Subtractive EQ BS

[UncleBob]

Something I regularly see with regards to clean digital eqs is that this one or that one "boosts poorly". A long time a go I tested this for myself, here's the method:

Take a loop (preferably something with a lot of high end info)
Copy it to two tracks
Put the same eq on both tracks
On the first track put an eq and boost a high shelf by 10 at 5khz Q1
On the second track put a low shelf, cut 10db at 5khz Q1
THEN turn the second tracks fader up by 10db

The result should be pure cancellation if you pan the tracks hard left and right in mono, flip the phase on one and sum them.

Obviously this won't work with any analog emu plug or eq's with funny q-curves, but try it out and let me know if my methods are screwed and some stock clean digital eq plugins really "boost poorly".

You should get the same result, as I did (with Fabfilter Pro-Q) boosting the full 30db with the one, and then cutting 30db in the other and making it up with 2 instances of Sonalksis FreeG.

 

[SoundsLikeFog]

If the coder did their homework, you are correct.

I always thought that the "boosts poorly"-arguments are suggesting, that most digital eq's sound like shit when boosting.. Not, if the math is broken or not.

 

[Nimvi]

I always thought that the "boosts poorly"-arguments are suggesting, that most digital eq's sound like shit when boosting.. Not, if the math is broken or not.

This is how I interpret it too. I use ReaEQ for all my clean EQ needs, but when I think that I need a boost, I am usually much happier with VibeEQ or something similar. So I could imagine someone saying "ReaEQ boosts poorly vs VibeEQ" in that situation, even though ReaEQ does exactly what I tell it to do. I agree it is a bit of an ambiguous choice of words.


Btw, I expected this thread to be about using subtractive EQ as much as possible instead of boosting... now there's a theme to get a heated discussion going
Obviously boosting is for pussies. Real men don't build a house. They cut away all the ground that doesn't look like one.

 

[UncleBob]

If the coder did their homework, you are correct.

I always thought that the "boosts poorly"-arguments are suggesting, that most digital eq's sound like shit when boosting.. Not, if the math is broken or not.

I always just thought it was odd that you should "always cut rather than boost", when there is zero difference in sound quality either way.

I remember the differences between different eq's (one track with ReaEQ and the other with Pro-Q) was always negligible as well.

 

[bryan_kilco]

Say you have a raw kick drum that needs some attack around 5-8khz.

You can't cut away low end to make that attack magically come out.

I've posted a thread yesterday about EQing and comping raw drums. Was curious as to what was "right" and "wrong", even though there are no rules....

 

[mva801]

The always cut rather than boost rule is complete and utter bullshit. EVERY knowledgeable engineer I've seen come through our studio fires up the API 550s and immediately starts adding 5 and 10k. Do whatever sounds good.

That said, different eq's definitely have different characters. adding a shelf at 10k on APIs definitely sounds different than doing it with the pro tools eq.... My guess is that people just don't like the character of the digital eq's quite as much. Which I can understand. Doesn't mean they aren't doing their job exactly like they're supposed to.

 

[~BURNY~]

It's a valid practice on shit prosummer grade mixers and EQ's where boost will cause more noise due to poorly designed amp stages.

On a high end Analog EQ, boosting will get you all the nice coloration, harmonics warmth and meat.

 

[mva801]

^^^ Which brings me back to do whatever sounds good. If you boost and it sounds like crap, don't do it. That's the crappy equipment's/ signal chain's fault, not eqing in general. I've just met a few people (usually inexperienced) who have it in their heads that you should never really boost an eq.

 

[Kohugaly]

argument [+10dB highshelf = -10db lowshelf * 10db gain] is not perfectly valid. It depends on the algorithm the EQ uses. If it's simple biquadratic filter with symmetric curves, than yes, they will be the same, but many EQs (like waves renaissance EQ) use different curves for boosting and cutting or use different algorithm from the very start. For example State Variable filter splits incoming signal into Highpass, bandpass, lowpass and lets you control each part separately. In that case boosting Low is not the same as cutting High and boosting gain.

My phylosophy of EQing is, that when removing (cuting) something, you what to do it with minimal artifacts (clean digital EQ). When you need to boost, it means that something is missing in the first place, so I use colorative EQ in that case, that also adds something extra(Usually ThrillseekerXTC by VoS). Clean digital EQing is simply frequency-dependent multiplication. If the sound isn't there in the first place zero*anything=zero (more precisely in case of EQing real signal "= more noise").

 

[The StabbinCabin]

I try to only use dramatic boosts with drums. The way I feel about it is that it makes more sense to take away what junk the mic picked up rather than to digitally add information to what isn't there. If I had a good hardware eq I'd be more open to dramatic eq on whatever. But IME, after focusing on making as many cuts as I feel I need and switch back and forth between mono/stereo- my mixes are way more open and coherent than from when I was boosting all over the place. This is excluding drums as mentioned.

I still like SSL/API type plugs for gentle boosts after I've carved away at things and do sometimes boost like 6-8db for added air. But I don't feel like it's bullshit to focus on cuts first for things like guitar/bass/vocals. That approach works best for me, especially when mixing in mono.

 

[Nimvi]

I still like SSL/API type plugs for gentle boosts after I've carved away at things and do sometimes boost like 6-8db for added air. But I don't feel like it's bullshit to focus on cuts first for things like guitar/bass/vocals. That approach works best for me, especially when mixing in mono.

Agree, I do the same thing and I'm quite happy with it. I think the whole trap with cutting vs boosting for newer engineers is that you have to develop an understanding of when what is required.

If you cut away all the gunk you don't want and there is obviously a lack of a certain frequency somewhere that would push that instrument to a more desirable position, I can't think of a reason not to exploit that.

The problem starts when a new engineer is boosting like crazy if he can't hear the individual instruments well because of severe masking and mud, which probably just results in more clashing and unappealing sounds. I think that is where that argument came from, but that's not a flaw of the EQ.
So it should probably be changed from "cut before boost" to "learn when to cut and when to boost".

 

[Mikaël-ange]

Cutting frequency over boosting isn't bullshit, and both method work.
Let's say a track have too much mid for you:
You can cut the mid out
You can boost low and high

That for the sonic stand point.

Now what most of you guy seem to didn't realize (could be wrong) is that when boosting a frequency, you reduce your available headroom inside the eq, so more chance to distort something.

The beauty of digital is that we have so much headroom to play with so that not really an issue anymore. But on the analog world that an other story.

One interesting thing (I least for me) I experiment a lot now is to try to cut frequency that can mask what I want to bring out of the track first, and boost what I want to emphasize second. Less boosting at the end so less phase shift.

Mix tend to become smallish sounding and harsh when over eq ed... Think about that...

 

[UncleBob]

Cutting frequency over boosting isn't bullshit, and both method work.
Let's say a track have too much mid for you:
You can cut the mid out
You can boost low and high

That for the sonic stand point.

Now what most of you guy seem to didn't realize (could be wrong) is that when boosting a frequency, you reduce your available headroom inside the eq, so more chance to distort something.

The beauty of digital is that we have so much headroom to play with so that not really an issue anymore. But on the analog world that an other story.

One interesting thing (I least for me) I experiment a lot now is to try to cut frequency that can mask what I want to bring out of the track first, and boost what I want to emphasize second. Less boosting at the end so less phase shift.

Mix tend to become smallish sounding and harsh when over eq ed... Think about that...

It's more from a technical plugin point of view, as in, boosting somehow causes the plugin to degrade the audio more than if you had to cut.

 

[Ermz]

The general point behind the 'digital EQs boost poorly' thing is not about whether you achieve the result through additive or subtractive means, but rather about a quality analogue equivalent of similar utility often sounding more euphonic in contrast to the plug-in. This is easily demonstrated in any studio with a sizable array of outboard gear.

 

[UncleBob]

The general point behind the 'digital EQs boost poorly' thing is not about whether you achieve the result through additive or subtractive means, but rather about a quality analogue equivalent of similar utility often sounding more euphonic in contrast to the plug-in. This is easily demonstrated in any studio with a sizable array of outboard gear.

I know, I wasn't really talking about methodology or analog versus digital, more just about when guys start carrying on about how one clean digital eq plugin boosts so "musically" over another, and how another clean digital eq boosts poorly so they prefer to only use it for cutting...........it's pretty ridiculous.

In other words, for example, the Fabfilter Pro-Q can boost 30db with absolutely the same result as cutting 30db, it's not like boosting with it somehow opens some nyquist rip in the space time continuum folding back so you have digital aliasing and modulation at 12khz completely destroying your audio universe..............as some over at the "blue" forum would have you believe.

With regards to any digital eq plugin (that doesn't have any funny filters) there's no difference between boosting or cutting.

 

[Ermz]

I agree, there's no reason there should be. It's all math. So long as the gain structuring is equivalent there should be no difference, unless the plug-in is coded to give some kind of harmonic break-up or saturation through boosting.

 

[Kohugaly]

exactly... it's all in the head... I believe this phenomenon comes from how human interacts with the plugin. I mean one GUI may lead you to create better sounding boosts easier thanks to way controls are layered and ranged, although the DSP algorithm is exactly the same.

 

[Melb_shredder]

90% of the time I'm making cuts. But there's a little boosting here and there on my mixes, but if I want anything above a tiny bit, ill throw on my outboard EQ. The difference between any of the analogue eq I use vs. digital for boosting is a noticeable one in favor of the analogue one all the time! I just don't like what the digital eq does when boosting frequencies. That cleanness is probably to the detriment of the digital eq in that particular instance

 

[Trevoire520]

One of the main reasons for the "cut don't boost" way of thinking is that louder sounds better to the ears, so often you can boost something and percieve it as an improvement when it's actually not. This is especially prelavent in newbie mixers.
Also remember the fact that most newbie engineers are recording their tracks too hot anyway, so any significant boosts are likely to push the tracks into clipping.

Personally I do a combo of cutting and boosting, just have to gain stage properly so that the levels is matched to that of the bypassed plugin so you can make informed decisions on whether you're actually making the right judgement.

 

[UncleBob]

One of the main reasons for the "cut don't boost" way of thinking is that louder sounds better to the ears, so often you can boost something and percieve it as an improvement when it's actually not. This is especially prelavent in newbie mixers.
Also remember the fact that most newbie engineers are recording their tracks too hot anyway, so any significant boosts are likely to push the tracks into clipping.

Personally I do a combo of cutting and boosting, just have to gain stage properly so that the levels is matched to that of the bypassed plugin so you can make informed decisions on whether you're actually making the right judgement.

Or you can be lazy and use this like I do, it's great for level matching pretty much anything:

http://www.kvraudio.com/product/mcdodo_s_super_volu_by_dodo_bird