Compression vs Limiting

a limiter is just a compressor with the RATIO approaching infinity. Limiters probably will always sound better if you're used to listening to commercial music where they compress all of the dynamics out of the track. I would not use a limiter on vocals unless I was trying to go for a specific genre like death metal where it's all screaming vocals anyway and you want the most RMS out of the track you can get.

Though it's not strictly the case, generally limiters will have a zero attack time, i.e. they let none of the attack through (and if a brick wall limiter, then an infinite ratio in that they will not let anything above the threshold.) Most general purpose compressors have an adjustable attack time, and often won't give you a zero attack time, so they won't necessarily guarantee that nothing goes above the threshold since they don't clamp down immediately.

In practical usage terms, you would generally use a limiter when you don't want to manipulate the envelope of the sound, but just want to make sure it doesn't go over a particular level, and you'd use a compressor to adjust the envelope of the wave form, i.e. let some amount of the attack through, then pull down on the rest, i.e. make it punchier, less punchy, sharper, emptier sounding, etc... acording to how you set the attack and release.

But the lines are fuzzy and there are plenty of squishy scenarios in between. One example that kind of demonstrates the difference very clearly because it uses both of them together is the case of a bass track.

Some genres have bass lines that are inhumanly steady and smooth. One way to do this is to put a limiter on the base track and pull the threshold down considerably, so that there are no dynamics hardly left at all. What you end up with if you look at the wave form is more or less a bunch of flat sausages where all of the attack of the notes is removed so they are just flatted off, because the limiter starts chopping it down immediately.

Then, you put a compressor after the limiter, and set the attack for say, 30 or 50 or so milliseconds. Then what happens is that the compressor now puts some dynamics back, because it lets the first bit of each of those sausages through, but pulls down on the rest each one. But, because you limited the sausages first, the peaks are all absolutely the same. So you get some punch back into it, but it's absoultely consistent and steady.

So in this case, the limiter was used to prevent anything from going over a certain level, and the compressor was used to re-sculpt the envelope of each note, to make it back into a more realistic sound that attacks then decays.

ORIGINAL: droddey

Though it's not strictly the case, generally limiters will have a zero attack time, i.e. they let none of the attack through (and if a brick wall limiter, then an infinite ratio in that they will not let anything above the threshold.) Most general purpose compressors have an adjustable attack time, and often won't give you a zero attack time, so they won't necessarily guarantee that nothing goes above the threshold since they don't clamp down immediately.

In practical usage terms, you would generally use a limiter when you don't want to manipulate the envelope of the sound, but just want to make sure it doesn't go over a particular level, and you'd use a compressor to adjust the envelope of the wave form, i.e. let some amount of the attack through, then pull down on the rest, i.e. make it punchier, less punchy, sharper, emptier sounding, etc... acording to how you set the attack and release.

But the lines are fuzzy and there are plenty of squishy scenarios in between. One example that kind of demonstrates the difference very clearly because it uses both of them together is the case of a bass track.

Some genres have bass lines that are inhumanly steady and smooth. One way to do this is to put a limiter on the base track and pull the threshold down considerably, so that there are no dynamics hardly left at all. What you end up with if you look at the wave form is more or less a bunch of flat sausages where all of the attack of the notes is removed so they are just flatted off, because the limiter starts chopping it down immediately.

Then, you put a compressor after the limiter, and set the attack for say, 30 or 50 or so milliseconds. Then what happens is that the compressor now puts some dynamics back, because it lets the first bit of each of those sausages through, but pulls down on the rest each one. But, because you limited the sausages first, the peaks are all absolutely the same. So you get some punch back into it, but it's absoultely consistent and steady.

So in this case, the limiter was used to prevent anything from going over a certain level, and the compressor was used to re-sculpt the envelope of each note, to make it back into a more realistic sound that attacks then decays.

Dean, I really like your thoughtful example here. I wondered about 2-stage compression (using the the compressor/make-up gain to bring up the lows and then the limiter to level off peaks, for drums). You've given the opposite for bass guitar.

I suppose you could accomplish either 2-stage scenario by using 2 instances of a compressor VST in the FX bin.

Ozone3 VST multi-band compressor module places: (1) limiter and (2) compressor and (3) expander on top of each other. Each has separate threshold, attack/release and ratio settings.

An expander: is the opposite of compression. Low compression ratios (<1) enhance loudness below the expanders threshold. But it may be used as a gate when its compression ratios are highest (>1 to infinity). Few of us would dare to expound on attack/release settings of an expander. Methinks, few of us delight in expanding low level noise, period.

Mixsit noted A/R settings for limiters can be much lower, I think I understand this as: you're not limiting precious transients that exist in attack-waveforms, just offensive peaks that pop up too high.

ORIGINAL: droddey

Though it's not strictly the case, generally limiters will have a zero attack time, i.e. they let none of the attack through (and if a brick wall limiter, then an infinite ratio in that they will not let anything above the threshold.) Most general purpose compressors have an adjustable attack time, and often won't give you a zero attack time, so they won't necessarily guarantee that nothing goes above the threshold since they don't clamp down immediately.

In practical usage terms, you would generally use a limiter when you don't want to manipulate the envelope of the sound, but just want to make sure it doesn't go over a particular level, and you'd use a compressor to adjust the envelope of the wave form, i.e. let some amount of the attack through, then pull down on the rest, i.e. make it punchier, less punchy, sharper, emptier sounding, etc... acording to how you set the attack and release.

But the lines are fuzzy and there are plenty of squishy scenarios in between. One example that kind of demonstrates the difference very clearly because it uses both of them together is the case of a bass track.

Some genres have bass lines that are inhumanly steady and smooth. One way to do this is to put a limiter on the base track and pull the threshold down considerably, so that there are no dynamics hardly left at all. What you end up with if you look at the wave form is more or less a bunch of flat sausages where all of the attack of the notes is removed so they are just flatted off, because the limiter starts chopping it down immediately.

Then, you put a compressor after the limiter, and set the attack for say, 30 or 50 or so milliseconds. Then what happens is that the compressor now puts some dynamics back, because it lets the first bit of each of those sausages through, but pulls down on the rest each one. But, because you limited the sausages first, the peaks are all absolutely the same. So you get some punch back into it, but it's absoultely consistent and steady.

So in this case, the limiter was used to prevent anything from going over a certain level, and the compressor was used to re-sculpt the envelope of each note, to make it back into a more realistic sound that attacks then decays.

thats a really cool trick with the bass ! hmmm all this talk of sausages is making me hungry though lol

kev

ORIGINAL: RichardHiorns

Is that a yes or a no?
Here's the question again: Can a limiter be used instead of a compressor on vocals? Is this normal?

You said you tried one that sounded good. The question might be focused a bit more to say 'which limiter' and when you include 'how deep will it need to go?' an L2 and an LA2 in limit mode might be two different animals.

ORIGINAL: RichardHiorns

Is that a yes or a no?
Here's the question again: Can a limiter be used instead of a compressor on vocals? Is this normal?

What's wrong with the vocals that you think you need either one of these? What are you trying to accomplish?

If your vocals float between too loud and too soft, then perhaps a compressor would help you even that out, if that's what you're going for.

If your vocals are mostly how you want them, but peak once and a while, use a limiter to keep them from going over whatever level you want to keep them under.

If the vocals are good enough as is, don't use either.

It's like you're asking "What's better, a spoon or a wrench?". It depends on what you are trying to do.

Is that a yes or a no?
Here's the question again: Can a limiter be used instead of a compressor on vocals? Is this normal?

As noted above, a limiter is, essentially, just a simple type of compressor. Put another way, just about any compressor can be set to act as a limiter, but limiters can't be set to perform every possible type of compression.

Any processing tool can be used on any audio if you like the result.

Limiters commonly serve two purposes, one of which is useful:

(1) Saving the need to adjust input levels properly in the first place to avoid clipping.

(2) Where input levels have been properly set, a limiter can avoid clipping caused by occasional sharp transients, of the sort that sometimes occur with percussion (as an example)