Reducing overloud vocal

[Guy Johnson]

Hello, long time no here! Must come by more often.

Anyway, there's a CD I like - but the vocal is way too high, it drives me nuts. So I've tried a MS compressor including trying with eq'd side-chain with limited (geddit!?) success.

Any suggestions?

G

[Drew Stephenson]

Are you using a broadband or multi-band compressor?

[CS70]

Hello, long time no here! Must come by more often.

Anyway, there's a CD I like - but the vocal is way too high, it drives me nuts. So I've tried a MS compressor including trying with eq'd side-chain with limited (geddit!?) success.

Any suggestions?

G

I guess it depends on the mix. As Drew writes, a multiband compressor makes it easier try to find where the "loudness" of the voice is, by putting EQ, band settings and mixback in the same box. Even if lots of the voice's beauty and tone is in the midrange, loud (i.e. prominent) vocals are often achieved by a boost with a high shelf, or peaks around 1K, 2K or 5K, so one or two bands in these areas can be a good starting point.

Of course lots of other energy-inducing material resides there (cymbals, the whack of the snare, reverberation, ...), so by squashing them you can rob the track of its energy. Ultimately it all depends on whether the voice is really the loudest element in the band you identify and compress.. A spectral analysis could help - by looking at where the frequencies are when there's no vocals with respect on when there are, you might find out visually what band is a suspect (even if, of course, many mixes change the balance a bit and even the instruments when the vocals are present).

[Mike Stranks]

Don't have any 'this'll sort it' suggestions, but... welcome back!

[James Perrett]

How about Adobe Audition's Centre Channel Extractor tuned to the vocal frequency range?

[Guy Johnson]

Hi, thanks for those - tried the mulitband, and as expected it squashed the acoustic guitars. It's a laid-back unplugged sounding record.

I'll have a go at the central channel extractor idea.

[Drew Stephenson]

Out of curiosity, what's the record?

[Tim Gillett]

Yes interesting to know which CD. Also whether all agree the vocal is too loud. Some might have preferred it even louder...

[zenguitar]

Out of curiosity, what's the record?

"Loud singing with quiet acoustic guitars" by someone or other I never heard of

Andy

[Drew Stephenson]

Out of curiosity, what's the record?

"Loud singing with quiet acoustic guitars" by someone or other I never heard of

Andy

Not really your cup of tea Andy!

[Guy Johnson]

I'll keep the identity of the CD to myself, peeps, it's one I bought the other week.
Levels-wise - you put on a track, get the track going at a nice level and then when the vocal comes in, I want to turn down the volume.
I'll prob. have another play around with it today.

[Sam Spoons]

How about using a "vocal zapper" type app to produce a 'backing only' k@raoke track then mix that back in with the original to boost the level of the backing? I'm pretty sure there will be artefacts and they may well spoil the quality of the recording but it's a long time since I've tried such a process so they may have improved.

[Drew Stephenson]

Or a vocal isolater and mix it back in with the polarity reversed.

[Mike Stranks]

Have you thought about one of the new-ish 'disassemble your master back to its component parts' bits of software?

Don't know how successful it would be or if the cost is prohibitive.... jus' a thought...

[Wonks]

I'd simply throw the CD away. Life's too short.

[Hugh Robjohns]

I agree, Wonks.

I think as 'audio peeps' we all tend to have strong views and preferences over how a mix should sound, and consequently struggle when the person making the recording takes an approach that doesn't fit with expectations.

I've certainly returned or junked many commercial CD's in the past because I couldn't accept the quality of recording or performance. In some cases I've replaced the original disc with a 're-mastered' version of my own making where I've tweaked the EQ or dynamics to create a more acceptable (to me) rendition. But when the core problem is a mix issue (rather than a tonal or dynamics problem) there is very little that can be done to achieve the desired re-balance, and it's rarely worth the time to try in my experience.

In Guy's case, if a multi-band compressor can't sort it out I'd forget it and move on.

H

[Guy Johnson]

Thanks for the thoughts.

I've done a few tweaked CDs for myself - eq and dynamics, or imaging - I love 'Hurt' sung by Johnny Cash. But NOT with two acoustics panned hard Left & Right - it's obscenely distracting from the song and singer.

Another wish:

I'd like a plugin that is an expander, with no time constants (maybe enough not to track waveforms) - just an expansion algorithm based on ratio... perhaps with a drawable graph controlling the ratio that the signal get expanded by at what levels. There are still so many recordings that I can only listen to while driving, they are so squashed. At the moment I use PSP master compressor on expansion mode, with limited success.

[BikerDude]

Hello, long time no here! Must come by more often.

Anyway, there's a CD I like - but the vocal is way too high, it drives me nuts. So I've tried a MS compressor including trying with eq'd side-chain with limited (geddit!?) success.

Any suggestions?

G

Perhaps a combination of a good multiband mixer and normalization.
If you can isolate the vocals in the mixer and bring them down it might help but that's not usually exact science. It won't be 100% discrete so at best you'll be pulling out some of everything in some segment of the mids. Isolate some other backing parts and bring them up.
Normalization will bring up the parts without the vocal to a level closer to the level of the vocals. But when both vocals and backing are present they will stay at the same percentage wise. It just levels off the loudest overall and quietest overall parts. Flattens.