When To Use Multiband Compression

Hey,

Regarding optimizing the bass frequencies and overall level (perceived loudness) on a project...

If I can make the bass sound and behave like I want during mixing of the tracks (handling compression on those tracks themselves), is there any compelling reason to add multiband compression on a final stereo mix? Seems at that point it can act as a four band EQ, if that would benefit the mix.

Definitely a multi-band com can act as a dynamic EQ. Effectively, since you can reduce particular frequencies based on how prominent they are at any time, not in a fixed way as with a regular EQ, it really does act like a dynamic EQ more than a compressor. It definitely changes the tonal characterstics of the sound based on it's level. So it's really a different kind of beast than a single band compressor (though some single bands also allow you to EQ the side band which provides kind of similar capabilities.)

I've been reticent to use them, because it's soooo easy to make a mess of things. But I'm slowly finding good uses for them. One recent one that worked out swimmingly is on a snare. Snares often seem to have extremely different characterstics based on how hard you hit them. EQ them so that they sound snappy when hit at moderate volume and they can be ear splitting on a nice hard hit, and vice versa. So I put a multi-band on it with fairly substantial gain reduction above 4K and below about 250, with a little less from 250 up to about 500. This worked out really well. The harder you hit it, the more it pulls down on the high end and low end, so lighter hits get plenty of body and crack, while hard hits don't sound tubby or ear harsh (also used a soft knee to soften it as it compresses more, i.e. to adjust the punch downward, not upward since I'm trying to get rid of punch as it gets louder.)

People use it in a similar way for, say, heavy distorted guitars. You want to pull out a lot of low end on them, becasue they can be huge on the bottom and sit all over the bass and anything else. But, according to how you are playing them might have more or less low end content, i.e. a palm muted part might have different low end than a slamming out chords part might. So put a multi-band on it and clamp down on the lows more and more as they are present more and more. You still may roll it off some with EQ as well, but there again it lets the less low endy parts have more, so that they don't sound thin, while clamping down on the parts that ahve too much.

So, anyway, in that sort of application I think that they are pretty dang nice, and not too hard to get right. On the mix buss it's a lot tricker. I do sometimes use one, but I keep it light. I've sat there for hours tweaking and just making things worse and worse. But with a little bit of gain reduction to just smooth things out, it's pretty useful.

I definitely always try to get it right in the mix. It might be hard though to necessarily get the same effect in the mix, since with the multi on the mix bus you are basically dynamically adjusting all of the instruments that happen to have content in a particular band. That would be pretty hard to do in the mix itself. So there might be things you could do a lot easier with the multi on the mix buss. But I think that the goal would always be to get it right in the mix, for those of us who can iteratively mix and master until we get it the way we want.

I would try and tweek using eq, and a standard compressor/limiter etc as opposed to using a multiband compressor.

I've never been a big fan of multi-band compressors...especially when you are listening on critical systems that can reproduce the full spectrum of frequencies with low driver compression (for example I've got a pair of JBL TR125's for my home stereo that I listen to music for pleasure on that is paired to a stereo arrangement of 10" subs powered all by a more than ample 300wpc mixer and it is dead easy to tell the difference between well mastered/compressed tracks and those that cram every ounce of power into the last 3 db's in sacrifice of dynamics...).

Multi-band compression will always sound better on crappier systems because it becomes difficult to discern what is natural compression resulting from the speaker driver's own incapabilities and the compression that you are applying on purpose. just my 2 cents.

Very interesting, and informative, thanks for the info Dean.

lets_split

Multi_band_workshop


From what I've gleaned from reading post from M.E.'s , MBC's are a tool of last resort. If you find problems that significant , perhaps you should go back to the mix and see what you can do there. Why can't you get the results your looking for there??

Cheers

Hey,

I've been experimenting by compressing the lowest bass band (about 120 and down I believe) a little bit and raising that output of that band until it has the punch that I like. I'm not touching any of the other bands, but will be trying the same approach on the highs just to see what happens.

For those of you who have used MBC on a final mix, are there any general procedures that you take?

ORIGINAL: doncolga

Hey,

I've been experimenting by compressing the lowest bass band (about 120 and down I believe) a little bit and raising that output of that band until it has the punch that I like. I'm not touching any of the other bands, but will be trying the same approach on the highs just to see what happens.

For those of you who have used MBC on a final mix, are there any general procedures that you take?

I do pretty much the same thing (in Ozone), exept I use three bands, the highest starting at 8000-12000 Hz. I compress the first band to get the right punch like you describe, then I compress the third band with a 2:1 - 4:1 ratio and add a small amount of makeup gain to squeeze out some sustained, top-end 'air'. I've not to date been able to achive any positive effects by compressing the middle band, so I leave that open.

I've also found that I have to tweak the attack time of the lowest band comp to somewhere between 1 and about 20 msec, depending on the material.

Sven

Sven

Dean, I really appreciate your sensitive input.

bass frequencies and overall level (perceived loudness) on a project...

I always use MBC at the mastering level, but only after applying general (1) EQ and (2) compression and limiting.

MBC (in Ozone3) helps me

1) Add dramatic soft/hard-knee limiting where necessary on up to 4 Q(s)
2) Add dramatic compression and, hence, 'orchestral' balance of octaves on the 4 Q(s).

I always employ EQ and General Compression/Limiters first for a generalized optimization, only.

1) if I blow off MBC, I blow off 50%+/- of sensitive 'orchestral' optimization (to my ears).

2) methinks, EQ and GC cannot gait out subtle 'arbitrary' noise as nearly as effectively as a good MBC

3) Compression is so fine-tuned with MBCs, you really get a more dynamic diversity at various freqs.

Hey,

Thanks for the responses folks. I have really enjoyed experimenting with the MBC and really feel like I've got a super powerful new tool that can easily sculpt the overall sound of a project. I'm finding that I can easily fatten up a song with even very *subtle* use of the MBC. Squeezing the bottom a little, then dialing in the punch that I want with surprisingly small adjustments; its very easy to pull back or add as needed in a broad sense, which I personally like. Same thing on the high end; even as little as a 1 dB increase, it's really easy to add air without using EQ. I'm not even touching the middle two bands. This is a big adjustment my mixes really needed...better control of the bass and bring out air.

This seems to be the best my mixes have ever compared to commercial projects. I didn't put in a high pass filter to chop off really, really deep stuff, so I'm looking forward to going back and adding that component.

Donny

ORIGINAL: AJ_0000

ORIGINAL: j boy

The problem with multiband compresion is the (unavoidable) amplitude distortion at the crossovers.

The Waves Multi Dynamics plug-in claims to eliminate that issue. I suspect the same is true of the LP64 in Sonar, which is what I've been using. I don't hear any artifacts at all, and it sounds considerably more transparent than any of the other (single-band) compressors I have at my disposal, although I accept that I could be missing something.

AJ,

I think some of the crossover distortion is not an issue if:

1) Crossovers are judiciously chosen for instruments
2) DCB's of compression are incremented modestly btw bandwidths
3) Vox's are completely compressed beforehand ... and given wide bandwidth(s) where feasible
4) Expander/Gating of lower dcbs is employed (per bandwidth) as necessary.
5) You know your problem bandwidths and your excellent bandwidths that you wish to optimize.

That is excellent! I've found myself unconsciously doing like you; compressing the lows and adding just a little air (volume) to enhance the highs