EQ question

  • Thread starter Thread starter geneticfunk
  • Start date Start date
G

geneticfunk

New member
OK, when I'm EQ'ing, what bugs me is that when I apply boost to any region, the overall volume (mathematically speaking, this is the power of the signal, right?) increases.

So, when I apply boost to a region, and it sounds better, it's hard to tell whether the improvement in the way it sounds comes from the simple increase in volume, or if it's the increase in the boosted frequencies relative to the non and less boosted frequencies that is responsible for the improvement in the way it sounds.

So, I have an idea of how to get around this problem, and I'd be interested to hear any thoughts and suggestions about the idea and also about how a person could actually put this idea into practice. If anyone has any better alternative, I'd be even more interested to hear about those =)

Anyway, my idea is to somehow make the overall volume (power?) of the flat and the EQ'ed signals equal to one another. This would require some sort of constant recalculation and adjustment of the EQ'ed signal's overall volume (power?), as the EQ was applied and tweaked.

So, what do you think?
 
well that's a good idea in theory, but you might be thinking a little more than you have to.


You have to take into account that a 2 db boost at 4khz is not the same as a db boost at say 340hz.

So to answer your question correctly, you're actually hearing both types of changes.

An increase/decrease in volume and a change in EQ.

Our ears don't ear things flat. This is a result of the Flecther Munson Curve or otherwise known as the equal loudness curve.

It's the same curve that tells us to generally mix at a certain level and how our ears react to EQ at different levels.

It should just be as simple as listening and then tweaking. If you have to, bypass your EQ from time to time just to make sure you didn't cause an unessesary change in volume.


It's more artistic than it is technical.
 
Yes, we generally equate louder to better, and it would be great if there was a CPL (constant percieved loudness) circuit that could follow an EQ to compensate for the change in percieved loudness (but defeatable).
One thing to keep in mind .... subtractive EQ usually sounds better than additive, so try more cuts and less boosts. Then you'll have the opposite problem! Dohh!

-RD
 
Well hell, why don't we just encorporate the equal loudness (i.e. Fletcher Munson) curves, along with the position of the monitor speakers (due to the dependence on perceived loudness on spatial considerations), and/or the 468 weighting curve (http://en.wikipedia.org/wiki/Equal-loudness_contours) into our CPL circuit?


From the link above:
The 468-weighting curve, originally proposed in CCIR recommendation 468, but later adopted by numerous standards bodies (IEC,BSI,JIS) was based on BBC Research using a variety of noise sources ranging from clicks to tone bursts to pink noise, and incorporates a special Quasi-peak rectifier to account for our reduced sensitivity to short bursts and clicks. It is widely used by Broadcasters and is by far the preferred weighting to use for all forms of noise measurement, enabling subjectively valid comparisons of different equipment types to be made even though they have different noise spectra and characteristics.



I definitely agree that if one is attempting to do art with their EQ'ing, then this EQ'ing is ultimately an artistic process. I just want to be able to change one variable at a time in my art.
 
P.S.- This applies to not just EQ, but all sorts of effects processing too.

Also, a sort of related question... do level meters on pro audio gear/computer applications usually incorporate a CPL type feature, or do they just "tell it like it is" physically?
 
geneticfunk said:
P.S.- This applies to not just EQ, but all sorts of effects processing too.

Also, a sort of related question... do level meters on pro audio gear/computer applications usually incorporate a CPL type feature, or do they just "tell it like it is" physically?
Click to expand...

Peak meters tell it like it is, absolutely. RMS meters will give you a closer approximation to perceived level. On a lot of software, and even some hardware, you can select the meter ballistics.

-RD
 
It should be simple enough to play with the volumes to compare the two sounds. Try monitoring around 85 db SPL, the hearing curves level off around there.
 
geneticfunk said:
Well hell, why don't we just encorporate the equal loudness (i.e. Fletcher Munson) curves, along with the position of the monitor speakers (due to the dependence on perceived loudness on spatial considerations), and/or the 468 weighting curve (http://en.wikipedia.org/wiki/Equal-loudness_contours) into our CPL circuit?
Click to expand...


Well that's a good question. But you see, the curve is general, no two ears are ever alike. So it's not something I always remember when I EQ, just a nifty thing to keep in mind.


But really the best approach for this sort of thing is just that, personal interpretation.

In the end it's really not supposed to be complicated, just rather time consuming. The more acute your earing becomes, the harder it is to satisfy EQ settings.
 
Yeah, it does vary from person to person, but those fletcher munson curves, as well as the curves derived from more recent experiments, represent the curves of many people, averaged together.

It's not hard to grab the faders and readjust the levels to try to get back to the original volume level, but during that process my ears are forgetting what the original sounded like. It would be much better to be able to compare one and the other without this extra step in between, throwing your ears off.

Even engineers make use of these averaged curves. Here's from that link posted above again:

The 468-weighting curve, originally proposed in CCIR recommendation 468, but later adopted by numerous standards bodies (IEC,BSI,JIS) was based on BBC Research using a variety of noise sources ranging from clicks to tone bursts to pink noise, and incorporates a special Quasi-peak rectifier to account for our reduced sensitivity to short bursts and clicks. It is widely used by Broadcasters and is by far the preferred weighting to use for all forms of noise measurement, enabling subjectively valid comparisons of different equipment types to be made even though they have different noise spectra and characteristics.
 
Robert D said:
Yes, we generally equate louder to better, and it would be great if there was a CPL (constant percieved loudness) circuit that could follow an EQ to compensate for the change in percieved loudness (but defeatable).
One thing to keep in mind .... subtractive EQ usually sounds better than additive, so try more cuts and less boosts. Then you'll have the opposite problem! Dohh!

-RD
Click to expand...

I was referring to the term that Rob defined in the earlier post in this topic, standing for Constant Perceived Loudness.
 
one thing ive been doing is using the FreeFilter Direct X plugin OR the HAR-BAL plugins..

with those , you load up a WAV from a CD of what you consider to be the best example of eqing for the genre you are working on at the moment, and the plugins show you what the ME did in terms of EQ to that track. You can save those settings and apply them to your own work. I use Alison Krauss and Union Station, Live for bluegrass, the who's tommy for rock, and for classical I use Decca releases...

Now I dont use the settings to the T, but it is a good baseline/learning tool from which to do my own work.

I know, OT...it has helped me greatly though.
 
it has helped me a lot.

EQ is such a touchy thing, what sounds good to one may not sound good to another.

so I just try to learn from albums I love, and professionals that I trust.
 
geneticfunk said:
Well hell, why don't we just encorporate the equal loudness (i.e. Fletcher Munson) curves, along with the position of the monitor speakers (due to the dependence on perceived loudness on spatial considerations), and/or the 468 weighting curve (http://en.wikipedia.org/wiki/Equal-loudness_contours) into our CPL circuit?
Click to expand...

I believe that this was the general idea behind the "loudness" button found on some of the older stereos.

Kinda sucked.
 
You must log in or register to reply here.
Back
Top