Subtractive EQ - how many cuts?

rokdan

New member
Hi all, first post here :-)

With the endless supply of eq's these days, when it comes to subtractive eq and eliminating harsh frequencies, how many cuts is good practice?

I mean an acoustic guitar for example, you can probably find at least 5 bad frequencies when boosted with narrow Q etc - do you cut all of them? Or just 1 or 2?

I know the ears are the golden rule, but what is common practice - and what is ridiculous over the top cutting?

The big desks usually have 2-3 sweepers - is that enough? Or is that a limitation on the actual hardware?

Cheers
Dan
 
I think the answer you're going to get is: as many as needed.

I'm no expert, I've just seen this play out before.
 
I will add to this. You have to listen and determine the sound based on the mix. If you EQ an instrument alone, then bring it into the mix, it will have different results than just using the EQ in relationship to the mix.

5 bad frequencies on an acoustic? I guess I need to go back and revisit my guitar tracks. I know of two I pretty much treat, down at the low end where it tends to rumble and some slight roll off at the top (I don't like bright AG). Other than that, I wouldn't know where the other three are.

EQ to/for the mix.
 
Pretty much any tight range of frequencies is going to sound bad when boosted.

Sometimes it's advised to 'scan' with a narrow notch to help you find problem frequencies, which seems to be what you're describing, but that's on the assumption that you're aware of a problem and your ears aren't experienced enough to just know straight-off where it is.

If you scan a perfectly good recording like that you could convince yourself there are problems that just aren't there.
Could that be happening?
4 band eq is almost always more than enough for me.
 
I rarely even EQ an acoustic guitar, and when I do it's almost always a boom at 200-250Hz that I cut, anything from 1-4dB, this has the same effect as boosting high mids, but only do in on boomy sounding guitars recorded with an LDC.
 
Hi all, first post here :-)

With the endless supply of eq's these days, when it comes to subtractive eq and eliminating harsh frequencies, how many cuts is good practice?

I mean an acoustic guitar for example, you can probably find at least 5 bad frequencies when boosted with narrow Q etc - do you cut all of them? Or just 1 or 2?

I know the ears are the golden rule, but what is common practice - and what is ridiculous over the top cutting?

The big desks usually have 2-3 sweepers - is that enough? Or is that a limitation on the actual hardware?

Cheers
Dan
Everyone's situation is different (Room, guitar, mic, sound you're looking for, etc......). The only thing that matters is the final product. If you need to cut 25 frequencies, then so be it. Nobody listens to a finished product and knows what was done to achieve the sound they're hearing. Just do whatever you feel you need to do to get the best sound you can. If it means 30 cuts, so be it. If it means no cuts at all, so be it. Just do what you need to do to get the sound you want.

Of course, if you find you need to cut/boost a lot of frequencies, that's probably a clue that you have a room, mic, or guitar problem. But that doesn't mean you shouldn't do whatever you need to do to get the best sound out of a recording.
 
I've stacked EQs before, but I realize using a lot of cuts and boosts means I did a lousy job tracking. Cut whatever it takes, but be aware, it may be easier to retrack and only have a few problems. I've also noticed that things are cleaner and better if I cut frequencies, then compress, then boost. Compressing the boosted frequencies does not work well (at least for me).
 
When you do the boost to find the bad stuff thing, you can always find certain frequencies that sound bad that don't need to be cut. The question you need to ask is "when do I stop adding eq points?"

After each eq adjustment, listen to the track. Is there something wrong with the frequency balance? If there is, figure out where (too much low, too brttile, etc) and sweep that area to find the problem.

If you just keep adding eq points and sweeping, you will.end up taking things out that shouldn't be.

2-3k is always annoying, but it balances some of the low end. If you take too much out, it will get muddy, which will lead you to take out 200hz, which will now leave it too midrangy, so you take out the 800hz, etc...

It's very easy to chase your tail doing this. If it takes more than three parametric point and high and low shelf/cut, there was something wrong with the recording.
 
Pretty much any tight range of frequencies is going to sound bad when boosted.

Sometimes it's advised to 'scan' with a narrow notch to help you find problem frequencies, which seems to be what you're describing, but that's on the assumption that you're aware of a problem and your ears aren't experienced enough to just know straight-off where it is.

If you scan a perfectly good recording like that you could convince yourself there are problems that just aren't there.
Could that be happening?
4 band eq is almost always more than enough for me.
Jay and Steena' nailed it.
To me mixing is circular; You go after and solve the big obvious stuff first, revisiting and refining as you go. As the mix gets better you're tuning and zeroing in on finer things. Sometimes entering those areas where it's 'different but not necessarily better, or worse lose sight of the big picture, or raw qualities that made it fun and interesting.
 
Thanks all! Seems as though I maybe over-thinking/analysing... yes Steenamaroo I was busy scanning with a narrow notch... I was happy with the original tracked acoustic, but as you said, I was convincing myself otherwise!! Also I must do more in-mix eqing...
 
Jay and Steena' nailed it.
To me mixing is circular; You go after and solve the big obvious stuff first, revisiting and refining as you go. As the mix gets better you're tuning and zeroing in on finer things. Sometimes entering those areas where it's 'different but not necessarily better, or worse lose sight of the big picture, or raw qualities that made it fun and interesting.


Right. And after experience and time working out the recorded tone, and the room, and the mics, and the......We start to find what works to a degree. But then every song, track, or artist is different.

Then we all just sit back and make new decisions for each.

I surely don't have a direct answer to the OP's question. I can only assume there is no correct answer. Just do what works and then do it again. :)
 
Thanks all! Seems as though I maybe over-thinking/analysing... yes Steenamaroo I was busy scanning with a narrow notch... I was happy with the original tracked acoustic, but as you said, I was convincing myself otherwise!! Also I must do more in-mix eqing...

My personal advice would be to do less 'in-mix' eqing, and more time spent on getting the track sounding good when recorded. Mic placement/room/whatever.

IMO, Eq is something that should be used to 'sweeten' if you will, the way an instrument sits in a mix. Not to define it's tone.
 
Jay and Steena' nailed it.

Thanks all! Seems as though I maybe over-thinking/analysing... yes Steenamaroo I was busy scanning with a narrow notch... I was happy with the original tracked acoustic, but as you said, I was convincing myself otherwise!!

I do it myself sometimes.
That's OK, as long as I catch myself doing it. ;)

IMO, Eq is something that should be used to 'sweeten' if you will, the way an instrument sits in a mix. Not to define it's tone.
Quote of the week nominee, although I think Greg's going to win this week.
He's had a damn good week though. :p
 
Always try to fix it at the source (eg if floor tom is ringing use somethign to dampen it). Otherwise, when you're using surgical EQ, as much as needed (or permitted by the EQ).

If you're have to significantly dip 3 or more frequencies, then you probably have a serious issue with the source or location that you need to address.
 
Everyone is pretty on point here... I'll just add something.

I generally use Waves Q10 for EQ. I can instantiate one plug, and use as many bands as needed, or as few as needed, with plenty there to add as I go, again, if needed.

Handy to have a plug that operates that way.... lots do

I avoid the use of modeled plugs when cutting out bad frequencies, as most models lock you into a specific Q, db of cut/boost, curves/types, or specific bands, ie. 60-100-150-so on.... also, some add color you may not want when getting surgical

And, a EQ like Q10 will let you target exact frequencies, curves, Q's and db of cut or boost, as needed.

Most of this I write while keeping surgical subtractive EQ in mind.... and I quite often get by without any boosting at all, but the same things keep true

Now.... One more thing I'll add, is that while thinking of EQ'ing... subtractive in particular:
Example:

Say I have a 250 or 500 range in a vocal I want to cut cause it's sorta muddy or honky.......
I can cut some EQ
Or, if the mud or honk is not persistent, and say it's only picking up during vowels or specific spots, rather than EQ, I might use C4 or C6 to target that frequency, bypass other bands so it's only effecting that frequency/band, and use it like a DeEsser only in that range

This is handy when the frequency in question is not persistent, but only at certain spots. Simply cutting that range with EQ can make the rest of the vocal thin or otherwise unpleasant or not suited to the mix.... follow?

Good lord.... so many ways to skin the cat :)
 
Best practice, of course, is no cuts at all. Sort you instrument, acoustics and microphone so you like the sound as it's recorded.

In reality, a bit of tweaking is often called for but use it sparingly and don't cut anything unless you're sure it improves the sound. I have to say that it's all too easy to suck the life out of your recording by getting over analytical and over fussy. When you listen to the same stuff over and over, you start to hear problems that aren't there or, even worse, are actually integral parts of the music. Real life DOES have some frequencies louder than others. What sound "harsh" on the 34th listen to a recording may well be just the natural sound of a guitar.
 
Hi all, first post here :-)

With the endless supply of eq's these days, when it comes to subtractive eq and eliminating harsh frequencies, how many cuts is good practice?

I mean an acoustic guitar for example, you can probably find at least 5 bad frequencies when boosted with narrow Q etc - do you cut all of them? Or just 1 or 2?

I know the ears are the golden rule, but what is common practice - and what is ridiculous over the top cutting?

The big desks usually have 2-3 sweepers - is that enough? Or is that a limitation on the actual hardware?

Cheers
Dan

These are limitations of the actual hardware but with a good purpose in it.

People spent time to record almost perfectly, so 3 were more than enough :)

There's really no answer to your question, it's like asking "how much salt should I put to food".
Each food is different, cooked different, used different ingredients, from different land, with different cooking hands and aspiration.

I am being ironic, I really used this example to help you understand why this question can't be asked.

What you can do is to boost with a band sweep around, find something that sounds s**ty and reduce it without being noticeable.
 
Back
Top