When mastering, where should i cut off in the low end?

[tone-def]

Nice tip livecollective.

[AK]

Yeah, totally get that mate and, like I say though, to be honest I have never done anything except work on the parts in the mix, I am not, don't wanna be and have no desire to be, a mastering engineer of any description. Getting the mix right is where it's at for me and I see no reason to work at the mix as a whole but thats just me.

[steevio]

I've heard some people say it should be around 30hz on the master-channel? But to me it seems like it doesn't make a difference? And should i do it with an EQ or the auto-filter (ableton)?

when you are working with frequencies below 30 Hz, you are starting to head into the territory of inaudible frequencies, we cant really hear much below 20Hz, so you may think you arent making a difference, but by removing very low frequencies you are allowing much more headroom for bass frequencies in the audible octaves above, which gives your bass more definition.

saying this, i never cut-off these frequencies myself, i just dont put them there in the first place.

i start with the fundamental tone of the track first, whether it be coming from the kick or the bass and work upwards.
so say my lowest bass note is E (41 Hz) there will be nothing below that anyway, so whats the point of cutting at 30 Hz ??

the only time you should be considering cutting low frequencies, is if you dont know whats going on down there, like say you used some sample for your bass which has undefined sub frequencies in it. If you have some unruly subs warbling around down there from samples, its best to spectrum analyse your track and identify the frequnecies that shouldnt be there and remove them, but there is no specific frequency at which you should cut, every track is different.
rules of thumb are useless in mastering.

if you synthesize your bass, you should never have to do this.

[ray parry]

I have no idea about the harmonics on say a Piano but if I'm creating a synthesized sound, I imagine the lowest note I hit, is the lowest note and that there aren't frequencies below that. Well that was and has been my understanding all the time.

All sounds have harmonics in conjunction with other sounds.

These relationships between wavelengths and frequencies of the various harmonics for a guitar string are summarized in the table below. If a frequency of 400hz is present it has the ablity to effect the volume of frequencies above and below it regardless of where it was recorded or how if the note is happening at the same time in phase with the note. For example going up on the frequency scale a note at 400 hz would effect: 800 hz, 1200 hz, 1600 hz, 200hz.....etc...
That can happen going up or down. This is why rolling off your low end can be detrimental to the mix unless you have a real problem down there. You need to know the math and the theory before you do stuff like that otherwise you can fck up and entire mix and feel of a track by applying a blanket solution. This is why in order to be a real pro engineer you need to have a good working knowledge in math and electrical engineering. That's why they're called engineers. Pro engineers have a name for people that call themselves engineers that just have a good ear but no technical knowledge of the theory, they call them knob twiddlers. Knob twiddlers can come up with some amazing mixes and a good ear can bring you a long way but a great ear and working knowledge of the theory can win you grammys.

I had a long talk with Pole about mastering during the day at mutek a few years ago, and he said the most important thing about eq'ing while mastering (this doesnt include removing harmonic distortions) is to recognize (just as Torque said) the relationship between frequencies. That often the best way to boost say 100hz, would be to suck a db or at a frequency on either side of the tone you want to boost. This creates a psychoacoustic phenomena used by lots (if not most) successful mastering engineers. Seems like a basic idea, getting it down pat is tricky but this single piece of advice really opened up what I was able to do mastering other peoples work.

so no boosting 100hz? just take away freqs around it?

[kdgh]

is it also the other way around?!

let's say.... you want to dip 100hz... you boost let's say 90 and 110hz?
i know for EQ's (mostly software) it's better sounding when you're cutting then boosting, though the algorhithms are getting REALLY really good lately.

[steevio]

also a well known mastering technique;

if you want to cut a frequency which is causing problems, use a very narrow Q filter to surgically remove the frequency, but at the same time use a wide boost around the same frequency to lift the frequencies around it so that it sounds natural and avoids creating an obvious hole in the spectrum

[hydrogen]

This is very interesting to me... about cutting frequencies... but a wouldn't narrow cut naturally create resonance in the surrounding frequencies by itself? sort of like a free boost?

[steevio]

yes but it sounds unnatural.
a very wide gentle boost makes it almost invisible