Substractive EQ is essentially the same as additive EQ?

[Artyomusico]

So here is a thing:
1) create two tracks with the same loop
2) invert the phase of one of them, there are no sound now (they cancel out)
3) now put an EQ8 on each of them

(do not change any parameters other than the gain of a freq)
4)on track 1 use Low shelf and just cut say 2dB
5)on track 2 use Hi shelf and boost 2dB
6)now gain match the track 1 by raising 2 dB on master gain of EQ or by raising the track fader
7)they now cancel each other once again

Any explanation of why does it works this way? Is there actually no difference between the subtractive and additive eq other than in the latter you have to raise the gain.

[cacti]

I think people worry about additive EQing because its ass excess gain to the mix perhaps?

[Angstrom]

I think there are a couple of reasons that subtractive is always taught as the primary method to use:

firstly - because if you boost, the signal is immediately louder, and your brain will tell you it's "better" just because of that. Even thought you may not have helped balance your mix in reality.
So the idea is to cut and then re-level the channel's gain to where you want it, avoiding the psycho-acoustic trickery which will throw you off.

A also tend to think there's a second (historical) reason , which is that additive EQ was a lot more of a problem back when everyone was working with a purely analog signal path. Imagine a multi-miked guitar amp/bass/kit combination running into a mid-range 1980s desk with analogue delay & flange effects.
I have no real idea what the noise floor would be, but I'm gonna take a flying guess at something well above -60db. Relentless boosting on each channel could lead to a situation where noise was increased. Disciplined cutting would help control that.

it's a theory anyway.

[Artyomusico]

I see. It makes perfect sense to me.
But now, imagine we are working with an EQ that at certain amounts of gain adds some nice saturation to certain frequencies. What do we do now? Do we cut so we do not hear any distortion, or do we raise and add some grit to the sound. How do these things really work? I tried to put a saturator with hard clipping and -8dB of gain after an EQ8 and it kinda sounds like a Pulteq (or at least the PSP emulation of it - NobleQ) to me now. But its just a saturation at the end of the chain....duh
Am I missing some magic ingridient, or is it just all about slightly different eq curve shapes (less prononced)?

[spookydirt]

I'm sure I heard somewhere that in nature frequencies only get attenuated, not boosted, so cutting freqs sounds more natural to our ears than a boost would. Which is logical.

[Khazul]

To make sense of this you need to learn a bit about filters and resonance, damping etc - at a guess maybe try looking up something like "filter damping" in google. At Q<=0.5, then boosting is resonance free - this mean that low/high cut is also artifact free and will transparently sum with the opposite filter. At Q>0.5, then you get some resonance at the cutoff, so it wont transparently sum wih its opposite to yield the original signal again.

This is why the basic rule about broad boost and narrow cuts comes in. Narrow (high Q) boost add a lot of resonance. If you want a narrow artifact free (ie no resonance) boost/band pass, then it is possible in a round-about way withg digital filters (such as used by ableton's EQs) using a notch filter/narrow cut as a starting point.

[ARDJ]

you're mixing(pun) up two different steps here.

1.) production - when you're PRODUCING and SOUND designing you're allowed to boost whatever you want. 30 years ago this would be done on a guitar amp, bass amp, etc before it even hit the mixing board. This is the guitarist/producer's call on how bright or what tone the sound has. Boost away, there's nothing wrong with adding frequencies at this stage and experimenting should be don here.

2.) mixing -- ok now you've got the sound how you want it, but how does it sound in RELATION to the rest of your elements. Does it stick out too much? does it compete with your vocals? lead? pads? chords? Where do you want it to sit in relation.

Now because of how music is written these days, you need to ask yourself am I sound designing or am I mixing? Only then will you know what's the most appropriate step.

[starfckr]

you're mixing(pun) up two different steps here.

1.) production - when you're PRODUCING and SOUND designing you're allowed to boost whatever you want. 30 years ago this would be done on a guitar amp, bass amp, etc before it even hit the mixing board. This is the guitarist/producer's call on how bright or what tone the sound has. Boost away, there's nothing wrong with adding frequencies at this stage and experimenting should be don here.

2.) mixing -- ok now you've got the sound how you want it, but how does it sound in RELATION to the rest of your elements. Does it stick out too much? does it compete with your vocals? lead? pads? chords? Where do you want it to sit in relation.

Now because of how music is written these days, you need to ask yourself am I sound designing or am I mixing? Only then will you know what's the most appropriate step.

Absolutely true.

I am never shy of using lots of boost on several bands when designing new sounds. But it`s a completely different thing getting every track in the mix to sit right.

Subtractive EQ is often more relevant when mixing, because subtracting selected frequencies on one instrument, can get another instrument to stick out, without affecting the tonal qualities of either of the instruments. Boosting frequencies in the mixdown will often lead to that instrument being louder. And, it sticks out, yes, but it also tends to be louder in comparison to everything else - so you start putting up the faders on other instruments, and then you have a gain-war.

[Tarekith]

Part of it is psycho-acoustic too. Our ears tend to have an easier time spotting boosted frequencies than when we cut frequencies. So if the goal is to transparently tame some elements so they mix with others in a track, then cutting freqs will tend to be more unnoticed.