Dynamic sidechain EQ?

[driven]

Had one of those late-night ideas, wanted to see if it's a solved problem before putting in too much research/experimentation time.

Is there a way to do dynamic, sidechain EQ so that it pulls frequency from target tracks specified by a spectrum analysis of the triggering track? (Rather than pulling "Hi" or "Lo" in bands, it would pull what is needed, adjusting EQ on the fly.)

FYI, this (very nice!) video describes how to do a sidechain EQ that relies on Hi and Lo bands, but I want the frequency pull to be determined on the fly -- seems way more elegant and precise. Or is the Hi- and Lo- band approach completely Good Enough in practice?

http://www.youtube.com/watch?v=TSIVVxoxEqM

[lvehon]

Here's a few things to try. You will need max for live for the first one. Duplicate your track and put a bandpass filter on the duplicate, or make a rack that's similar. Center the bandpass around the frequencies you are having trouble with, you may even want to boost them. Turn duplicate track audio off. Put an envelope follower (Max) on the same duplicate track and map it to the gain knob in eq8 that is around the problem frequencies. Adjust the envelope follower to get the right amount of reduction/boost. You might also want to play with the eq in the side-chain section of the new live 9 compressor. Set side-chain input to the same track then turn the eq on and go from there. There's always the good old multi-band compressor as well.

[lvehon]

http://www.soundradix.com/products/surfer-eq

[driven]

Phrased another way, I want to take the Spectrum of one track and carve that shape out of another track, on the fly as it changes -- without me having to manually identify which frequencies to massage, which may change over time. I'm not even 100% sure this is a good idea, as it may take too much decision-making out of my hands.

Yep that Surfer EQ looks like exactly what I want -- but dang $200. Seems like this should be possible within the tools available within Ableton Live (Suite). Maybe not, gonna dig a bit..

I know nothing about Max for Live. Maybe this is a reason to dig into it.

[thisissami]

@driven - did you ever figure anything out to your liking with this? I'm curious about the same things that you were asking about a number of years ago.

[Stromkraft]

@driven - did you ever figure anything out to your liking with this? I'm curious about the same things that you were asking about a number of years ago.

The envelope follower solution above works, but of course there is very little intelligent spectrum analysis going on. It needs to be set up properly and still is a little crude.

Surfer's EQ requires iLok 2 (hardware dongle) so that's out for me and many others.

Real-time side-chain input dynamic EQs and similar does the job and there are some alternatives these days. Like for instance:

• TDR Nova GE
• MeldaProduction MDynamicEq and MAutoDynamicEq
• Voxengo GlissEQ
• Blue Cat's Dynamics plugin & Stereo Parametr'EQ stereo equalizer
• HOFA IQ-EQ V3
• And I believe FabFilter ProQ

And there's also solutions like Wavefactory's Trackspacer and Izotope Neutron.

I don't see any native solutions, but I wasn't aware of the Envelope Follower until a few years back. I've used this for bass ducking, but the analysis set up falls on the user as is true for most of the dynamic plug-ins too but those are more versatile I think.

[Khazul]

I use fab filter pro-mb for this kind of thing.

Its a very nice multi-band compressor / dynamic EQ with a side chain input.

A common use for me is sticking it over an instrumental sub mix to keep a vocal clear (vocal feeds the side chain). Gives lots of control and nice UI that gives lots of visual feedback making it easy to use.
Its also a great multi-band compressor for post-mix processing as well - again the UI and they way it works make it very easy to use compared to most multi-band compressors (which typically cause more trouble than they fix if you are not used to them).
As a dynamic EQ - its good enough for de-essing duty too.

Previously I have made an effect rack using only live plugins to do ducking in a frequency range - it worked quite well for the typical vocal over instrument sub mix type job - certainly good enough to not bother spending a fortune on good side chain dynamic eq plugin.

The basic idea is you use an EQ to band cut a part of the audio range. The output of that is inverted (utility plugin - phase invert) and summed (via effect rack) with the original audio. Now you have two versions of the audio - everything except the band cut part and the other is just the band cut part. They should sum to perfectly reproduce the original audio (if they don't, check EQ modes and reduce the Q factor in the EQ). Stick a side chain compressor in the EQ chain. Feed side chain from vocal and tweak as needed. You must keep the Q factor on the EQ below 0.7 to prevent any non linearity. Also do not use the high quality EQ mode as it will not sum perfectly (last time I tested - maybe this has been fixed).

Just be careful with the compressor settings as the result can sound awful - err on the side of subtle.