Learning about limiters

Limiters are useful for controlling peaks.
- you start mixing your tune and adjusting EQ with things in mono more or less.
- you tweak eq to remove sounds you don't want and to ensure you can clearly hear things you do want.
- you may add compression on various tracks to help make them more clearly audible more of the time.
--Distorted electric guitar is something that rarely needs compression the tube amp really is the compressor.
 
Limiters are useful for things that have very loud peeks compared to average volume. A limiter is just a compressor at a very high ratio.
 
Examples of where to use a limiter on a single track might be a bass that is being slapped. Normally, the sound of the bass is fairly soft compared with when the bassist slaps the strings. You want to hear the slap, but you need the average volume to be near the volume of the slap. So you limit the sound so the slap is closer to the average sound of the bass.
 
Limiting applied at the end of the chain helps make it possible to get the average volume you want for the track without over compressing the individual tracks.
 
So... if you mix a track and then do bounce to tracks (lower the gain on the master bus until there is zero clipping) you will see the wave form for your track. There will be a few loudest moments for the track. If you normalize to 0 or -.1 now, you have the track as loud as it can go without clipping and without limiting.
 
Limiting will shave these loudest moments allowing you to bring up the average loudness of the track. Different limiters do this differently. Some try to be more transparent (make it hard to know there was limiting applied). Some add some distortion at the point of limiting that helps the listener know this would be going louder, but it's limited so that the listener can feel the excitement that goes with the loudness without having the loudness.
 
You can also use limiters in other creative ways to massively distort things. In general, when you use a limiter to crush a vocal (like a decimator) to get a cool sound, you are not trying to be transparent. You are going for something new.
 
Limiters on the master bus make the most sense when you understand average loudness for tracks and have a target loudness. In that case, the limiter is like a mixing gauge. If you are having to turn it up too much, you need to double back to your mix and try to tighten it up. If you are only shaving a few dB off the mix, your are probably in the range of transparent.
 
Look ahead options on limiters help the limiter be more transparent but increase latency. This means if you aren't adding new tracks and you want it to be transparent, turning up the look-ahead setting probably makes sense.

iZotope has resources published that are written pretty universally. The "Mastering with Ozone" PDF is a good resource, but the others are worth reading as well.

https://www.izotope.com/e...-resources/guides.html

To answer your question more directly, a limiter is often classified as a compressor with > 10:1 ratio, although not strictly defined. Taming peaks/master buss protection is often useful (i.e., the vast majority of the audio passes through without triggering the threshold), but some also use the limiter to crush the crap out of the audio (often not recommended, especially on the master buss).

https://youtu.be/d7PYjalQ2vQ
 
I'm far from being a proficient user, but I like FabFilter stuff and I always have an instance of Pro-L on my master bus, even if it doesn't have to do any work. I guess that, at the end of the day, I use it to help raise the total level of my project and capture any rogue peaks that I am too lazy to automate out.
 
I generally set the output gain to -0.5 and alter the input gain so that I am sitting at around 0dB when using the K-14 scale. Seems to work okay for the sort of stuff I make...

Some general info you may find interesting re using limiters:
 
http://productionadvice.co.uk/brickwall-limiters/
 
http://www.soundonsound.c...s-should-i-use-limiter
 
http://www.practical-music-production.com/audio-limiter.html
 
I tend to use them only in a few places. Firstly they can be set so there is no gain increase at all but the output ceiling can be set to say -0.5 or so and what the limiter will do in that situation is simply stop any loud peaks from going in the red which is not bad.  I tend to open up tracks in an editing program and make the fine tuning adjustments there.  I do something similar only in that I don't reduce the gain of a peak just limit its upper level only. 
 
You can raise the rms level of a track or buss using a limiter followed by gain.  If you set a limiter to say limit its output ceiling to say -3db (note you are not adding input gain here either) then you can follow this with a gain addition of say +2dB with another plugin. Then the track will be 2 dB louder rms wise but still only reach -1 dB output ceiling.  (This will sound different as well to pushing the input gain by 2dB going in and limiting to -1dB output ceiling. The reason why is when you do it this way you are now pushing many more peaks up to the output limiting point so hence it may sound a little more limited. Better to limit just a few peaks and then add overall gain to the whole track. This is still something I do in the editor though)
 
Sometimes they can be good on a very dynamic vocal track.  In this mode I might use a limiter followed by a compressor.  The limiter will have a high threshold and not do much most of the time but they are excellent for catching a rogue peak for example.  The compressor that follows will often have a low ratio (ie 2:1) in this case and the threshold set so there is only 2 to 3 dB of GR.
 
The other obvious limiting place is in mastering when the client is asking you to create a loud master.  I feel that not all limiters are created equal in this mode.  For me I use PSP Xenon because it is one serious limiter and expensive too.  With this you can actually adjust how the transients sound so you have a fair degree of control over the limited sound which in many cases is not often the case.  Xenon sounds best when only limiting to say around 4 dB or so. Once you start going over this the mix can break up and sound bad.  If I have to raise the rms value of a mix from say -20 up to say -10 then I get all three stages involved in doing that.  The EQ will often allow you to add some gain followed by the compressor and this is the important place say another 3 dB or so and that leaves the remaining 3 to 4 dB to be provided by the limiter.
 
Some like using them everywhere in a mix but that is because there are red lights coming on everywhere.  As I use the K system for tracking and choose a great reference like -20 (and meter everything through a VU meter on the way in) I never see a clip light coming on anywhere! So for me I don't need them for that job.
 
 

I am always using rms or VU metering after the limiter as well say in mastering.  But what you have to do here is to re calibrate the meter though.  If your master starts out at say 0 dB for -14 dB FS (premastered for example) then when the mix is at -14 (rms) the VU meter should be peaking at around 0 dB VU which is great. (meter must be calibrated so -14= 0 dB VU) But as you add rms level to a master the the VU is going to pin over into the red.
 
If I am aiming for say a -10 rms master now this is 4 dB higher.  So I reset the VU now (after the limiter that is) so -10 FS = 0 dB VU.  So now a -14 rms pre master will show -4 on the VU and as I make the master louder with EQ, compression and limiting the signal slowly comes up to 0 dB VU which now we know is representing a -10 rms master which is pretty damn loud!
 
Some seriously mastered very loudness wars masters can get up to as high as -6 rms but this is taking things way too far. The mix is usually breaking up by this point now.  No limiter on earth can get you up to say -6 rms and still sound clean and transient and punchy so it is a silly level to aim for.