The No.1 Website for Pro Audio
 All  This Thread  Reviews  Gear Database  Gear for sale     Latest  Trending
Subtractive EQ
Old 28th January 2008
  #1
Lives for gear
 
drumzealot's Avatar
 

Subtractive EQ

I always feel like I have to preface questions by pointing out that I have searched for info. I found a number of comments scattered among other threads, but I think a thread dedicated to this topic would be very helpful.

I’m not primarily looking for advice regarding what I do (although that is welcome). The primary goal of this thread is to gather various opinions and especially practical info about how and why you use subtractive EQ. Step-by-step descriptions would be very helpful!
I don’t have a control room in my home, which places obvious limitations on how much EQ I use while tracking. Plus I usually play everything myself and record one instrument at a time. After tracking I listen to the whole thing in mono, do a fast rough mix by adjusting the levels, and then EQ each track individually with the other tracks muted. Sometimes, I just jump right to EQing the individual tracks.

First, I’ll assess the need for compression on each track.
Second, I use a HPF and try to find a spot where the track does not loose any of its fullness, but eliminates any un-wanted rumble or noise. Obviously, an instrument that produces a lot of low end gets little HPFing. For the next step, I disengage the HPF.
Third, I sweep through the frequency range (boosting 10 db or more) and isolate any weird and/or ugly sounding resonances. I’m not sure if these are the result of the room, the mic, or whatever, but I usually find some frequency that sounds sorta like feedback or sounds noisy and scoop it out. Typically, I find these resonances in the low mid-low and high-mid frequencies. Often I can narrow the scope of the curve to a fine point with the goal of scooping out the uglies without sacrificing the integrity of the track. Then I’ll re-engage the HPF and assess each track individually, making adjustments where needed.
Fourth, I listen to the whole tune and assess. Without getting too focused on detail, I listen for relative clarity in:
-instruments that have a lot of low end (bass drum relative to bass guitar)
-instruments that have lots of mids and highs (guitars, snare, keys, vocals, woodwinds)
-high end detail (cymbals, vocals, brass, snare)
Once I feel that the mix is closer to what is desired I’ll get progressively more detail oriented.

If I didn’t add compression to a track in the first step, I might do so at this stage. After getting the compression where I want it, I’ll try EQing the mix buss. If I’m not pleased with the results I’ll scrap the mix bus EQ and try to address any issues on a per-track basis (post comp), while toggling between the mix and a solo’ed track. Usually post comp EQ is additive: HF shelf, maybe boost some mids on a guitar track, etc.
Old 28th January 2008
  #2
Lives for gear
 

Quote:
and then EQ each track individually with the other tracks muted. Sometimes, I just jump right to EQing the individual tracks.
I agree with your overall approach and sequence of events........except for the above. I used to do the "a la carte" approach to EQ, but these days, I get a mix as close as I can without EQ, then start working my way down the tracks, EQ'ing with the entire mix up, occasionally soloing a track if it's unusually problematic. I've found that mixes are falling into place much faster for me now since I changed this tactic. I'm also becoming more sensitive to minute changes in pan settings rather than EQ to get individual tracks to stand out.

To address the thread topic more specifically, the main reason I use subtractive EQ before additive EQ is to maintain headroom in the gain structure of a mix.
Old 29th January 2008
  #3
Lives for gear
 
Sigma's Avatar
getting 15 lbs of sheit in a 10 lb bag
Old 29th January 2008
  #4
Lives for gear
 
drumzealot's Avatar
 

Quote:
Originally Posted by Sigma View Post
getting 15 lbs of sheit in a 10 lb bag
That's an interesting approach. Intelligent. Insightful. I’m sure this post will be referenced for ages to come.
I’ve printed this and will meditate on it. The next time I’m trying to figure out subtractive EQ I’ll think, “15 pounds of sheit…10 pound bag…WHAT DOES IT MEAN???” I’m sure enlightenment is right around the corner.
Thank you so much, Mike, for bestowing your coveted wisdom upon us.

For your invaluable contribution to furtherment of sound production I and the Board have unanimously chosen you, yes you, for this week's a-hole award.

May a toothless hag gnaw on your grandmother’s Wheat Thins.
Old 29th January 2008
  #5
Lives for gear
 
drumzealot's Avatar
 

Quote:
Originally Posted by river View Post
To address the thread topic more specifically, the main reason I use subtractive EQ before additive EQ is to maintain headroom in the gain structure of a mix.
Yes, but how do you execute the subtractive EQ? How do you decide what to subtract?
Old 29th January 2008
  #6
Lives for gear
 

Quote:
Yes, but how do you execute the subtractive EQ? How do you decide what to subtract?
First thing I do is to indentify if there are two or more source tracks competing with each other in a mix, make a decision as to which one wins out over the other(s) for any given frequency range, then go to work on the one(s) that need subtractive work. I tighten up the bandwidth all the way, boost gain and sweep through that area looking for a hot spot. When I find a frequency that might be problematic, I pull gain back on that band into negative numbers (usually not more than -4 dB) and slowly open up the bandwidth until I start to lose the overall character I want that track to have, then close it back up until the character returns. I'll often push gain back towards zero also until the masking fequency problem starts to re-emerge, then cut gain again until it disappears. General rule is that the lower the frequency, the tighter the bandwidth will remain. Getting back to the a la carte philosophy, what a track sounds like EQ'd in solo doesn't make it sit well in a mix relative to the other tracks competing in the same frequency range. The cumulative effect of tracks sharing the same frequency ranges can make any individual track sound fatter with respect to low end, duller with respect to upper mids and highs, so it's important to judge EQ changes with all tracks in the mix up, rather than by themselves. Also, low end and low mid problems can be contributed to with addition of reverb, so EQ'ing the effect is an effective technique for clearing up mud, as is high passing.....but I notice you said tracks with a lot of low end get little HPF. Consider that with digital, the bandwidth goes down to about 5 Hz. High pass filtering bass and kick at around 25-30Hz will clean up a bunch of subsonic crud that eats up enormous amounts of headroom and will make them both much tighter sounding. I often high pass the main mix bus also. Hope this makes sense/helps?
Old 29th January 2008
  #7
Gear Addict
 
Lee Knight's Avatar
 

Subtractive EQ. It's training yourself to identify that the issue, for instance, isn't that the voice needs a shelf boost at 6k, but rather, all that 250 (or whatever freq) is masking the nice, natural highs that are already there. That's a big part of why scooping a kick or toms works so well. Take away the mask and see what's really there.

So how?

Well, eq'ing in solo is a disaster. You need to identify what the issue is in context. BUT... once you've identified your snare needs more crack in the context of everything playing, for instance, solo it now.

Is there a frequency that is overtly cloudy? Boxy? Tubby? Identify the problem in solo. Do the old boost and sweep to zero in and learn numbers to sound. Now that you've got your problem frequency, bring everything back to flat.

Unsolo.

Now, in the context of the mix again, turn down the problematic frequency while boosting over all level of the snare. This simultaneously turns down the "boxy", "cloudy" _____ fill in the blank, while giving a natural high and low end boost. You're not actually boosting highs, but boosting overall level, but without the problem frequency, you are in fact bringing up the level of "the good" while bringing down the level of "the bad".

The same holds true the other way around. Guitars not fat enough? Maybe actually they're harsh.
Old 29th January 2008
  #8
Lives for gear
 
drumzealot's Avatar
 

Quote:
Originally Posted by river View Post
...but I notice you said tracks with a lot of low end get little HPF. Consider that with digital, the bandwidth goes down to about 5 Hz. High pass filtering bass and kick at around 25-30Hz will clean up a bunch of subsonic crud that eats up enormous amounts of headroom and will make them both much tighter sounding. I often high pass the main mix bus also. Hope this makes sense/helps?
I guess "little" is a relative term. I ussually end up the the HPF landing somewhere in the 20-30Hz range.

Thanks for the helpful info!
Old 29th January 2008
  #9
Lives for gear
 
amishsixstringe's Avatar
 

Quote:
Originally Posted by Lee Knight View Post
Subtractive EQ. It's training yourself to identify that the issue, for instance, isn't that the voice needs a shelf boost at 6k, but rather, all that 250 (or whatever freq) is masking the nice, natural high already there. That's a big part of why scooping a kick or toms works so well. Take away the mask and see what's really there.

So how?

Well, eq'ing in solo is a disaster. You need to identify what the issue is in context. BUT... once you've unidentified your snare needs more crack in the context of everything playing, for instance, solo it now.

Is there a frequency that is overtly cloudy? Boxy? Tubby? Identify the problem in solo. do the old boost and sweep to zero in and learn numbers to sound. Now that you've got your problem frequency bring everything back to flat.

Unsolo.

Now, in the context of the mix again, turn down the problematic frequency while boosting over all level of the snare. This simultaneously turns down the "boxy", "cloudy" _____ fill in the blank, while giving a natural high and low end boost. You're not actually boosting highs, but boosting overall level, but without the problem frequency, you are in fact bring up the level of "the good" while bring down the level of "the bad".

The same holds true the other way around. Guitars not fat enough? Maybe actually they're harsh.

Very Nice post.

Neil
Old 29th January 2008
  #10
Lives for gear
 
robot gigante's Avatar
I started having the best results once I became able to identify the sound of peaks or resonances at different frequencies without having to sweep much for them, which for whatever reason is sometimes deceptive for me. But once I was able to listen to a mix and instantly recognize stuff like a bump in the low mids with the guitar that's muddying up the mix for example, things became a lot easier. So I think ear training is key. Once you learn to figure out the exact thing that creates a problem, you may find yourself using EQ less- maybe just a tiny dip here and there in some cases- which I think is a good thing.

Maybe all of this goes without saying because it's so obvious, but I think that when doing the sweep method on every track, the big picture is easy to lose sometimes, especially since it might be one or two (or more of course) problematic tracks that are masking the others or making them sound a certain way. You don't want to be cranking highs on a lead vocal track or doing drastic subtractive EQ on it, for example, if other tracks are getting in the way of the vocal being able to cut through the mix. That's the ass-backwards way to work and it will make the mix sound bad.

So the skill to be able to identify the one or two things (or maybe more of course) that are preventing the big picture from getting through is really important imho, and equally important is the skill to assess whether or not the problem should be addressed with a something like a volume change and not EQ at all. Sometimes we might find a 'nasty resonance' while sweeping that is actually a key part of an individual track's sound, and all one needs to do to make that track fit is to drop it down a couple db and not EQ it at all. I think therefore that the mute button and the faders are two of the most important tools that we have when it comes to subtractive EQ, if that makes any sense.

edit- I just read Lee's post and I think he is kind of saying the same thing. Good post.
Old 29th January 2008
  #11
Lives for gear
 
Sigma's Avatar
Quote:
Originally Posted by drumzealot View Post
That's an interesting approach. Intelligent. Insightful. I’m sure this post will be referenced for ages to come.
I’ve printed this and will meditate on it. The next time I’m trying to figure out subtractive EQ I’ll think, “15 pounds of sheit…10 pound bag…WHAT DOES IT MEAN???” I’m sure enlightenment is right around the corner.
Thank you so much, Mike, for bestowing your coveted wisdom upon us.

For your invaluable contribution to furtherment of sound production I and the Board have unanimously chosen you, yes you, for this week's a-hole award.

May a toothless hag gnaw on your grandmother’s Wheat Thins.
both my grandmothers are food for poe's conquering worm.

wow..anger management classes in your future?..learn to spar before putting curare on your spear..your barbs are dull
i hope you are not so tight azzed around your clients..levtiy puppy levity

that being said

getting that 15 lbs of sheit in a 10 pound bag really was an back cover synopsis

the referece was made because the best answers were already given [river and lee] when arrraingements tend to be heavy..masking comes to play..when paning doesn't solve the masking issue..subtractive eq does

when top end seems to be lacking it may really be masking from too much low mid freq

reassessing eq after verb and fx and after that eq'g verb and efx are the norm

the great sound solo'd may not be this best sound in the mix

the big fat piano sounds great alone or with playing with a vln but with drums, power chords, heavy bass etc it would sound lost and tuddy up the mix


hence "getting the 15 lbs of sheit in a 10 lb bag"

lighten up dude..personal affonts don't work with me..i'm too old to take umbrage..

metimucil and viagra maybe
Old 29th January 2008
  #12
Gear Nut
 

There is some really great info in this thread, some of it should be sent to the wiki.
Old 29th January 2008
  #13
Gear Nut
 
mynameischance's Avatar
Quote:
Originally Posted by amishsixstringe View Post
Very Nice post.

Neil
+1
Old 29th January 2008
  #14
Gear Nut
 
mynameischance's Avatar
Course it's all dependent on the source material, But I've been doing a lot more experimenting with subtractive eq'ing after I received the most NOBLE of suggestions:

A/B like crazy!
Find bands, songs, production quality that you enjoy and ref against it. Reach for it.

I have a mixing ref playlist on my iPod with the songs I think most pertain to what I'm going for. Whatever field of my song I happen to be concentrating on, I'll tend to play a few seconds, bars of these mixes and it sort of sets my compass. BTW, I pull the iPod into an aux and set the level of the aux in the general db of what I'm pushing on my mix. Since you're comparing mastered vs unmastered, no reason for the ref to be too loud. Hard to compare otherwise.

And what i've quickly found is the whole notion of bottom in bass, or kick is so fundamentally less than what your heart desires. Listen to your favorite tracks! It's not nearly as much as you think!

Heck, even just last night, I solved a kick drum issue not by bringing out 100hz or 80 hz (where you think that beautiful thump would be) but rather applying a low pass all the way up to 60hz. It was a particular sound I was going for, but wow what a difference. And the kick was still pushing, just not crud.

With vocals I can go as high as 300-400 (again song dependent). Guitars not so much. With my overheads and hihats, I apply the hi pass while tracking. makes a huge difference....
Old 29th January 2008
  #15
Moderator
 
TonyBelmont's Avatar
 

Quote:
Originally Posted by drumzealot View Post
Yes, but how do you execute the subtractive EQ? How do you decide what to subtract?
Using your ears and experimenting. That's what it really comes down to.
Old 29th January 2008
  #16
Lives for gear
 
drumzealot's Avatar
 

Quote:
Originally Posted by Sigma View Post

lighten up dude..personal affonts don't work with me..i'm too old to take umbrage..

metimucil and viagra maybe
May a toothless hag gnaw on your grandmothers Wheat Thins is heavy?
Evidently your not old enough to remember that Johnny Carson bit.
Old 29th January 2008
  #17
Gear Head
 

Apparently,
subtractive eq is what started it all, the eq's really just added noise if you boosted the frequency so cutting was really the only way.
I've heard a lot of engineers swear by never going into the +db section and just adjusting the output or compressing, limiting, or something else after.

As much as I'd like to always be one of those guys, I do add db here and there- makes life easy. But concerning subtractive: I always try to cut before I boost.

Concerning soloing (while you've already gotten several comments on it...ill add mine) I solo mine occasionally and sweep just to see if there is something really pleasant in there Im NOT hearing and if so, I try to emphasize it. Otherwise, ala carte really is like having a bunch of talented blind people build a building together.

Love your post, and while I'm only 21...I feel like a lot of young engineers don't know anything about cutting freq's unless its for resonance control (and sometimes...not even then)
Old 30th January 2008
  #18
Lives for gear
 
Sigma's Avatar
Quote:
Originally Posted by drumzealot View Post
May a toothless hag gnaw on your grandmothers Wheat Thins is heavy?
Evidently your not old enough to remember that Johnny Carson bit.
i was never home that early
Old 30th January 2008
  #19
Gear Nut
 

Drumzealot-

I appreciate your post here. Kind of like people have been saying I think it's real important to mix in context. The "size" and timbre of a sound is extremely contextual. Kind of like those visual "tricks" where an identical color appears lighter or darker based on what color is next to it.

I think many beginning engineers, myself included, begin by trying to increase what we want to hear instead of decreasing what we don't want to hear. Hence a lot of my first mixes included a lot of additive EQ. The longer I work with music, though, the more I become convinced it's about what you take away. Especially since we're faced with "15 pounds" that need to be fit in a "10 pound" sack nearly every time we sit down to mix something.

So I try to begin by asking myself "who is going to be the big boy here" bearing in mind that mixed music only really gets to have about 2-4 big boys at any one time. On the recommendation of one of the resident Gearslutz engineers that means I usually begin with the vocal. I already know the "space" the drums are going to occupy and their formulaic and transient nature makes them simple to deal with. The real question, and where the art of subtractive EQing come in, is in the 300-3000Hz range. Here I find no one solution to really fit all situations but in general I get a lot of mileage out of asking myself, "how much can I take away from this 'blank' before it doesn't sound like a 'blank' anymore?" In other words, I experiment with subtracting frequencies from my sounds (in context) and try to ascertain the effect my subtraction is having on my ability to hear other instruments. If the instrument I've subtracted from still sounds like a 'blank' I consider it a done deal. I should add that I usually start this subtraction process with the instrument that plays the least important role.

I guess all of this is my long way of saying I love subtractive EQing, and consider it one of my principal mixing techniques. Again, I agree with all the guys saying "ears, ears, ears." Unfortunately this doesn't allow for an easy answer to your original question but I hope my "questions" might be of some help to people.
Old 30th January 2008
  #20
Gear Guru
 

Quote:
Originally Posted by Relicrecordings View Post
I've heard a lot of engineers swear by never going into the +db section and just adjusting the output or compressing, limiting, or something else after.

As much as I'd like to always be one of those guys, I do add db here and there- makes life easy.
I hear a lot of people swear they never watch TV, or that they only watch PBS. tutt

Everybody boosts sometimes. I never got the superiority complex associated with subtractive EQ.

Not to take anything away from subtractive EQ, but in the end it's just another curve.
Old 30th January 2008
  #21
Gear Maniac
 
Mr. Man's Avatar
 

Boosting EQ increases the noise floor and phase distortion (in analogue land). Cutting is better because it doesn't introduce those artifacts.

One of the differences between boosting and cutting is it's a steak knife versus a scalpel. I like to 'amputate' or cut only what is not needed or flattering and boost what it really needs much as others, I'm sure, do. I don't give a d*mn about the Q factor – only if it sounds good.

And A/Bing is vital. Psychoacoustically your ears become accustomed to sound after 30 second even if it sounds bad!

Then there's the whole "should I use a shelf, bell, or graphic EQ?" mess. Bottom line is, if it sounds good, it is good. Don't worry about where the knobs are pointing even if they are all backwards.
Old 30th January 2008
  #22
Here for the gear
 

Fantastic thread, thanks guys.
Old 30th January 2008
  #23
Gear Guru
 
Drumsound's Avatar
I didn't read the whole thread so forgive me if this has been covered. DON''T EQ IN SOLO!
Old 30th January 2008
  #24
Lives for gear
 
amishsixstringe's Avatar
 

Quote:
Originally Posted by Drumsound View Post
I didn't read the whole thread so forgive me if this has been covered. DON''T EQ IN SOLO!
You need to. Threads like these are the reason I still come to this board. And yes, the don't eq in solo thing has come up once or twice...in every eq thread there is. I'm not trying to take stabs at you, by the way. Just saying...read the thread. it's good.

Neil
Old 30th January 2008
  #25
Lives for gear
 

Hey..... a friendly little piece of advice here guys.

If you are kinda' new here or just sorta' new...
don't jack with people who give you answers that you don't understand.

You be surprised at who you might piss off.

I knew exactly what Sigma meant.
I also know that he was probably working in his dad's VERY famous studio when you were watching Johnny Carson.
Old 30th January 2008
  #26
Lives for gear
 
BLueROom's Avatar
 

Quote:
Originally Posted by Lee Knight View Post
Subtractive EQ. It's training yourself to identify that the issue, for instance, isn't that the voice needs a shelf boost at 6k, but rather, all that 250 (or whatever freq) is masking the nice, natural highs that are already there. That's a big part of why scooping a kick or toms works so well. Take away the mask and see what's really there.

So how?

Well, eq'ing in solo is a disaster. You need to identify what the issue is in context. BUT... once you've identified your snare needs more crack in the context of everything playing, for instance, solo it now.

Is there a frequency that is overtly cloudy? Boxy? Tubby? Identify the problem in solo. Do the old boost and sweep to zero in and learn numbers to sound. Now that you've got your problem frequency, bring everything back to flat.

Unsolo.

Now, in the context of the mix again, turn down the problematic frequency while boosting over all level of the snare. This simultaneously turns down the "boxy", "cloudy" _____ fill in the blank, while giving a natural high and low end boost. You're not actually boosting highs, but boosting overall level, but without the problem frequency, you are in fact bringing up the level of "the good" while bringing down the level of "the bad".

The same holds true the other way around. Guitars not fat enough? Maybe actually they're harsh.
excellent post. great advice and well stated. if there was a rep system here, you'd get rep'd
Old 30th January 2008
  #27
Lives for gear
 

Quote:
Originally Posted by Relicrecordings View Post
Apparently,
subtractive eq is what started it all, the eq's really just added noise if you boosted the frequency so cutting was really the only way.
I've heard a lot of engineers swear by never going into the +db section and just adjusting the output or compressing, limiting, or something else after.

As much as I'd like to always be one of those guys, I do add db here and there- makes life easy. But concerning subtractive: I always try to cut before I boost.

Concerning soloing (while you've already gotten several comments on it...ill add mine) I solo mine occasionally and sweep just to see if there is something really pleasant in there Im NOT hearing and if so, I try to emphasize it. Otherwise, ala carte really is like having a bunch of talented blind people build a building together.

Love your post, and while I'm only 21...I feel like a lot of young engineers don't know anything about cutting freq's unless its for resonance control (and sometimes...not even then)
I use subtractive eq a lot like everyone else.... all the reasons have been stated here already.....please don't suggest to people who want to learn from this post that additive is bad....sometimes + 20 is what you do if its needed....however I would say learn subtractive before additive....and....stop suggesting that eqing while a track is soloed is incorrect...there are a lot of issues you can (and should) deal with when you are listening to it alone...It is an important part of the process..

Nick

BTW I'm not pointing the finger at the person I quoted...just general. Peace.
Old 30th January 2008
  #28
Lives for gear
 
Unclenny's Avatar
Once I started to approach EQ as frequency specific gain control it took away some of the 'voodoo factor'.

Also......if you're still stumped by these applications and you don't quite trust your ears yet throw up a spectrum analyzer (PAZ for instance) and watch what happens when you muck out the barn at 300.
Old 30th January 2008
  #29
Lives for gear
 
loujudson's Avatar
Back in the 60s when I actually got some education in audio old school style, the word for boosting a frequency was "equalize" and the word for cutting was "filter" - so you can streamline the language if you like, by filtering instead of EQing. We were also taught (and learned by experience) that filtering is usually better than EQing. Partly it depends on the quality of the gear, partly on phase shift, but mainly taking away is cleaner than adding to.

But always it is better if the player can adjust, or the mic, or the room, the usual blah blah.

<L>
Old 30th January 2008
  #30
Gear Head
 
darkswan's Avatar
What types of tracks might you be inclined to use a band pass to compartmentalize frequency vs. HPF?

Instead of only a 30hz HPF on bass or drum - what if you set a band pass at say 300hz on the same track.

Does this make the music sound unnatural? especially if you discreetly mix an entire project with band pass filtering?
Post Reply

Welcome to the Gearslutz Pro Audio Community!

Registration benefits include:
  • The ability to reply to and create new discussions
  • Access to members-only giveaways & competitions
  • Interact with VIP industry experts in our guest Q&As
  • Access to members-only sub forum discussions
  • Access to members-only Chat Room
  • Get INSTANT ACCESS to the world's best private pro audio Classifieds for only USD $20/year
  • Promote your eBay auctions and Reverb.com listings for free
  • Remove this message!
You need an account to post a reply. Create a username and password below and an account will be created and your post entered.


 
 
Slide to join now Processing…
Thread Tools
Search this Thread
Search this Thread:

Advanced Search
Forum Jump
Forum Jump