Really short question for you guys mastering. How much do you cut completely off the high end? I know all about the low end, but I am unfamiliar with the high...are there frequencies not detectable to the ears for high like there is for Low (Under 30 hz)?

 

Yes. I can't hear over 16k & i mix mostly in 48k.

DEPENDING on song, if there are glitches, say a sampler/synths starts aliasing in the HFs, i might apply a gentle rolloff at anywhere from 18k-23k, but most of the time i don't.

Since you probably can't hear >16k, it's mostly an aesthetic issue (unless there are exceptional circumstances where HFs start impacting the mix), but it might be 'good practice to', & might give you 0.1dB of headroom.

If you want to start ROing, i suggest 20-21k.

 

Last edited: Jul 29, 2016

Maybe it's just my ears, or the music I'm making, but I've found that some high-end frequencies sort of "amass" and complicate each other, like low-end frequencies do, and especially after compression, though to the effect of stridency. I try to roll off the frequencies not primarily characteristic of the element (and therefore superfluous). For example, I find at what frequency range the hi-hat is most active or necessary, and roll frequencies off around that. I'm not concerned with frequencies inaudible to the human ear, but with frequencies which interfere with the audibility or intelligibility of other elements of the mix.

 

As long as it sounds good to my ears, and it looks OK with a spectrum analyser? when it comes to mastering, I will seek to keep the dynamics as close to the original as possible. Just make it louder.
Cutting or boosting (at mastering stage) just because you think there ought to be some, is wrong in my eyes.

 

I will sometimes roll off at 17, if the sequencers are hissing in the high end, though I can't here over 14 myself. As you age the minute hairs in the inner ear weaken and break. Unlike birds, they don't grow back.

 

Dave Gamble from dmg audio recommends to use his dmg equilibrium with Butterworth bell/peak at highs. I tried this, and it was really amazing. Since that, I use only this eq and this method. Because I really can't find better method to smooth most harshness at highs. This really works in >90% cases.

 

Depends.
I EQ each channel individually LP/HP and then check the busses and then you can still play with with mixbus or master LP/HP.
I dont go over 14k.
Cheap software EQ has a hard time taming the high frequency range,may sound fatiguing even in the 11-12k range.
With quality EQ like AIr EQ you can really get so much more out of your material without sacrificing the low/high end.

 

Last edited by a moderator: Jul 29, 2016

well, in the low end it makes sense to cut to free some energy. even if that practically means I'm cutting frequencies that are not directly perceivable for me.

mp3 encoding obfuscates what others do, that is why mp3 is bad for reference tracks.
Why I say that? that could be where this method derived from, I don't know.

But technically it does not make sense, Mp3 encoding is extremely clever.

If you add a low-pass, you're actually putting a low-pass on top of another low-pass
and we know every process has it's trade-offs, even with a deluxe super linear zero cross-point silky eq

essentially it isn't about good or bad but knowing and understanding what happens

 

I do lots of eq and tons of roll offs, notching, on the single channels, then by the ends I cut the low freqency and do a subtle down slop on the high end to kill some of the air.

 

When I mix sounds. I usually cut a fair amount of a lot of sounds, or at least roll it off gently with a less sharp slope, but still cutting the high end out. I usually cut kicks and stuff for the high end off around 15k. The kick sometimes sounds 'worse' and the overall character is hurt, but when I add the snare, hi hats, etc... the kick doesn't seem so hurt because the other sounds just eat the 15k+ range up easily. So I have to cut a lot and slope a lot of 15k+ on many sounds off. When mastering i can boost the overall high end, but when mixing it's all about making sure they don't muddle up with each other like frequencies around 50z.

Sometimes i will use a bus a sound use a steep low cut all the way to like 10khz, then add a reverb on the same bus and make sure the high end of the reverb is clear and then add another EQ and just tame the high end reverb. Put it fairly low in the mix and that usually compensates for any loss in the high end and adds some, almost, glistening air the the mix.