Multiband compression during mastering on dance style music

[eddu]

Ok so i got a decent mix, i can differentiate kick and bass perfectly in the mix and everything sounds fine at the bottom of the spectrum.

However, when i send my mixes to some mastering engineer i find the lows like more tight, more compact maybe, and i am almost sure they did some kind of mb compression.

So...what settings are you used to use when compressing the low band during mastering for dance music (basically prominent kicks on every 1/4 and funky basses or gorgeus basses ala minimoog with some pultec or other coloured eq)?

Thanks

[bob katz]

Originally Posted by eddu

Ok so i got a decent mix, i can differentiate kick and bass perfectly in the mix and everything sounds fine at the bottom of the spectrum.

However, when i send my mixes to some mastering engineer i find the lows like more tight, more compact maybe, and i am almost sure they did some kind of mb compression.

So...what settings are you used to use when compressing the low band during mastering for dance music (basically prominent kicks on every 1/4 and funky basses or gorgeus basses ala minimoog with some pultec or other coloured eq)?

Thanks

Honestly, compression on the bass drum can produce the opposite effect to what's intended, and the loudness race has really killed the danceability of dance music. When a client lets me run a dance master at a lower level it kills the rest of the pack! If you truly find the lows more "tight" (in a good way) and you hear the bass and bass drum perfectly, then you have found a mastering engineer who can take your bass drum and not ruin it or even improve it. That's a rare breed! Chances are you may not have had the bass drum at an optimum level relative to the rest and a lot of what he/she did in mastering was in optimizing EQ as much as compression.

If multiband compression is used, part of the key is attack/release time. The beater of the bass drum is in the upper harmonics, the fundamental is, well, at the bottom! If you're trying to make it fatter, I'd use a fairly long attack time and a medium or fast release time, and I'd try to confine that to the range from, say, 40 to 70 Hz, so as not to screw up the beater. And use a much lighter compression, or even no compression in the higher, beater range. Again, this depends a lot on what you send us!

We listen, if the beater is light, I might even expand and not compress the range from, say, 80 to 120 Hz or higher. Or, if the beater needs more definition, I might try some subtle harmonic distortion, either with an analog or digital-based solution, assuming it doesn't screw up the rest of the sound or make the vocal harsh. It's a compromise in that respect, so if you doubt the accuracy of your bass drum and bass, send us stems so we won't screw up the midrange while we're finding the best sound for your bass drum and bass. Anyway, if the bass drum is boomy, I guess the first thing I'd try is a delicate high pass, since a lot of those club woofers are very "flubby" in that range, and the bass drum definitely can sound tighter if it's boomy or heavy. But this has to be done delicately, since the "boomers" are an important part of the feel of the music, you switch the high pass in and out, and like the eye doctor you say, "better now? or better now?"

So there are really three ranges to be concerned with the bass drum, the subharmonics (the feel, say, below 40 Hz), the boom (40 to 70 Hz), and the beater (which really goes up to about 2 kHz, but in a master you can't touch that range without affecting the vocals and melody instruments, so in general, when concentrating on the bass drum, we might compress or equalize no higher than 120 Hz to avoid that conflict).

I don't know if this helps, it's VERY general, and it's all by the ear. If what comes in sounds very good and needs delicate work, that's the best solution. If the disc can be cut at not too hot a level, that's even better, because compression or bad compression really can kill a kick drum.

[eddu]

Originally Posted by bob katz

If multiband compression is used, part of the key is attack/release time. The beater of the bass drum is in the upper harmonics, the fundamental is, well, at the bottom! If you're trying to make it fatter, I'd use a fairly long attack time and a medium or fast release time, and I'd try to confine that to the range from, say, 40 to 70 Hz, so as not to screw up the beater. And use a much lighter compression, or even no compression in the higher, beater range. Again, this depends a lot on what you send us!

Ok,but if you choose long attack it will let the kick pass thru without compressing then compress the rest (bass) which will probably lead to a pumping effect, specially with short release..am i wrong or am i missing someting?

[AlexLakis]

Using a longer attack time will allow for the initial hit and impact of the bass drum to slide thru uncompressed, while the residual bass gets compressed and/or expanded. This will allow you to retain it's impact, while controlling the overall level of the bass. Some pumping can occur, just play around with the attack/release settings until you hear something you like! Try starting your release settings around 9 o'clock or so.

P.S. Isn't dance music supposed to "pump?"

[eddu]

If i wanna do pumping i do sidechaining compression (kick-bass or kick-everything else). That really pumps everything the way i want. I ´d never had thought about making the song pump during the mastering.

Anyway i think i wouldnt like to loose any part of the bass just after the kick. Imagine a funky bassline with a note one octave up on the 2/16th. Its bottom will be reduced by the mb comp with these settings.

Thanks for the tips

[Tantrum The Cat]

Very cool thread! I've done loads of small-label dance music (mostly commercial house & poppy Euro-Trance stuff), and I'd LOVE to see more house & techno people posting on here!

[bob katz]

Originally Posted by eddu

Ok,but if you choose long attack it will let the kick pass thru without compressing then compress the rest (bass) which will probably lead to a pumping effect, specially with short release..am i wrong or am i missing someting?

Just long enough to let the attack of the beater go through. It's an art... I can't quote a number since each brand of compressor is different and so are the drumsets you feed them. But we're talking about long enough (50, 75, 100, 150 ms, just to give you some ballpark number) to let the transient part of the attack go through. Then you hit it with a medium ratio to help it stay fat.

Yes, if you make the release too short, it may pump. But remember, it's multiband, so it will be hard to pump unless you're trying for that effect. As a rule of thumb, the release should be short enough to fully release between the beats, but not much shorter :-).

[Cojo]

Originally Posted by eddu

If i wanna do pumping i do sidechaining compression (kick-bass or kick-everything else). That really pumps everything the way i want. I ´d never had thought about making the song pump during the mastering.

Anyway i think i wouldnt like to loose any part of the bass just after the kick. Imagine a funky bassline with a note one octave up on the 2/16th. Its bottom will be reduced by the mb comp with these settings.

Thanks for the tips

My experience is that you woun't loose the bass after the kick when using a long attack time if the treshold level also is set right. If so, then the bassline will not be effected by the compression since the treshold level is just reached on the kicks.

/Cojo

[eddu]

Originally Posted by bob katz

Just long enough to let the attack of the beater go through. It's an art...

I thought we were talking about the "boom" band (40-70). the beater was in the next 2 bands, as you said..

[gsilbers]

dance music is kinda compresed in nature, as just by comparing a jazz waveform and dance waveform the dance tune looks like a squarewave

so , fo r mastering, do a lot of dance mixes arrive already pre-mastered(sticking an L2 om the master bus) ?

is that a good / bad thing or depends on how mucha compresio nwas used?


do mastering engineers like to recive 2-3 band stems of a song? (as to better control different freq)


is there any trick for a bigger kick drum when mixing and mastering besides the alrerady menstioned? kick is soooo loud in dance!



=

[bob katz]

Originally Posted by eddu

I thought we were talking about the "boom" band (40-70). the beater was in the next 2 bands, as you said..

Well, there's an overlap. I find that too fast an attack screws up the "boom" band nearly as much as the "beater band".

BK

[bob katz]

Originally Posted by gsilbers

dance music is kinda compresed in nature, as just by comparing a jazz waveform and dance waveform the dance tune looks like a squarewave

so , fo r mastering, do a lot of dance mixes arrive already pre-mastered(sticking an L2 om the master bus) ?

Those are the mixes I often turn back! When the mixing engineer concentrates on getting a good sound and not on the absolute loudness, it's the best collaboration between mixer and mastering engineer. A peak limiter on the 2 bus is a dangerous tool for the mix studio.

is there any trick for a bigger kick drum when mixing and mastering besides the alrerady menstioned? kick is soooo loud in dance!

Great samples! Get great samples. Mike it yourself!

BK

[bob katz]

Originally Posted by AlexLakis

Using a longer attack time will allow for the initial hit and impact of the bass drum to slide thru uncompressed, while the residual bass gets compressed and/or expanded. This will allow you to retain it's impact, while controlling the overall level of the bass. Some pumping can occur, just play around with the attack/release settings until you hear something you like! Try starting your release settings around 9 o'clock or so.

P.S. Isn't dance music supposed to "pump?"

Yeah.. part of the art of the compressor. As long as the impact is obtained through judicious transients the body of the drum can pump and contribute to the "punch". But a punch from a hard fist consists of both parts, you have to have both to have the impact.

BK

[Mike Caffrey]

Originally Posted by eddu

Ok so i got a decent mix, i can differentiate kick and bass perfectly in the mix and everything sounds fine at the bottom of the spectrum.

However, when i send my mixes to some mastering engineer i find the lows like more tight, more compact maybe, and i am almost sure they did some kind of mb compression.

So...what settings are you used to use when compressing the low band during mastering for dance music (basically prominent kicks on every 1/4 and funky basses or gorgeus basses ala minimoog with some pultec or other coloured eq)?

Thanks

They may have, but I'd bet it's some form of EQ. I tend to notch out a few tiny slivers of certain low frequencies to shape the low end the way I want. That definitely makes a tighter sound without affecting the dynamics.

Another possibility is elliptical EQ - going M/S and getting rid of extraneous low in info from the sides.

[gsilbers]

Originally Posted by bob katz

Those are the mixes I often turn back! When the mixing engineer concentrates on getting a good sound and not on the absolute loudness, it's the best collaboration between mixer and mastering engineer. A peak limiter on the 2 bus is a dangerous tool for the mix studio.


BK

interesting.

so, how much headroom would be appropiate for turning in dance music for mastering. ? as opposed to other styles.


you also mentioned somewhere , or hint at , staying away from software compressors. is that because of the harmonic dist. that is added with analog compressors which gives a better sound, or is there something else about software comp?

[Empty Planet]

A big thank you to BK for the detailed info. I'm finding your book most illuminating, but it's just so great to get info like this on GS.

I have still to generalize the lessons I've learned from mixing. I've worked a lot on guitars with this approach (pick attack and the body of the sound as distinct elements, handled differently) but for some idiotic reason I never applied it to kick drums. Duh.

Thanks to all the posters. Very helpful.

Cheers.

[Jules]

I like to solo the bass band, extend it up to the frequncy point of the track I feel is NOT bass, then mess with the attack & release ratios etc.. - what I am often looking for is to find an attack time where the kick attack isnt ruined (so a late attack time)

Then I adjust the mid band up to what I think is the HF zone of the track - and mess with it in solo...

Adjusting the HF band, I often look for a de-essing action, because I am a moron with esses on my vocals.. and they almost always need some taming .

[Mike Caffrey]

Originally Posted by Jules

I like to solo the bass band, extend it up to the frequncy point of the track I feel is NOT bass, then mess with the attack & release ratios etc.. - what I am often looking for is to find an attack time where the kick attack isnt ruined (so a late attack time)

Then I adjust the mid band up to what I think is the HF zone of the track - and mess with it in solo...

Adjusting the HF band, I often look for a de-essing action, because I am a moron with esses on my vocals.. and they almost always need some taming .

Do you find that you end up de-essing the guitars/keys/cymbals or anything else that has HF content?

Have you ever tired this in M/S?

I'm at the point now, where I still do some 2 buss treatments in stereo, but I can barely conceive or working in anything other than M/S for most functions.

[bob katz]

Originally Posted by gsilbers

interesting.

so, how much headroom would be appropiate for turning in dance music for mastering. ? as opposed to other styles.

[/quote}


All of it! :-). Actually, mix till it sounds good! My reasons for recommending at last 2 or 3 db of peak headroom at the top apply to all genres. But the meter isn't the full story. Recommending 2-3 dB at the top is just a safe thing to recommend, it will probably carry through with less unnecessary processing when it gets to the mastering studio, especially if we are going to do any analog processing, as it would feed a D/A converter quite safely, assuming no damage was done on your end prior to the -3 dBFS peak.

you also mentioned somewhere , or hint at , staying away from software compressors. is that because of the harmonic dist. that is added with analog compressors which gives a better sound, or is there something else about software comp?[/QUOTE]

Someday I suspect software compressors will be able to produce the identical sound that we can get from an analog compressor. That day is fast approaching. In the meantime, be cautious. Some of the software emulations seem to distort in a ugly way far earlier than their analog counterparts when you start to use more gain reduction and aggressive time constants (fairly fast release). Some of this is aliasing distortion, if the software compressor does not double or triple or quadruple-sample, or if you are not running at double sample rate. Some of this I think is lack of expertise of the software designer at emulating the characteristics of the RC circuits on the analog side. And some of it may be my fault, inabilty to work by the numbers when I'm trying to get a tricky sound! I think I have the numbers down pat when I'm just trying to subtly enhance something that was mixed well. What separates the men from the boys is the ability to enhance something that was not mixed well but you know can be enhanced with "attitude". That is, the "ultimate mix bus compressor" that the client did not have.

When it comes to "optical" (which is usually gentle and not used for "attitude") I know that great strides have been made at emulating the speed with which an optical compressor returns to unity gain after it's been pushed, and I've heard some very nice optical emuations from PSP and from Waves in the Ren comp. In fact, I'm prepared to say that you can do optical pretty darn well now with the PSP and get a very "warm" sound quite easily now, at least at 96 kHz.

But the more aggressive sounds that you can get from an API or a Cranesong Trakker, I've not heard done as well, I start to hear grunge from the digital versions. They seem to "choke up" earlier than their analog counterparts. And it's a lot harder to keep them "warm and fat" in the aggressive modes, especially faster release times.

Some of my counterparts say the Weiss DS1-MK2 can do punchy and fat just as well as an analog unit, but I have not been able to make it do so, even though I use it for other stuff all the time and I own TWO Weiss units! I do have to try though, in theory, it's all time constants, and it's certainly a very low distortion digital circuit with (theoretically) as fast or slow a reaction time as anyone could want. Same with my TC Electronic System 6000. I have to spend a day with it and shoot it out against the Trakker really trying to get "that sound".

I guess it's because when I'm looking for aggressive, for an "attitude", for hip hop or rock, with only a few minutes time I can get "the sound" quickly with the Trakker, and not hear any ugliness. But I really have to coax it out of the digital versions and I'm not entirely sure if I've found it or fooling myself. Stay tuned on this. It may be possible with the TC or the Weiss and I may not have found the formula. But i'm pretty convinced "aggressive" can't be done with any of the current plugins. I could be proved wrong on that!

[creegstor]

Originally Posted by Mike Caffrey

Do you find that you end up de-essing the guitars/keys/cymbals or anything else that has HF content?

Have you ever tired this in M/S?

I'm at the point now, where I still do some 2 buss treatments in stereo, but I can barely conceive or working in anything other than M/S for most functions.

Mike (or anyone else who knows!). How does one 'go MS' if they're working with samples tracks etc. that are in stereo? Thanks in advance.

[gsilbers]

Originally Posted by creegstor

Mike (or anyone else who knows!). How does one 'go MS' if they're working with samples tracks etc. that are in stereo? Thanks in advance.

check out kvrs webpage for cheap or free utility plugins thatll let u do that

[gsilbers]

Originally Posted by bob katz

you also mentioned somewhere , or hint at , staying away from software compressors. is that because of the harmonic dist. that is added with analog compressors which gives a better sound, or is there something else about software comp?

Someday I suspect software compressors will be able to produce the identical sound that we can get from an analog compressor. That day is fast approaching. In the meantime, be cautious. Some of the software emulations seem to distort in a ugly way far earlier than their analog counterparts when you start to use more gain reduction and aggressive time constants (fairly fast release). Some of this is aliasing distortion, if the software compressor does not double or triple or quadruple-sample, or if you are not running at double sample rate. Some of this I think is lack of expertise of the software designer at emulating the characteristics of the RC circuits on the analog side. And some of it may be my fault, inabilty to work by the numbers when I'm trying to get a tricky sound! I think I have the numbers down pat when I'm just trying to subtly enhance something that was mixed well. What separates the men from the boys is the ability to enhance something that was not mixed well but you know can be enhanced with "attitude". That is, the "ultimate mix bus compressor" that the client did not have.

When it comes to "optical" (which is usually gentle and not used for "attitude") I know that great strides have been made at emulating the speed with which an optical compressor returns to unity gain after it's been pushed, and I've heard some very nice optical emuations from PSP and from Waves in the Ren comp. In fact, I'm prepared to say that you can do optical pretty darn well now with the PSP and get a very "warm" sound quite easily now, at least at 96 kHz.

But the more aggressive sounds that you can get from an API or a Cranesong Trakker, I've not heard done as well, I start to hear grunge from the digital versions. They seem to "choke up" earlier than their analog counterparts. And it's a lot harder to keep them "warm and fat" in the aggressive modes, especially faster release times.

Some of my counterparts say the Weiss DS1-MK2 can do punchy and fat just as well as an analog unit, but I have not been able to make it do so, even though I use it for other stuff all the time and I own TWO Weiss units! I do have to try though, in theory, it's all time constants, and it's certainly a very low distortion digital circuit with (theoretically) as fast or slow a reaction time as anyone could want. Same with my TC Electronic System 6000. I have to spend a day with it and shoot it out against the Trakker really trying to get "that sound".

I guess it's because when I'm looking for aggressive, for an "attitude", for hip hop or rock, with only a few minutes time I can get "the sound" quickly with the Trakker, and not hear any ugliness. But I really have to coax it out of the digital versions and I'm not entirely sure if I've found it or fooling myself. Stay tuned on this. It may be possible with the TC or the Weiss and I may not have found the formula. But i'm pretty convinced "aggressive" can't be done with any of the current plugins. I could be proved wrong on that![/QUOTE]




this topic is pretty interesting. i think others are not jumping in cause its under dance mastering. so i start a new thread with this topic . i hope u dont mind if i add the quote above in the new thread. others can jump in and give other insights.

[gsilbers]

also, for mastering dance do you use unique or different spatial (stero spread etc) techniques?



most dance mixes are mono heavy, cause i think DJ are kinda afraid of club/venues sound system to be screwy by having setups that include only mono source inputs, or the speakers are placed in awkard ways like the right speaker is near the DJ booth and the left speakers are in another room/ambient. so a lot of dance music is mono and then have the extra fx, those that pan lot, sweeps etc to be stereo.
do you get a lot of this type of mixes?

how do you do by spatial treatments on different genres? (generalizing i know but gives an idea)

[Jules]

Originally Posted by Mike Caffrey

Do you find that you end up de-essing the guitars/keys/cymbals or anything else that has HF content?

Have you ever tired this in M/S?

I'm at the point now, where I still do some 2 buss treatments in stereo, but I can barely conceive or working in anything other than M/S for most functions.

I haven't got that advanced yet with my 'junior mastering'.. I presume you mean I could de-ess the mono element in my mix (to attack the central lead vocal) and leave the sides of the stereo (overheads etc) - un touched by HF reduction?

Is there a way to do this in PT?

I also have a TC finalizer 96k whick has some sort of M/S function built in.... but haven't studied it's application much at all.. (my bad) Can the TC box help me?

[JohnNy C]

Originally Posted by gsilbers

also, for mastering dance do you use unique or different spatial (stero spread etc) techniques?



most dance mixes are mono heavy, cause i think DJ are kinda afraid of club/venues sound system to be screwy by having setups that include only mono source inputs, or the speakers are placed in awkard ways like the right speaker is near the DJ booth and the left speakers are in another room/ambient. so a lot of dance music is mono and then have the extra fx, those that pan lot, sweeps etc to be stereo.
do you get a lot of this type of mixes?

how do you do by spatial treatments on different genres? (generalizing i know but gives an idea)

I always mix my music (House, trance, progressive) with heavy panning. And ALWAYS check your mix as you go along in Mono so you know what it will sound like in most clubs that run Mono systems. But honestly, a thick Trance track with heavy mono mixing sounds so boring to me. And all my huge influences in the scene do some great panning in their mixes.

I also feel that if you don't need to use compression when it comes to 4 on the floor music, don't. I try to get my kick sound right form the start. And I have a huge tendency of mixing as I compose. So by the time my track is done, it is pretty well mixed except for a couple odds and ends. I find by working this way, I use less dynamic effects because i really focus my time on getting the right sound WHILE composing rather than after.

[just.sounds]

Originally Posted by Jules

I haven't got that advanced yet with my 'junior mastering'.. I presume you mean I could de-ess the mono element in my mix (to attack the central lead vocal) and leave the sides of the stereo (overheads etc) - un touched by HF reduction?

Is there a way to do this in PT?

I also have a TC finalizer 96k whick has some sort of M/S function built in.... but haven't studied it's application much at all.. (my bad) Can the TC box help me?

Yes there is! the M is the mono sum of 2 the stereo channels and the S the difference. so if you pan the 2 tracks to the middle you get the m and when you apply that out of phase to the original you get the S

to go the other way round you can use the standard m/s mixer configuration
one channel M. split the S to two channels panned left and right (exactly the same level) and put one out of phase. (be carefull wich one because you can swap LR this way) now you can set the levels to adjust stereo width.

http://kmt.hku.nl/~eelco/blok2stereo.html
here you can find an exell page that shows what you do to the stereo image (also handy when you are recording stereo)
there is some other information about what happens to stereo when you feed it to a dolby decoder but that is another story.

Good luck!

[Nordenstam]

Another bass-friendly trick is to high pass filter the sidechain on the compressor. (the "big" button on some HW comps). Half a dB to a DB at 50 to 100 hz is often enough. This makes the comp trig less on the bass end, leaving it slightly less affected by the comp while it's still doing it's magic on the rest of the sounds.

Cudos for JohNyC for mentioning the best trick in the book - getting it right without compression! =)

[Mike Caffrey]

Originally Posted by creegstor

Mike (or anyone else who knows!). How does one 'go MS' if they're working with samples tracks etc. that are in stereo? Thanks in advance.

I use the Dangerous S&M Box. I'm sure there are other ways, but I don't know what they are.

[Mike Caffrey]

Originally Posted by Jules

I haven't got that advanced yet with my 'junior mastering'.. I presume you mean I could de-ess the mono element in my mix (to attack the central lead vocal) and leave the sides of the stereo (overheads etc) - un touched by HF reduction?

Is there a way to do this in PT?

I also have a TC finalizer 96k whick has some sort of M/S function built in.... but haven't studied it's application much at all.. (my bad) Can the TC box help me?

Yes that's what I meant. You can do similar things during mixing as well.

I'm sure there's a way to do it ITB or with the Finalizer, but I don't know what it is.

They key is having inserts in your M/S chain.

[robot gigante]

Originally Posted by gsilbers

is there any trick for a bigger kick drum when mixing and mastering besides the alrerady menstioned? kick is soooo loud in dance!

Here are a couple tips on getting the kick to punch through (mixing not mastering):

-Make sure there's enough 'tic' in the mid/high mid frequencies to let the listener's ear find the kick.... I would search the archives and find out where Dave Pensado talks about it, useful info there. I find that there are a few ways to do it which are all close to the same thing: one would be to parallel compress the kick (nuke it) and use EQ so that one of the kicks is basically midrange tic and blend it with the uncompressed kick, two would be to layer in a ticcy kick sample (highpassed) with the main kick, and three is what some producers that have given me dance stuff to mix do though I don't do it myself- use a high hat tic along with the kick.

If you have the right sample from the get-go that's cool, but having the option to adjust the level of the 'tic' depending on the mix is cool. You don't need very much though, just enough to make the kick pop out a little, not to make the tic sound audible.

-Find out what things are fighting with the kick and sidechain a compressor to duck them a bit when the kick hits- you can have it sound pumpy or not depending on how you set things (sometimes pumpy is good sometimes not). A couple dbx 160xt's on a bus are good for this.


In mastering it seems like I can tighten up the low end a bit with wideband compression, either using parallel compression or sometimes not. I think the mix pumping a little sounds pretty good anyways if you do it right, it's not the same sound as a sidechained comp. I'd bet it's that and/or EQ, not multiband compression.