no substitute for capturing the audio properly in the moment.

[huejahfink]

Cellotron's method of reducing side gain is usually the first port of call for me, but it sounds like your recording might be from a single mono mic, so if that is the case this will make no difference.
This is where I consider the big guns.

Waves LinMB - used as a multiband expander.
I've used it many many times to reduce the reverberant parts of vocal recordings.
Reduce the gain on each band, and switch the range to positive.
You will want a fast attack and fast to moderate release.
Set thresholds so that it boosts the vocal sound in that band and then falls smoothly after.
Can take a bit of time to set up, and possibly some threshold automation.
The more prominent the vocal is, generally the better the results.

[RJHollins]

Originally Posted by brookshire7

Working with some video footage of a lecture. Due to the poor room acoustics and mic (low-quality camera-mounted shotgun mic), the audio track has a cold, cavernous quality. Any recommendations for bringing out some warmth, clarity and presence. Using Adobe Soundbooth and/or Sound Forge Pro 10. Thanks!

There are several tools that are called on to address issues. How you approach it, and skill level, and the expected results may very well be challenged and compromised.

In restoration work, tools like EQ, gates, noise reduction, spectral editors, can all be called on..

Specifically, there are certain NR packages that utilize 1000's of gates across the spectrum. Over the years you get to test, pull, experiment.

I have found that 'gating type' NR has an 'ability' to start shutting down the room sound. Eventually getting to a dry 'source' sound. I have then gone back in with room simulators, convolution reverbs, etc, to reconstruct a new stage. Very detailed, time intensive.

There was mention, somewhere, of a new plugin released that supposedly 'removes' reverb. Forgot the name, or where read. Might be of interest ?

[Tha Govna]

Originally Posted by RJHollins

There was mention, somewhere, of a new plugin released that supposedly 'removes' reverb. Forgot the name, or where read. Might be of interest ?

Probably SPL's DeVerb.

I've used the sustain control on SPL's Transient Designer, which DeVerb is based on, to remove room verb from a rap vocal.

Worked like a charm in that situation.

[Riccardo]

Zynaptiq Unveil could also be useful. I haven't used it on speech yet but we did use it during mixing (n the other room) to reduce ambience in very "roomy" upright bass track.

[huejahfink]

Appreciating hearing about some of these more case-specific 'focusing' solutions, thanks.

[chronmaster]

SPL DeVerb is definitely a great plugin for pulling the room sound down. Same with reducing the sustain on the SPL Transient designer. The Waves collection of noise gates are also pretty simple to use and can be really effective for reducing room sound.

[RJHollins]

SPL's DeVerb ... yea, that was the name I heard.

Not heard of Zynaptiq Unveil ... I'll look at both of these for more info.

Thanks !

[OpusOfTrolls]

+1 for multiband dynamics. Go nuts.

[ed littman]

any noise reduction software can control the reverb to a degree. The material may sound prosessed but voice may be more focused as desired.
Ed

[finetuner]

Originally Posted by Riccardo

Zynaptiq Unveil could also be useful. I haven't used it on speech yet but we did use it during mixing (n the other room) to reduce ambience in very "roomy" upright bass track.

Thanks for the tip.
I'm demoing it right now on some 'verby tracks.
Impressive! This is the best reverb/backdrop removal tool i've heard so far.

Denoising for these purposes can be very tricky as it's hard to obtain a usable noise fingerprint. On a full mix that is.

[Denis Goekdag]

Yes, denoisers won't really work as the reverb *level* depends on the signal level, and denoisers rely on a fixed absolute level print to do their thing. Also, the reverb will typically have a similar spectrum as the source.

[Thomas W. Bethe]

You might have a look at this Product Review - Zynaptiq Unveil

More info and demo here http://www.youtube.com/watch?v=C7bLQ...layer_embedded