Advanced spectrum side chaining

Hi

Now I don't know if this is even possible, but if you don't know what side chaining is, it's essentially taking a away from b. A possibly being a sub-bass line and B being a sub-bass drum. Every time the drums hit, the sub-bass line "ducks" or rather, "gets a lot quieter depending on your settings".

So, it can be really useful, but even if you haven't used it before, you will have heard it before.

I can't remember who uses excessive amounts of it, but it's apparently very popular within the french house dance scene and I think eric prydze or some dickshit like that turned it around on its head so 'every' instrument was being cancelled out by the drum. Cock fuck.

Right so anyway, great feature if you don't use it in the popular shit way, but as I'm a fussy little shit, I am asking for more, and this 'more' relates to my topic title.

Advanced spectrum side chaining

I want to be able to 'take away' A from B, but leave B still going, with *important part* B's standard frequencies, taking away A's clashing frequency with B

So, this bass line is harmonic and it's using many notes. It uses 40 to 45hz and 80 to 100hz. The drum is using only 40 to 42hz, I'd like to be able to have a side chainer (perhaps the built in one within ableton live, or if not, then a 3rd party one) to be able to 'take away' the conflicting frequencies - in this case the 40 to 42hz - instead of just totally lowing the whole volume of the whole spectrum.

If anyone knows more about me in this, or if they know how to accomplish this I'd be very grateful.

use an EQ or something on one, let the sidechain plug on the other control the gain knob on the EQ?

  Adjective said:

use an EQ or something on one, let the sidechain plug on the other control the gain knob on the EQ?

or

get this

  dild0 said:

  Adjective said:

use an EQ or something on one, let the sidechain plug on the other control the gain knob on the EQ?

 

or

 

get this

thanks

not going to be able to use it cus a) can't find a crack or anything, b) it's only 45 quid to get a whole pack of the plugins - though i don't have that cash right now.

OK, here's how you do it (not with FFT):

Say you want to side chain compress sound A against sound B's 100-200hz frequencies.

Send sound B to a bus, twice - invert the phase of one copy 180 degrees. You will hear nothing if it's worked, because the frequencies are phasing each other out. Using an EQ/filter start to boost the frequencies of one copy in the range that you want (100-200hz). If you're doing it right, you should start to hear only the sounds in that frequency range (sort of a ghetto bandpass). OK, this is your source for side chaining - you don't want to hear this sound in the main mix, but you want to use these frequencies to duck the volume of sound A.

Side chain compress sound A against your bus with this phased EQ'd version of sound B.

If this does not make sense to you, it's because you're stupid. Not because I've totally failed to clearly explain it, yeah?

This is some reasonably advanced audio shit. If you're smart you'll work out how to make a stereo spread enhancer like this too.

Learn about Phase.

Uh, hope this works for you?

  esquimaw said:

OK, here's how you do it (not with FFT):

 

 

Say you want to side chain compress sound A against sound B's 100-200hz frequencies.

 

Send sound B to a bus, twice - invert the phase of one copy 180 degrees. You will hear nothing if it's worked, because the frequencies are phasing each other out. Using an EQ/filter start to boost the frequencies of one copy in the range that you want (100-200hz). If you're doing it right, you should start to hear only the sounds in that frequency range (sort of a ghetto bandpass). OK, this is your source for side chaining - you don't want to hear this sound in the main mix, but you want to use these frequencies to duck the volume of sound A.

 

Side chain compress sound A against your bus with this phased EQ'd version of sound B.

 

If this does not make sense to you, it's because you're stupid. Not because I've totally failed to clearly explain it, yeah?

 

This is some reasonably advanced audio shit. If you're smart you'll work out how to make a stereo spread enhancer like this too.

 

Learn about Phase.

 

Uh, hope this works for you?

that's fucking brilliant, black magic!!

post here moar plz.

  esquimaw said:

OK, here's how you do it (not with FFT):

 

 

Say you want to side chain compress sound A against sound B's 100-200hz frequencies.

 

Send sound B to a bus, twice - invert the phase of one copy 180 degrees. You will hear nothing if it's worked, because the frequencies are phasing each other out. Using an EQ/filter start to boost the frequencies of one copy in the range that you want (100-200hz). If you're doing it right, you should start to hear only the sounds in that frequency range (sort of a ghetto bandpass). OK, this is your source for side chaining - you don't want to hear this sound in the main mix, but you want to use these frequencies to duck the volume of sound A.

 

Side chain compress sound A against your bus with this phased EQ'd version of sound B.

 

If this does not make sense to you, it's because you're stupid. Not because I've totally failed to clearly explain it, yeah?

 

This is some reasonably advanced audio shit. If you're smart you'll work out how to make a stereo spread enhancer like this too.

 

Learn about Phase.

 

Uh, hope this works for you?

Where did you learn this from and I don't understand when you mention a few words, such as 'Send sound B to a bus' and 'phase of one copy 180 degrees' 'ghetto bandpass'.

Also what do you mean by 'stereo spread enhancer' and where can I learn about Phase.. do you mean the standard effect of phasing?...

'http://en.wikipedia.org/wiki/Phaser_(effect)'

Thanks. Also, "you're stupid if you don't get it" is unprofessional on your side, I don't get it because the terms you use I do not know.

  esquimaw said:

OK, here's how you do it (not with FFT):

 

 

Say you want to side chain compress sound A against sound B's 100-200hz frequencies.

 

Send sound B to a bus, twice - invert the phase of one copy 180 degrees. You will hear nothing if it's worked, because the frequencies are phasing each other out. Using an EQ/filter start to boost the frequencies of one copy in the range that you want (100-200hz). If you're doing it right, you should start to hear only the sounds in that frequency range (sort of a ghetto bandpass). OK, this is your source for side chaining - you don't want to hear this sound in the main mix, but you want to use these frequencies to duck the volume of sound A.

 

Side chain compress sound A against your bus with this phased EQ'd version of sound B.

 

If this does not make sense to you, it's because you're stupid. Not because I've totally failed to clearly explain it, yeah?

 

This is some reasonably advanced audio shit. If you're smart you'll work out how to make a stereo spread enhancer like this too.

 

Learn about Phase.

 

Uh, hope this works for you?

this not solves his question. he want to duck only the overlapping frequencies between audio A and audio B

joshier,i don't think you can do that in ableton. the only way to do it is to control some highpass filter frequency knob or the gain of a bell with a sidechain plugin (as adjetive says) and this is possible only in FLstudio.

Yo, there's a sidechain on the autofilter in live 7, it will do exactly what you want

  joshier said:

  esquimaw said:

OK, here's how you do it (not with FFT):

 

 

Say you want to side chain compress sound A against sound B's 100-200hz frequencies.

 

Send sound B to a bus, twice - invert the phase of one copy 180 degrees. You will hear nothing if it's worked, because the frequencies are phasing each other out. Using an EQ/filter start to boost the frequencies of one copy in the range that you want (100-200hz). If you're doing it right, you should start to hear only the sounds in that frequency range (sort of a ghetto bandpass). OK, this is your source for side chaining - you don't want to hear this sound in the main mix, but you want to use these frequencies to duck the volume of sound A.

 

Side chain compress sound A against your bus with this phased EQ'd version of sound B.

 

If this does not make sense to you, it's because you're stupid. Not because I've totally failed to clearly explain it, yeah?

 

This is some reasonably advanced audio shit. If you're smart you'll work out how to make a stereo spread enhancer like this too.

 

Learn about Phase.

 

Uh, hope this works for you?

 

do you mean the standard effect of phasing?...

 

no. he doesn't.

abort. abort.

esquimaw's way would work if you did it just right... EQ and filtering in software involves discrete fourier transforms no matter how you do it, it's just transparent to the user.

If you can't flip the waveforms, and you have a good filter or EQ you can also just:

1. Take your source sound. Make 1 copy of it. Now you have sounds A and B

2. Bandpass-filter A so that it blocks out the range you want to sidechain

3. Notch-filter B so that it only plays the range you want to sidechain

4. Sidechain-compress B

i.e, in Reason, you could just make two NN19 samplers with the sound, use the vocoder EQ as a makeshift BP/Notch filter, and use the compressor to sidechain the second sampler.

Using VST's, just find a good filter that lets you specify the range of BP/Notch, or use an EQ with a good resolution of frequencies

  esquimaw said:

OK, here's how you do it (not with FFT):

 

 

Say you want to side chain compress sound A against sound B's 100-200hz frequencies.

 

Send sound B to a bus, twice - invert the phase of one copy 180 degrees. You will hear nothing if it's worked, because the frequencies are phasing each other out. Using an EQ/filter start to boost the frequencies of one copy in the range that you want (100-200hz). If you're doing it right, you should start to hear only the sounds in that frequency range (sort of a ghetto bandpass). OK, this is your source for side chaining - you don't want to hear this sound in the main mix, but you want to use these frequencies to duck the volume of sound A.

 

Side chain compress sound A against your bus with this phased EQ'd version of sound B.

 

If this does not make sense to you, it's because you're stupid. Not because I've totally failed to clearly explain it, yeah?

 

This is some reasonably advanced audio shit. If you're smart you'll work out how to make a stereo spread enhancer like this too.

 

Learn about Phase.

 

Uh, hope this works for you?

I just wanted to mention that if you have Ableton Live you actually don't need to do the phase inversion, you can just bus to a bandpass, disable the bus (audio is still computed), and use the bus as a source in the side chaining. Also, the side chain feature is on all the new compressors in ableton live 7, just gotta unhide the tab.

  acid1 said:

Also, the side chain feature is on all the new compressors in ableton live 7, just gotta unhide the tab.

I know i've already mentioned it, but they've also included a sidechain on the autofilter and gate, for those not already in the know.

  joshier said:

Also what do you mean by 'stereo spread enhancer' and where can I learn about Phase.. do you mean the standard effect of phasing?...

 

'http://en.wikipedia.org/wiki/Phaser_(effect)'

 

Thanks. Also, "you're stupid if you don't get it" is unprofessional on your side, I don't get it because the terms you use I do not know.

That was me making a lame joke about how my explanation was not particularly clear.. I was a bit hung over when I typed it. I don't think a silly tone of voice comes across well in plain text. This is (reasonably) advanced shit and I explain like a moron. Sorry dude!

Here's some links:

SOS synth secrets: filters and phase

Wiki - Phase (waves)

Wiki - Phase cancellation

Having re-read your original post, I don't think my solution solved your problem anyway.. I described how you might make 1 band of a multiband compressor, sort of.

If you have overlapping frequencies... The louder one will always be heard over the quieter one anyway, as long as they're coming from the same speaker. If they are 'in phase' the volume will just be louder.

This is the point where you have to realise that you're dealing with physics.. Sound is waves. This is probably not something I have the time to properly explain here. There's plenty of books/articles been written on the subject.. Go discover!

  Amen Warrior said:

Yo, there's a sidechain on the autofilter in live 7, it will do exactly what you want

it does exactly what we tried to say 'till now,excellent!

*goes legally upgrading ableton live to version 7*

i have a decent grounding in science, i knew what phase was. as a lad, i was placed in front of a pair of sine wave generators hooked up to a speaker and an oscilloscope, and that was my first hands-on lesson with it (rotating circles :). i just never really considered it when making music. it was a in different part of my brain. it HAD occurred to me to look into it here and there, i just never got around to it.

a lot of what i do has to with feeding fx back on itself, and as i've been reading this thread, a lot of rad ideas have been popping into my head. of course i'm visiting someone now, and my gear is miles away...

ekt needs more of this. class thread

here is an ableton rack that you could experiment with to achieve what you want, it uses esuimaws method to spilt the signal into low and hi, then has a side chain compressor on the bass. Gives very clean sounding ducking with the right settings.

http://tinyurl.com/qevsbdrtvhsrtvhrthrvthwrthrthv

my take:

you take an envelope follower (the first stage of a compressor) and use its output to control (negative) gain of the EQ, instead of controlling an attenuator, which makes a compressor.

that means: the output of the envelope follower will represent the loudness of the bassdrum signal, which will make the bassline's EQ (manually set to bassline's frequencies that clash with the bassdrum) cut the appropriate amount of signal.

you then connect the signal A (bass drum) to the envelope follower and signal B (bassline) to the EQ. now the bassline "ducks" the bassdrum just in the frequency range set by the EQ.

in the end, just mix the two signals (unmodified bassdrum and modified bassline)

it should work and it could be described kinda like an inverse vocoder :)

actually, if you wanted a completely automated solution (no manual EQ setting), you'd need a (multi-band) vocoder, that ATTENUATES the carrier signal instead of AMPLIFYING it.

that means you can do it all pretty simple and completely in analogue realm (no fourier required)

im really mad at myself i ignored this thread!

when i was more into doing conceptual experimental music stuff i had this idea of doing subtractive (as opposed to additive) real-time FFT processing of two sounds A & B controller one another, this sounds very similar to what you are asking for.

So far i havent really found anything capable of this with high quality amount of bands like 512 besides Max/msp. I tried doing it with Plogue Bidule for the longest time but never could figure out how to reverse the math.

anybody with max/msp skills willing to take a crack at this? i'd love to be able to use something that does this.

i never even thought of the utilitarian aspect of how a filter like this could be extremely useful for ,mixing until i read joshiers idea.

kokoon's breakdown of what needs to be done is perfect, although i personally would be coming at it from a more FFT convolution direction. From another perspective i would say this filter is basically inverse convolution