Snare and Compression

a sophmoric question

So, I’ve been n-tracking for two years now. Fairly regularly, at that. So, I’ve spent some time (at length actually) messing with the snare’s compression settings and here’s what I’ve observed:

I like the sound of the heavily compressed snare (at times) enough to want to explore it some. However, when I compress it (say 10:1), the output clips in the compressor and the mains clip as well. If I lower the output to the point where it doesn’t clip, I can no longer hear the snare (at least at a reasonable level) in the mix – it sounds whimpy and thin. I’ve messed with the attack a little, and the more ms I add to the attack, the more it clips – presumably because I"m letting the transients pass thru which would be the loudest part of the sound.

So, obviously, I’m doing something wrong. I don’t pretend to understand the finer points of compression, and maybe that’s the starting point. I’ve trusted N-track’s settings for the most part (on the drums anyway), and haven’t really had much need to diddle with it.

Oh, one other thing: All drums are sequenced and not played live, so I have no idea how they were originally recorded. I’m using Jamstix for what it’s worth – a really decent drummer, seeing that I’m not. It has it’s own compression, but that’s not really the point. I don’t use Jamstix compressor, and would really like to know how to use N’s compressor a little better. :)

Your turn.

Bill


(billthecat @ Jul. 19 2007,03:08)
QUOTE
I don't use Jamstix compressor, and would really like to know how to use N's compressor a little better. :)

Your turn.

Bill

This is a good starting point:
Jezar's How to use a compressor

I’ve got sort of a beginner’s primmer on compression here.

Hey guys:

First of all, many thanks for the replies. I really do appreciate this forum and the wealth of information that exists here.

Secondly, I apologize for not doing a good job of explaining the situation. I think I’m familiar with the basic purposes of a compressor. I use them extensively on vocals (mainly the ratio), guitars (the ratio and attack) and certainly on bass (again, ratio and attack).

I think it’s the threshold’s function that’s getting me. I understand its purpose, just not the function. The purpose (perhaps only to show my real ignorance) is to set a level at which the compressor kicks in. I get that. But the functions served seems only to be to drive the snare (or whatever I lower the threshold on) into the red zone – it clips. If I start futzing with it, I do one of two things: I redline the track and the master channel, or I bury the snare out of the mix altogether. Additionally, I change the sound of the snare (or whatever). This is the part that interests me. I love the sound of a really compressed snare – I like the sound when it’s clipping the master levels. But obviously that’s not going to translate well at mixdown.

Now, to be fair, I read thru both of the articles Mark and Captain Damage posted (again, I want you both to know that I appreciate it alot!), and I went back to some tracks and futzed with them some more - particularly the threshold. Same deal. Either over-driven, or lost. And it didn’t seem to matter really, how much or how little I tweaked the values. If I lower the threshold, I MUST lower the output, or it clips. And by the time I’ve lowered the output on that track, the snare is lost.

The attack only seems to change the timbre of the snare – doesn’t seem to do much to the level (excepting the transients, which presumably are what is changing the timbre). The release seems to have little to say in this discussion.

Again, I’m referring to N’s compressor, so that we all should have a reference to agree on. I’d be happy to share sounds, settings and the like, if anyone wants to take a crack at fixing me…

Thanks again, for all the input.

Bill

I think you need to experiment while soloing the snare. If you’re getting frustrated when it gets lost in the mix, you need to eliminate the rest of the mix. If you have to lower the output when you lower the threshold then do so, but then raise the track volume to compensate. Also, try another compressor plugin. Although they basically function teh same, different plugs will have different sounds and user interfaces that might make a difference to you. I don’t use the n-Track compressor so I can’t comment on that specifically.

That does sound strange. I can’t speculate what’s going on. I do have jamstix and N-Track Compressor so if you want to post your song file (just drums no wave files to keep it small) I’d be willing to take a look and see if I have any ideas for you. Maybe save it at the point where it’s clipping?

I find the n-track compressor a little kooky. I’m not sure it it’s me or the compressor but hey…

I use the Kjaerhus Classic compressor quite a lot and it seems to behave exactly as I expect it too. Having a nice simple “Gain Reduction” light to show when the compressor is working is very helpful rather than trying to figure out what Flavio’s graph is doing. That way I tend to use my ears to mix rather than my eyes and brain.

One tip with compressors that someone once gave me is to make sure that the compressor’s output volume is set so that when the compressor is bypassed the volume of the material is the same. That way you are hearing the effect of the compressor, not the additional volume (which always tends to make things sound “better”). When you’re done A-B-ing then you can play with the output volume.

n-Tracks compressor has a “compensate” function that will cause the volume to go up as the threshold is lowered. There is a button to turn it off, and it defaults to being on.

That said, I’ve noticed quite a few software compressors that don’t act like analog counterparts. With a set threshold and a ratio of 2:1 (for example) I’d expect the end volume to go down when the ratio is changed to 4:1, 8:1, 20:1, ending in hard limiting. Many don’t. I don’t get it. I’ve also noticed that lowering the threshold also doesn’t cause the end volume to go down. Not all compressors are like that, but the ones that do must be employing some kind of compensation, just as is the default in the fasoft compressor, even if they say they don’t.

There is at least one compressor out there that gets the ratio backwards – 1:1 is hard limiting and 20:1 is virtually no compression at all. This is just wrong, but that’s the way it’s labeled. No, I don’t remember which one it is. It was years ago.
I prefer plug-ins to act similar to outboard counterparts, but that’s not always the case.


(phoo @ Jul. 20 2007,09:54)
QUOTE
Not all compressors are like that, but the ones that do must be employing some kind of compensation, just as is the default in the fasoft compressor, even if they say they don't.

If I was a cynic I might suggest that they do that to make their compressor sound "better", but I'm not, so I won't.

I agree with phoo on this, it does get confusing after using rack mount compressors all these years.

The problem is simply that the compressor is letting the attack transients through. All you have to do is limit these transients using a limiter.

The longer you set the attack, the worse the problem is. The shorter you set it, the better it should be, but some compressors won’t let you set it short enough to control the snare’s attack – that’s in the range where a “limiter” is expected to work.

I would have expected Phoo to know this, being a drummer, but perhaps he’s just never used real heavy compression on his drums. 10:1 is pretty heavy! And the higher the ratio, the worse this problem gets.

You may find that if you compress enough to squash the attack, you’ll lose the sound you want. Most likely, you’ll want a combination of compressing (to get the sound contour you want in MOST of the snare’s duration) and limiting (to avoid clipping – but still allowing the attack to jump up over the “body” of the snare. Frankly, if you don’t do this (allowing you to lower the snare’s body significantly in the mix), the snare will either have a muddy attack or else will be so overpowering that it will mask all the music happening at the same time.

I suggest that even though you dig the sound of the heavily compressed snare, be wary against overuse of the technique – focusing too much on one instrument will degrade the sound of the others, and snare is particularly noisy thus capable of masking or smearing other detail. However, in a very sparse arrangement this could be very dramatic, somewhat like the heavily reverbed snare in “Bridge Over Troubled Water”.

Something else that occurs to me is that your release setting might be so long that the compressor doesn’t have time to open up before the next beat causes it to clamp down again, thereby giving the effect of turning the volume of the whole track way down. 250 miliseconds is and eighth note at 120 bpm.

Something else: are you compressing the entire drum track or just the snare? If you’re compressing the entire track, then anything on the track - snare, kick, hats, etc., - can cause the compressor to clamp down.

Great discussion. I really appreciate all the input. So, Jeff, to your point, does anyone know how the n-track compressor fits into the effects chain? Is it first or last? Would I need to use a different compressor in the effects windows, or is putting a limiter in simply all that needs to be done?

Captain Damage: I’m compressing the snare only. Everything’s on it’s own separate channel, and I think that it’s set to 200ms.(I think that’s the default for the soft knee?) Besides, the cymbals seem to get a little bright when I compress the whole thing. One other thing I’ve noticed is that the kick doesnt seem to need the same type of compression that the snare does. Anyway, that’s the long trip around your question.

So, I think the next thing to do is to try the limiter after the compressor, and then maybe try a different compressor as well.

The kick will want different compression settings from the snare.

While it probaby has nothing to do with your volume problem, I think 200ms is too long a release time for snare. The envelope for a random snare sample in my collection (extracted from Drumkit From Hell 2) falls below -12dB in less than 40ms. From 100ms to 200ms there is only one peak above -30dB. So I’d try cutting your release back to 100ms - 150ms. But that’s an aesthetic thing - it will change the sound of the snare, maybe in a way you don’t like.

I think before you plug in a limiter you should try a different compressor plug. Some compressor plugins have built-in switchable limiters, BTW.

Also, from your description, I’m wondering if you might try a distortion plugin rather than compression. Give it a try. If you like it, go with it.

<!–QuoteBegin>

QUOTE
I would have expected Phoo to know this, being a drummer, but perhaps he’s just never used real heavy compression on his drums.

Of course know this. My comment had nothing to do with the attacks getting through, but compensation of the output and the fact that many plug-ins aren’t doing what we might think they are doing.

When I set the threshold of a brick wall limiter at -40db I expect the max output to be -40 db (assuming no compensation). When the output jumps up to 0 db no matter what the threshold is I can tell there is compensation going on. That’s not always a bad thing if that’s what you want, but it can be very unexpected.

When I look at a resulting output in a wave editor and I can see that the stuff UNDER the threshold is louder than the original and there has been no gain changes, regardless of the attack, then I know there’s been fiddling with the levels below the threshold, and it’s most likely a raising of the overall level. Gain compensation.

One good sounding mastering limiter plug-in actually has a fixed threshold of 0 db yet the slider labeled “threshold” was nothing more than an overall input gain. The “threshold” is used to adjust how hard 0 db is banged against. Whatever the input it’s adjusted to hit the limiter. That’s automatic gain, not threshold.

Fasoft’s compressor defaults to 2ms attack and 200ms release, but it also defaults to having compensation gain turned on (red stair-step looks like it’s going down button from left to right) and a ratio of 1:1 (no compression). The default threshold is -19.9db.

Stick that on a track as-is, change the Ratio slider and you will see that the volume goes up even if the track’s wave is well below the threshold. To be sure, change the threshold to something like -6dd. Use a track that has nothing above -6db. Changing the ratio enough and the track could go into clipping.

That has nothing to do with the attack nor release. Changing them won’t make a hill of beans in this case.

Hi Gents:
This is a nice discussion going on here… I hope I don’t screw it up…

Anyway, I’ve been working with Snare tracks that need to be controlled with compression and limiting…

Now I’m hoping I’m not promoting brands of plug-ins here… but I’ve been playing with the Kjaerhaus MPL-1 Pro SE Limiter for a while now… I have several Limiters but I find I’m reaching for this limiter, lately…

The threshold control on this one/limiter, seems to work… However, I haven’t been using this limiter with real snare tracks… I’ve been using it with “Sample’d” tracks…

With Sample’d snare tracks you only have limited adjustments/playing around that you can do with that type of snare track… If you start playing around with the dynamics of samples as opposed to a real snare it can get to sound ugly… even though the sample is supposed to be the real thing…

Why is that?

Bill…

Nothing wrong with reccommending a good plugin!

A good plug deserves a good plug :)

Cheers
Jeff


(XonXoff @ Jul. 20 2007,10:25)
QUOTE
compressor is working is very helpful rather than trying to figure out what Flavio's graph is doing.

Actually, the graph helped me to understand exactly what the settings compressor do - the ratio and threshold are clearly defined as characteristics of the graph...