1. Register NOW and become part of this fantastic knowledge base forum! This message will go away once you have registered.

Boosting/compressing low/high level audio in specific areas of a recording?

Discussion in 'Recording' started by BCMusic, Jul 21, 2010.

  1. BCMusic

    BCMusic Guest

    I may be explaining this in a way you may not understand, so please try to read carefully. I have a few recordings I'd like to try an idea on. Basically I'd like to take low level words (anything lower than 40db) and boost them up to 60db, and any high level words (above 70 or 75db), compress them to 60 or 65db. Is this even possible? Basically I just want everything level, and not low in some parts where you have to turn the volume up, or high enough where you need to turn the volume down... It gets to be a huge PITA when I go and do it manually, especially if it's a 16 bar take... I have a few audio editing/production/mastering software packages and a myriad of plugins. If I need to buy any plugins/software in order to achieve this, then please let me know. Thanks!


    P.S. - As it stands, I have no outboard gear, so if this is not possible without it, then I'll pass for now...
  2. Boswell

    Boswell Moderator Distinguished Member

    A compressor will do exactly this in principle, but the results may not be to your liking. However, if you already have a compression plug-in, it won't cost you anything to try it. Set the threshold so your quietest passages are uncompressed, and the ratio to be very steep (20:1) or infinite (limiter). Use the make-up gain to restore the output level.
  3. rwogh

    rwogh Active Member

    Multiband Compression

    The first response is absolutely correct. Since you're looking to just get all frequencies to the same dB, you just need a basic compressor.

    However, there are such things as multiband compressors that allow you to apply various compressions to different frequency ranges.

    multi-band compressor Search | Musician's Friend
    Dynamic range compression - Wikipedia, the free encyclopedia
  4. rwogh

    rwogh Active Member

  5. RemyRAD

    RemyRAD Well-Known Member

    What you are asking for is what could be described as " AUTOMATIC VOLUME CONTROL". In inexpensive battery-operated cassette recorders and certain communication microphones. This was adequate for better vocal intelligibility. It wasn't better for fidelity. In fact it was low-fi. It also highly filtered all of the low frequencies and high frequencies out, to eliminate overly obnoxious background noise from traffic, explosions, air conditioning, etc.. Of course we still heard all of that background trash rush up between each word and pause. Now while folks are talking about spectral processing utilizing a multi-band compressor/limiter, this unfortunate noise oriented scenario can still be quite problematic. Sometimes all one needs is a single frequency weighted single band compressor. Or a compressor with a " sidechain" insert for an equalizer. This way you can make the compressor only sensitive to most of vocal frequencies between 300 & 3000 cycles or hurts or Hertz or whatever kind of bondage & discipline you want to beat your vocal into.

    Now, no one has even brought up the subject of including some kind of " Downward Expansion", to keep the noise rush under some control. We don't always want you to sound like you are drowning and gasping for air. So some hardware compressor/limiters feature an additional built-in noise gate, whose threshold can be set differently from the compression. So a lot of this depends on whether you are making music or doing forensics? A noise gate or downward expander can be a much greater challenge to tweak than a compressor/limiter. If incorrectly adjusted, you might as well be watching Vince with his Slap Chop, in an audible kind of way. You'll love his nuts.

    I'll take my KEPEX after my 1176 thank you. Of course you can also virtually draw your own in the most rudimentary of most dynamics processing included processing with the lowest end of professional software, such as the $50 entry-level version of Sound Forage by Sony, formerly Sonic Foundry which is now making something else. The idea is to automatically turn off the microphone to make the vocal more intelligible and then quickly turning it down so you hear none of the background trash. Don't even think about the noise reduction plug-ins in the software. Although removal of some steady-state background noise, and before any compression is initiated, isn't a bad idea. But you have to go lightly and you have to only sample a short bit of only the steady-state background noise. This takes a software photo of the noise to be removed. So, it works good on continuous steady noises such as air-conditioning, buzzing and your cat purring. It won't do much for the TV, screaming kids and dog barking in the background. The compression will only makes it worse. Sometimes there isn't anything you can do with one of these scenarios to improve a bad recording. But there you go anyhow.

    Forget about the multiband it's not applicable here
    Mx. Remy Ann David
  6. Adler

    Adler Guest

    I have to say that editing these parts manually will yield the most transparent results. However, as this process can be quite tedious, it may make more sense to try a downward compressor (usually referred to simply as a compressor) in conjunction with an upward expander. The compressor will bring down the loud sections and the upward expander will bring up the soft sections. I use Flux Solera which allows both processes to be performed in parallel.

    Chris Adler
    Mindtree Studios
  7. Adler

    Adler Guest

    I agree multiband processing is not the way to go in this case.

    Chris Adler
    Mindtree Studios
  8. Davedog

    Davedog Distinguished Member

    I assume that you have these vocal tracks as separate entities from the rest of the mix and are simply trying to get them to 'stick' when and where its needed. Tedious as it may be, fader riding is my preferred method, although the Waves bundles have a fader-rider plug-in that works quite well. If you are trying to get this out of an already finished mix then downward expansion behind a compressor is the ticket.
  9. RemyRAD

    RemyRAD Well-Known Member

    Here is one of the other obstacles to try to do what you want. If in fact your vocal track is on its own individual track, kudos to you. Of course, the compression & limiting trick is easily accomplished with just about any software. But here's the rub. When you compress or limit rather drastically you also bring up all the junk in the background that you don't want such as headphone bleed, HVAC, heavy breathing. Most of this can be easily controlled if you understand how to " draw" your own compression/limiting/downward expansion curve. You can tell the software (by drawing the line with dots) so that below his quietest passage, the volume turns down below a set threshold. It's a touchy adjustment. But once you find the right spot, your highly compressed & limited vocals will take on a new word tighter quality than ever before. But it's that downward expansion threshold that requires a great deal of finesse to make it function properly so, you'll need to practice that. It's something I use regularly. Many of the bundled software includes some kind of dynamics processing. Many of them feature a graphical user interface that shows you the compression curve. The trick is to set a dot at the lowest usable sound. Then, you can drag the rest of that curve nearly straight down or at a slight angle. Many hardware devices have similar capabilities built in such as DBX and others. But the same can be had in your bundled software without having to shell out big $ for somebody else's plug-in that figured out how to do what I am now describing. Adjustment of the attack and release also makes this rather critical. If trying to do all of this at once is too much for you, just first add the compression/limiting. Then do a second pass with a unity gain compressor/limiter and just draw in your downward expansion below your set threshold. This is actually the best way to do it since you want your downward expander to have slightly different attack and release speeds from your compressor/limiter. And it's for this reason I use separate outboard compressor/limiters and separate downward expander's since I want the different attack and release times on both. This still can be done within software without having to purchase anything else.

    Thinking outside of both boxes
    Mx. Remy and David
  10. audiokid

    audiokid Staff

    I'm looking at the CharterOak SCL-1. Sounds like this comp is able to isolate specifics artifacts in greater detail without it effecting/ ducking the other frequencies, as much. It may be going into my Dangerous Master chain.

    Check this out.


Share This Page