Cutting ultra high-frequencies before/during mastering?

[metamorphosis]

Using spectral analysis on a track - I can see a lot of noise in the very high frequency range. I can solo out and "listen" to these frequencies, but by-and-large (save for the occasional crash cymbal harmonics) they are completely inauduble.

Is it mostly just junk up there? Should I clean it up?

I'm thinking it might make for a cleaner waveform before hitting the mastering chain (comp, limiter etc). It's got me wondering about reducing or removing subsonic frequencies too (say, under 10hZ).

Wasting my time? General rule of thumb?

It totally depends on the genre. If you're doing audiophile-style acoustic stuff, then no, leave it in. If you're going for brickwall loudness, you might want to consider getting rid of it.
I personally leave stuff in, I don't like the muffled sounds of a lot of modern recordings, though it suits some environments like radio and crappy computer speakers etc.

[MogwaiBoy]

As a mastering engineer, I HATE, HATE, HATE this trend to low pass an entire mix, especially when people think it helps them get louder songs. A lot of that information up high gives your song it's sense of space and depth, taking that away can make the whole track sound flat even. I can usually spot instantly when a producer sends me a track where they did this on the mix buss. When I ask for a version without that EQing, most of the time people are shocked at how much better it sounds (especially those who want really wide mixes) when they hear their master.

I get doing it for artistic reasons, to make something sound a bit more lo-fi or warm maybe. But unless you have a real need you can HEAR for low-passing the mix, my advice is don't. Horses for courses and all that.

What about the ultra-sonic stuff generally beyond human hearing, upwards of 20kHz?

I should probably pay more attention to the subsonic end of the frequency spectrum - more energy there.

I greatly appreciate the input of a professional Mastering Engineer, thank you

[Tarekith]

What about the ultrasonic stuff? If you can't hear it, who cares? Unless you have some REALLY weird anomaly where you have insanely loud peaks there, it's not causing any harm. Most of that is filtered out anyway if your end result is a 44.1khz wave or MP3 file.

[do_androids_dream]

As a mastering engineer, I HATE, HATE, HATE this trend to low pass an entire mix, especially when people think it helps them get louder songs. A lot of that information up high gives your song it's sense of space and depth, taking that away can make the whole track sound flat even. I can usually spot instantly when a producer sends me a track where they did this on the mix buss. When I ask for a version without that EQing, most of the time people are shocked at how much better it sounds (especially those who want really wide mixes) when they hear their master.

I get doing it for artistic reasons, to make something sound a bit more lo-fi or warm maybe. But unless you have a real need you can HEAR for low-passing the mix, my advice is don't. Horses for courses and all that.

Very, very interesting Tarekith. I'm quite the opposite on this subject. I find I'm lpf'ing probably 1 in 3 tracks because lots of brittle high end is exactly what - I think - *can* make a track sound amateurish and flat. I find that boosting high end is what less-than-confident producers often do to gain loudness, excitement etc. I often low pass at around 17 - 18 kHz with a gentle filter (6dB or maybe 12) to get rid of the irritating build up there (and by irritating I mean potentially ear fatiguing) which means that the track is going to sound great when played very loud - not harsh. Of course, this depends entirely on the tracks content.

[do_androids_dream]

@ Mogwaiboy - I hear very little 'wrong' with your track. There's potentially too much build up in the low mids - probably somewhere in the 300 - 700Hz region at a guess. I would be using eq or a dynamic eq to lighten this area a little and bring it into line with the rest of the tracks frequencies. I master a lot of material like this btw. Low mid frequencies are often a feature of ambient/dark ambient so I would be careful not remove any of the tracks feel in doing so. I might be experimenting with 'glueing' things a little more too.

As regards very high frequencies.. Firstly I don't hear much in the way of high end but if there is a lot of energy up there on the analyser that doesn't really sound very audible there's no harm in filtering it away. In this sort of music it's sometimes clicks in the percussion sounds that occupy this area and if it's not very musical just filter it away. Probably better to do it on the individual track though if you can pinpoint exactly where it's coming from.

[Tarekith]

Like I said, horses for course

[do_androids_dream]

Like I said, horses for course

Definitely - interesting to discuss though.

[Hermetech Mastering]

Yep, no rule of thumb here, every track and project different. I do usually leave my Bax EQ's hi cut filter at 70kHz, 28kHz or 18kHz though (especially if the client wants a more warm/analogue sound, or a really loud master). The first is inaudible to me, but you can for sure hear the second two. Reason to remove it even if it's inaudible is, I would suggest, to get a little more headroom for loudness, and you never know what kinda effect that stuff is gonna have on different playback systems/DACs etc. so better to take it out. But again, all project dependent.

[t3toooo]

Made my hands dirty and mastered this track ,quite like it,if you don't mind (since it was for download).A bit on the rush, i guess i did cut bit too much bass but that's also a matter of taste.

https://app.box.com/s/q8mvfx4b0nop4da7ezncodcb9aqra6tb


(Tell me if you want it removed)

[MogwaiBoy]

DUDE! I don't mind at all. Having a listen now : )

Man the hiss is fierce, isn't it? Check out the super high frequency oscillating in the outro... That just looks like a lost spaghetti noodle on the RX5 spectrogram I don't know how that got there but hey... it's an old song, our first as a group actually - but as you can see there's a lot wrong with this mix.

I'll also post some of my results with mastering it. I'm going to use RX5 to tidy up some of the extreme harshness (plus there is mega low bass content under 20hZ eating a lot of energy for no reason).

Some really interesting spatial/stereo separation happening on your master, man - enjoying it as a point of difference to what I'm capable of.

[camsr]

In mastering, it's situational. In mixing it can be used to give dynamic to things like hats and crashes. I use a slew rate limiter to cut the treble sometimes, it's like a distorting lowpass filter and creates a high frequency density.

[DJ Warmonger]

In mastering, it's situational. In mixing it can be used to give dynamic to things like hats and crashes. I use a slew rate limiter to cut the treble sometimes, it's like a distorting lowpass filter and creates a high frequency density.

Interesting, are there any slew-rate limiter VSTs around?

EDIT: ...or I could make my own in a few minutes

[do_androids_dream]

Some really interesting spatial/stereo separation happening on your master, man - enjoying it as a point of difference to what I'm capable of.

I wouldn't say it's interesting. It sounds like the mid channel is very much missing - something that needs correcting basically in terms of mastering.

Edit - I've just been comparing the original to this master. While it's great that folks are getting involved this really isn't a 'master' of this track. It now sounds like a different piece of music. For a start, the frequency response is entirely different (you must have done some very extreme eq work!). What processing did you do? It's very extreme to say the least.

This isn't the role of mastering - to change things so severely. One has to approach mastering from the point of view that the artist is already happy with the sound of his/her music on the system that they made it on in the room that they made it in - it already sounds like how they want it to sound. Your role as an engineer is to make 'corrections' that give the music the best chance of sounding 'good' wherever it is played and to try to enhance what the artist is trying to say. To make such a drastic change would fail in the role of mastering unless it is some kind of rescue job (in which case this would be discussed first).

[MogwaiBoy]

It is unfaithful to the dynamics/feel of the original track - you could say it's been chucked through the rinser. Instead of saying "you completely destroyed our track!" I said it was interesting... because it is : ) It's not something we're aiming for (uber loud) but it's interesting to hear that take on it. There are some interesting compression pumping effects around the bass pushing the highs away etc

I kept an open mind and appreciated (and was actually flattered by) the fact that someone took the initiative to have a go at it themselves.

[t3toooo]

I added a tad bit more "punch" in the bass/low mid region and i tried to let the rim shot/clap not to be so spiky/prominent although i like the swelling of the instruments up to the point when everything is in one line,loudness wise.
It's super broad but for this music i think it's all right.

Yeah,i describe myself not as a mastering engineer,my aim was to make the instruments more audible and give it some punch and probably some coloration.

https://app.box.com/s/gc8cwbmgs30y84gglm3riayamts4gsl1