Differentiating pleasant/unpleasant frequencies

[TPCalm]

Hello.
Let's take for example, distorted guitars.
As I EQ to open up space in a mix, I have trouble finding what makes some frequencies pleasant as opposed to being unpleasant.
To my ears, boosting and frequency sweeping with a tight Q by +24db sounds awful across the entire spectrum.

What am I missing here?

Is this process done mathematically by finding the fundamental frequency and attenuating odd/even harmonics?

Thanks in advance for your help.

 

[thefalloftheheretic]

Generally, as you're doing said EQ boosting, there will be one frequency that grates against your ears worse than the others.
Reduce that.

There are certain frequencies that generally need to be cut out (the 4.5 kHz notch being the classic example).

It's not mathematical, it's just all a matter of getting the bad to sound good.
Whatever works for you may not work for me, or vice versa.

 

[heshian46]

What is a good Q width to sweep at? Im typically doing it with the tightest Q possible, and +24dB, but is this really the best way to go?

 

[thefalloftheheretic]

What is a good Q width to sweep at? Im typically doing it with the tightest Q possible, and +24dB, but is this really the best way to go?

It all depends.
On REAPER's ReaEQ, I alter the Q depending on the track.
Narrowest is NOT always the best, I prefer to have a little bit of width to my Q so that I can hone in on the offending frequencies the best.

 

[Jordon]

If I do this, I'll usually do an 18db boost with a fairly narrow Q and SLOWLY sweep small sections of the spectrum at a time. Whatever overly-harsh or really resonant frequencies (for newbs, whatever frequency all of a sudden jumps up in volume) I find, I'll dip down by 2 or 3 db after slightly widening the Q so that it sounds more natural.

 

[Nimvi]

What is a good Q width to sweep at? Im typically doing it with the tightest Q possible, and +24dB, but is this really the best way to go?

Altough that's the gist, I think that the extremity of those settings is giving you troubles. If you boost the smallest possible range by 24 dbs, I think anything will sound terrible, like you've mentioned. I usually use a peak of about 8 to 12 db, with a pretty tight q (but not necessarily the tightest possible). For example, if you are using ReaEQ in Reaper, the smallest possible Q value is 0.01, but I would use it at about 0.10 to sweep.

An alternative is to do the same thing with a cut instead of a boost by the way. Just make a tight notchfilter and slowly move it around. If there is something irritating, you will feel relieved whenever you hit it. I find this method less exhausting to the ears, especially when working with distorted guitars.

 

[thefalloftheheretic]

Altough that's the gist, I think that the extremity of those settings is giving you troubles. If you boost the smallest possible range by 24 dbs, I think anything will sound terrible, like you've mentioned. I usually use a peak of about 8 to 12 db, with a pretty tight q (but not necessarily the tightest possible). For example, if you are using ReaEQ in Reaper, the smallest possible Q value is 0.01, but I would use it at about 0.10 to sweep.

An alternative is to do the same thing with a cut instead of a boost by the way. Just make a tight notchfilter and slowly move it around. If there is something irritating, you will feel relieved whenever you hit it. I find this method less exhausting to the ears, especially when working with distorted guitars.

Man, I never thought of that. I might have a new method.

 

[heshian46]

Good information fellas.

And of course we don't want to cut too much of the offending freq once we find it. Jordon, you mentioned cutting about 2-3 dB with a slightly wider Q than your sweeping Q. Is this typical? I'm finding that any more than that and you lose a good amount of necessary information in that area.

What are the rest of you guys doing?

 

[thefalloftheheretic]

Personally, I don't think "typical" really applies in the art of mixing and mastering.
It all depends on the source.
If your source has a huge 4 kHz spike, you might need to cut -12 dBs in that general area.

Does that make sense?

 

[heshian46]

Personally, I don't think "typical" really applies in the art of mixing and mastering.
It all depends on the source.
If your source has a huge 4 kHz spike, you might need to cut -12 dBs in that general area.

Does that make sense?

I agree with you, but im talking about the methods we use right off the bat to identify these problem areas.

Your still going to use a method you're familiar with first to find that 4 kHz spike.

 

[amarshism]

I would try and hear the problem first then use the sweep to help find it. Not always the other way around. This will improve your listening skills and you will soon find when you go to sweep youll start going straight to the problem area +/- 100hz rather than 1000 or more lol.

 

[Jordon]

Personally, I don't think "typical" really applies in the art of mixing and mastering.
It all depends on the source.
If your source has a huge 4 kHz spike, you might need to cut -12 dBs in that general area.

Does that make sense?

Exactly, it all depends on the source material and how it's interacting with things around it. The main reason I lean towards using a slightly wider Q when notching out frequencies is that it just sounds and feels more natural.

In the past, I've found that if you're trying to clean up the same frequency in multiple instruments (for arguments sake you tracked a lead vocal, backing vocals and an acoustic in the same room with the same mic at different times), approaching the cleanup with varying Q's and amounts of reduction will lend a more pleasing result, and you're less likely to wind up with a gaping hole in your mix at 4k or whatever.

Of course, I'm not saying that there isn't a need for ultra-surgical cuts. I was handed some overhead tracks recently that were an absolute mess above 6k. That was a fun evening.

EDIT: Then there are those fun times where you notched a frequency out that was annoying when the instrument was soloed, and you come to find out that's where all the life was in the track. :danceboy:

 

[thefalloftheheretic]

EDIT: Then there are those fun times where you notched a frequency out that was annoying when the instrument was soloed, and you come to find out that's where all the life was in the track. :danceboy:

Then of course are the times where the people you tracked wouldn't fucking listen to you and you end up doing 25 notches in their messy, off time guitar playing. And you end up being paid more. :hotjump:

 

[Jordon]

Then of course are the times where the people you tracked wouldn't fucking listen to you and you end up doing 25 notches in their messy, off time guitar playing. And you end up being paid more. :hotjump:

I love those moments.

 

[tonymiro]

I would try and hear the problem first then use the sweep to help find it. Not always the other way around. This will improve your listening skills and you will soon find when you go to sweep youll start going straight to the problem area +/- 100hz rather than 1000 or more lol.

This and for the same reason. Developing critical listening can be really useful. You don't need perfect pitch but it's really useful to have, for instance, a good idea what 3000Hz, 1500Hz, 1000Hz, 250 Hz, 50Hz etc sound like. It will also help develop an idea of how a HPF/LPF is likely to affect a mix and so on.

 

[TomMartin]

I would try and hear the problem first then use the sweep to help find it. Not always the other way around. This will improve your listening skills and you will soon find when you go to sweep youll start going straight to the problem area +/- 100hz rather than 1000 or more lol.

100% agree.

If you can't hear something is wrong in the first place, sweeping won't reveal anything to you. Fuck, maybe nothing IS wrong afterall!

If I hear something I don't like, I get a narrow q and a huge boost and sweep around until the boosted frequency matches the pitch of whatever was pissing me off before I started sweeping and boosting. Then cut it out.

Boosting and sweeping reveals absolutely nothing. Of course every frequency sounds terrible, you just boosted it by 24db! Rather than reveal what's wrong, use the method to track down the frequency that needs cutting AFTER you already heard it.

 

[Leon--]

i suggest you to cut all the unpleasant freq's with notch filters,and then you will see if you need to boost some of them,it's really important first to remove,and then add

 

[TPCalm]

Thanks, dudes. I've got more than enough to work with.