Low Cut increases volume?

[jasper]

So, this drum has a -5 db peak. But when I put an EQ8 on it, and CUT everything below 100 Hz, and cut everything above 19kHz,
the signal no peaks at -0.2 db.

Why does cutting frequencies out increase the peak volume of this sample??

note: This occurs or the Q on the EQ bands has no effect on this, whether the Q of the EQ bands is the default 0.71, or Zero,
the volume of the audio signal increases when frequencies are rolled off or cut.

(sorry, no time for video evidence).

[Tarekith]

Not all filters are perfect, it is totally normal to have a slight bump in volume above the cutoff of point, usually it's a larger bump the steeper the filter slope too.

[jasper]

Yes but when the Q (resonance) is at Zero
and the frequency is rolled up to 150Hz
and it’s a kick drum.
I was just wondering why the resulting
volume is 3dB louder.

[Artcutech]

from my experience doing cuts always results with an increase in volume, I don't think its an Ableton thing(the person I learned from was on logic using a proQ2) so when using your ears and cutting until you can hear an audible difference then pulling back as not to effect that actual sound, or even if you do effect the sound, you always get a boost in volume because you've cut the spectrum and now it's focused on a smaller specific area(obviously unless you cut so much that there's barely any audio, but I think you know what I mean), thats always how I've seen it but it was explained in much more detail and properly when I was learning about mixing, cutting at 25hz or 30 hz on a kick is gonna give you a louder kick, because there's less going on, or lets say a pad, if you cut frequencies where there are none up until the point there are you will still get a boost in loudness as the excess spectrum no longer is relevant

drop a mastered track thats been limited at -0.00 db just as an example, your showing -0.00db, now put a proQ2 on the end of the track with a 50hz cut, you would think it would be unaffected or even quieter, but it'll actually clip and go over -0.00 db because it makes the track slightly louder(in this case not in a good way) and of course the limiter in this case has no say so but doing a frequency cut you wouldn't think should make anything louder but it does

[Tarekith]

Yes but when the Q (resonance) is at Zero
and the frequency is rolled up to 150Hz
and it’s a kick drum.
I was just wondering why the resulting
volume is 3dB louder.

Like I said, this perfectly normal for most high and low pass filters, even when the resonance is 0. The shape of the filter curve is not the same as the image you see in the plug in window.

[pottering]

It is probably the Adaptive Q setting in EQ Eight.

[EDIT: No, I was wrong. Don't know what it is. Probably what Tarekith said.]

[Artcutech]

dude I feel like an idiot, I thought you said .3db and not 3db, please ignore my previous post as far as addressing your problem

[jasper]

Artcutech, your answer makes the most sense.
It does happen with other EQ’s and softwares.
Kind of a counterintuitive thing that when you
remove some of a signal the overall increases
its peak. Thanks.

[jasper]

Interesting...

[jlgrimes]

So, this drum has a -5 db peak. But when I put an EQ8 on it, and CUT everything below 100 Hz, and cut everything above 19kHz,
the signal no peaks at -0.2 db.

Why does cutting frequencies out increase the peak volume of this sample??

note: This occurs or the Q on the EQ bands has no effect on this, whether the Q of the EQ bands is the default 0.71, or Zero,
the volume of the audio signal increases when frequencies are rolled off or cut.

(sorry, no time for video evidence).

What's interesting is low cuts are also used to remove low end mud particularly subsonic freqs that wont be heard on most systems which can give you more headroom to make louder tracks with minimal distortion.

[Artcutech]

yup, definitely sounds like whats going on

[kparagraphic]

if you resample it you might find that the overall volume isn't really increased, there is just a short peak / bump created when the kick drum starts. i'm no expert but i'd say it's something along the lines of the filter algorithm adjusting to the sudden onset of sound.

[fishmonkey]

there's multiple reasons why a subtractive EQ filter might increase the peak level of a signal.

- an EQ changes the shape of the waveform, and the resulting waveform might have a higher peak level.

- the EQs most commonly used in mixing introduce frequency-dependent phase shifts to the signal, and changing phase relationships will change the way that parts of the signal interact with each other, and sometimes this will be additive, increasing peak level.

- some filter designs introduce boosts above the frequency that the filter is set to.

https://www.soundonsound.com/sound-advi ... eem-louder

[Hyperwind]

Same thing. Even with the default Q which doesn't have resonance peak on that graphic. Lowering Q helps getting rid of raised volume. Which makes you think whether this "neutral" Q is actually resonant.
Happens everywhere, not only on transient sounds such as kicks or bass but also on pads, stable leads etc.

[beatswilsonian]

Phase shift. Common problem on any material with low freq content. Will happen on any non-linear EQ you use. You'll find you have to move the cutoff pretty high to avoid the spike in peak level. -6db slope is basically the only HPF that won't introduce a pretty noticeable bump in peak level.

2 solutions, neither of which is perfect:
1) Linear phase EQ. Problem: Pre-ringing. For a static kick, what I personally do is print the high passed kick and manually cut off the pre-ring that you'll see when you zoom in.
2) Clipper/Limiter: Problem: Source integrity. Will change your sound slightly to moderately depending on settings and how hard you hit it.

TDR makes Infrasonic with some settings/modes to address this problem, but this is very sub-specific as the cutoff only goes up to 50hz or something like that. Even Infrasonic will introduce a small bump in peak levels, but not the 3db you're seeing.