Audio EQ - Should all the EQ numbers basically net zero?

submitted by cheese_greater

Like, if you take away 4db "from treble", should that be distributed roughly amongst bass and mid or solely either if you want it to be heavy for that range?

Is it like an equation that benefits from balancing it out?

Log in to comment


The Hobbyist

An equalizer does not have to sum up to any specific number. Each frequency range is basically being amplified or attenuated individually. You are boosting or reducing specific frequency ranges. If you reduce them all equally, then the end result is that your song is lower volume. If you boost them all, your song is louder.

Of note: boosting songs may cause occasional crackling sounds. If this is the case it is because the boosting is clipping the top end of the amplitude of your signal at various frequencies. So boost moderately. You are better off reducing some frequencies and leaving the rest normal and increase the volume of the source to compensate whenever possible.


Yeah, that last paragraph is important. I’m a professional audio technician, and way too many people will begin with boosts instead of cuts. But cutting is much easier on a technical level, because you’re just lowering the volume of something. Boosting is much more technically complicated, because you’re “adding” signal that doesn’t already exist. So you have to make that signal from something, and that’s much more technically difficult than simply turning the volume down.

Imagine you have a signal coming in at a baseline of 0dB. Cutting 6dB is easy, because you simply let less of the signal pass. But if you want to turn it up 6dB, you need to “create” that 6dB from somewhere, because it doesn’t already exist. You can’t just “turn it up” because it’s already turned all the way up at 0dB.


No. The eq should be the inverse of the room/speaker response; it’s irrelevant if the gains add to zero.

And, most music has way more energy in certain (generally lower) bands, so it’s not like making the numbers sum to zero results in the same total energy.


I think that's a very important point. If you "balance" the EQ like you describe but you boost a frequency band that our ears are more sensitive to and lower a band that we're less sensitive to by the same amount of dB... The mix might still sound louder than before.

So don't worry too much about the numbers and heed the great advice given in this thread.


No, you EQ to bring specific frequencies to the front or top, and to attenuate other less desirable frequencies from the mix. You want to use it with compression and other tools to build your sound space.


I think the problem is mostly in the naming. It's less of an *equaliser* in the maths sense but more of a *wibbly-wobbly-thinger*.

Admiral Patrick

I'm no audiophile, but I always adjust the EQ based on what I'm listening to and optimize it to my preference. Sometimes it just needs an adjustment to a specific frequency range (reduce the treble a bit, boost the midrange so the vocals are more apparent, etc).

A lot of the EQ presets will apply symmetrically, though. e.g. "Rock" on my phone's EQ will boost the lowest and highest frequencies equally while slightly reducing the middle range. The "Pop" preset does the exact opposite.

My non-audiophile advice is to just set them to what sounds best to you.


Whatever you want, babe. There's no wrong way to EQ.

You are not making an objective change.

irotsoma , edited

Best practice varies based on the specific use case, but generally you don't want to modify sound any more than you have to. The more you modify, the more it has the potential to distort the sound depending on the quality of the particular amplifier and other components. It can get complicated, especially with hardware/analog EQs because of physical quality of the wiring, connections, and components. So, in general, it's best to leave most things at 0db and change only the things you want to change. However, there are exceptions. Like if you want to change all but one of the frequencies by -4db, you're better off changing the main amp by -4db and increasing just the one you didn't want changing by 4db this the EQ is only modifying a small amount of frequency and the rest is passthrough. And generally the main amp is going to have less distortion than the EQ.

That being said, this is talking about cases where your tolerance for distortion is extremely low or you have a really shitty EQ, in which case it's probably better to just throw it out and forget the whole thing 🤣. Because most people aren't going to notice the difference of using the EQ for amplification or the main, even if they aren't the best quality.

So, if you really are as picky about stuff that doesn't matter much, but just want things to be as perfect as possible, like me, balance towards 0 as much as possible using both the EQ and the main. Otherwise, do whatever, probably no one will know the difference.

Also, I know the math isn't exactly right for the -4 and +4, but there are other things you need to know to get it perfect anyway for one inline -4 to equal another inline +4. But it's close enough....


No. It doesn't work like that at all.

You might as well turn up the volume knob to gain back the lost amplitude. That will maintain the mix that you just set to your liking. Just set it as you like it.

However, if you do boost the frequencies a lot so the signal starts clipping, then it begins to make sense to adjust the faders in relation to each other until it stops clipping and still have the "shape" that you like, and then use the volume knob afterwards again.

For instance, if you like a lot of bass and turn up the bass, then it'll likely clip. It might be better to turn everything else but the bass down and then boost the volume.

This is mostly an issue for the bass area. Our hearing is (logarithmically) less sensitive to low frequencies, so in order to turn up the bass we have to make it much louder than if we want to turn up the treble. The bass easily takes up the entire "headroom" available in the signal, resulting in clipping before it is amplified. The rule of thumb is that cutting is better than boosting.

Anyway, unless you're compensating for a bad speaker or similar, it's generally best to leave the EQ alone. Professionally produced music is already mastered to utilise the entire frequency spectrum in a balanced way so that it can be safely turned up without having certain frequencies dominate the output or to turn it down without losing the details.

Using an EQ post production is somewhat like salting a gourmet meal. Chances are that it's making it worse unless you know why you're doing it.

Obviously you can listen to music however you want, but please pay attention to what happens when you turn up the volume. It's likely that you'll want to use less EQ as the volume goes up.

cheese_greater [OP]

For sure, I'm experimenting with it for purposes unrelated to simple listening and appreciation purposes, more context and ambience related


I'm not sure what you're up to but if you are using software to do it, perhaps a multiband compressor would be more suitable than an EQ.

cheese_greater [OP]

I'm using something like that, does all kinds of multiple bands shenanigans


Interesting question. I’ve always used my ears because you can’t just plus and minus frequency ranges to get the sound you want since some frequencies may sound louder than others due to a number of factors (environment, hardware, ears, physics, etc).

I like to think of a flat EQ as 0, and you can add or subtract frequency ranges for taste/needs.

You really only want to be concerned about a sort of 0 level if you’re overall volume is clipping/distorting and you can ring it down for some reason.


On my phone, I can't set the volume high enough for things to be audible via the phone speaker, due to stupid OS limitations.

So, to mitigate that, I've pushed all the sliders in the equalizer to the top. It doesn't sound any different, just louder.

That's also what happens, if you don't balance the numbers. It's just overall slightly more or less loud. And the numbers for volume are completely arbitrary anyways, so no need to worry about it.