Question about frequency and volume

[Justin Bays]

I've been wondering this for a while...whenever i highpass something i actually lose headroom, why is that? Like if something is peaking at -5 dB, it will be peaking at -4 or -3 when i cut the bass out of it for example. Why is that?

Also I've noticed that low sounds peak higher on a spectrum analyzer..like if a high sound and a bass sound are peaking at the same dB in a meter, the bassy one will always hit higher up (vertically) on the analyzer, was wondering why that happens too

[Widowmaker]

might be because your resonance knob is turned up, usually resonance adds some dB's

what plugin you using

[AxeD]

might be because your resonance knob is turned up, usually resonance adds some dB's

what plugin you using

Seriously?


For lower frequencies to be audible they need to be louder.
So a lot of the volume in music is in the low end.
Has more to do with physics than with whatever plugin etc..

Dunno about visual spectrum analyzers, as I don't use any.

[press]

whenever i highpass something i actually lose headroom, why is that? Like if something is peaking at -5 dB, it will be peaking at -4 or -3 when i cut the bass out of it for example. Why is that?

you are gaining headroom in this scenerio. as per expectation.

[Justin Bays]

might be because your resonance knob is turned up, usually resonance adds some dB's

what plugin you using

Seriously?


For lower frequencies to be audible they need to be louder.
So a lot of the volume in music is in the low end.
Has more to do with physics than with whatever plugin etc..

Dunno about visual spectrum analyzers, as I don't use any.

i think you actually answered my 2nd question about the spectrum analyzer :p it makes more sense to me now

This is the part I'm confused about..if a lot of the volume is in the low end, then shouldnt cutting out the low end reduce the overall volume of the sound? I just played a saw wave at G2 - it peaked at about -10 dB..and then i cut out the subs of the sound by putting a 48 db per octave highpass at 100 hz and it was peaking at about -6 dB which means i lost 4 dB of headroom which means the sound is now 4 dB louder doesnt it? I was thinking it is maybe because all of the frequencies in the sound are getting focused in one area which is causing a loss of headroom but i dont really know

[Justin Bays]

you are gaining headroom in this scenerio. as per expectation.

but i'm talking about digital audio, where 0 db is the ceiling

[Widowmaker]

might be because your resonance knob is turned up, usually resonance adds some dB's

what plugin you using

Seriously?


For lower frequencies to be audible they need to be louder.
So a lot of the volume in music is in the low end.
Has more to do with physics than with whatever plugin etc..

Dunno about visual spectrum analyzers, as I don't use any.

seriously what?

was just offering an opinion sorry if it was "incorrect" no need to chat to me like i'm a moron mate.

[press]

you are gaining headroom in this scenerio. as per expectation.

but i'm talking about digital audio, where 0 db is the ceiling

LOL, fucking negative numbers. read it like ten times and still thought "im not seeing the problem here". my bad. this is indeed a peculiar situation, cant say ive seen that myself. resonance could be the cause in my opinion as well, nothing else really makes any sense to me. Are all your eqs and filters doing this?

[Justin Bays]

you are gaining headroom in this scenerio. as per expectation.

but i'm talking about digital audio, where 0 db is the ceiling

LOL, fucking negative numbers. read it like ten times and still thought "im not seeing the problem here". my bad. this is indeed a peculiar situation, cant say ive seen that myself. resonance could be the cause in my opinion as well, nothing else really makes any sense to me. Are all your eqs and filters doing this?

i usually use fab-filter pro-q but it happens with all of my EQ's...i'm using ableton live. I just use a highpass with no resonance or anything just flat.. it's weird :/

[press]

you are gaining headroom in this scenerio. as per expectation.

but i'm talking about digital audio, where 0 db is the ceiling

LOL, fucking negative numbers. read it like ten times and still thought "im not seeing the problem here". my bad. this is indeed a peculiar situation, cant say ive seen that myself. resonance could be the cause in my opinion as well, nothing else really makes any sense to me. Are all your eqs and filters doing this?

i usually use fab-filter pro-q but it happens with all of my EQ's...i'm using ableton live. I just use a highpass with no resonance or anything just flat.. it's weird :/

well the only other dumb work around i can think of is use a shelf to bring down all the frequencies up to the highpass cut off point to bring it all back down the decibel or two you need.

[AxeD]

Yeah that was kind of uncalled for Widowmaker and I'm actually partially incorrect because I read the op wrong.
I figured he must have meant the level went down and he didn't understand why exactly

The level can't go up when you're only cutting off bass though, that's impossible.
Isn't the filter inverted or something like that? Because this goes against logic.
Unless it's a dodgy piece of software, but I doubt that too.

[Justin Bays]

Yeah that was kind of uncalled for Widowmaker and I'm actually partially incorrect because I read the op wrong.
I figured he must have meant the level went down and he didn't understand why exactly

The level can't go up when you're only cutting off bass though, that's impossible.
Isn't the filter inverted or something like that? Because this goes against logic.
Unless it's a dodgy piece of software, but I doubt that too.

Yeah it happens with every single Eq and filter i try :/ it doesn't happen with all sounds but most of them... I just tried it with a saw wave playing a G2 and i lost like 3 db of headroom when i cut out the bass with different EQ's and filters.

I have a feeling it has something to do with masking..like the frequencies in the sound are getting stacked because it isn't being spread-out across the whole spectrum..i dont really know though

[VirtualMark]

It's probably a limitation of the EQ you're using. I don't remember exactly why but EQ's with a steep slope can actually boost a bit at the cutoff point. I know digital EQ's use FFT and that they're not too accurate with low frequencies.

Try a more gentle slope, say 12db instead of 48, see if you get the same results.

[stompzi]

The level can't go up when you're only cutting off bass though, that's impossible.

Actually, it can and often does.

When you cut or boost with your EQ, you're not merely applying a gain change to a specific set of frequencies, you're introducing phase changes which can indeed make the peak of the signal louder.

[Justin Bays]

When you cut or boost with your EQ, you're not merely applying a gain change to a specific set of frequencies, you're introducing phase changes which can indeed make the peak of the signal louder.

I think this is why


I tried it with a 12 db slope in pro-q and i still lost headroom

[drake89]

It's probably a limitation of the EQ you're using. I don't remember exactly why but EQ's with a steep slope can actually boost a bit at the cutoff point. I know digital EQ's use FFT and that they're not too accurate with low frequencies.

Try a more gentle slope, say 12db instead of 48, see if you get the same results.

I think this is it. try turning the Q way up on a HPF or LPF and you'll see visually that the 'crossover' frequency or whatever is at the top of the slope is boosted. It could also be a limitation of the VU meters. I still don't fully understand them especially since they work differently in the analog and digital realms. and they work different than our ears and hardware.

[stompzi]

It's a phenomenon that occurs with EQ/filters in general, not just the one he's using or only digital

[AxeD]

Hey thanks stompzi. That's interesting stuff. Seems like it's more of a problem with analog eqs but it occurs
in digital too.

I guess this would have to be the problem in ops case. But phase is kind of unpredictable or at least a bit tricky to calculate.
It would be quite a coincidence if it the phase difference is always either destructive or constructive isn't it?

[Triphosphate]

Oh yeah. Has this test been tried with a linear or graphical eq?

[Turnipish_Thoughts]

Yeah the filter/EQ boosting when cutting phenomenon thing. It's been explained already.

Bought something to mind though. afaict cutting out partof a sound won't necessarily lower the DB, peak amplitude is frequency specific isn't it? i.e. only if you cut the loudest frequencies in a sound will the peak amplitude drop. Cutting the tops out when they're lower than the lows won't actually change the amplitude of the sound at all, it'll just alter the timbre, right?

I might be completely wrong though...