(24-05-2015, 09:57 PM)pauli Wrote: [ -> ]+1. I've come to realize that if I'm trying to broadly change the character of a mix during mastering, something's amiss in the mix itself. Lots of internet advice suggests a linear phase high shelving boost at mastering for "air" and "sheen" and a mastered sound. I think that usually only works if you're using really high quality mastering processors... lots of stuff between 15 and 20kHz, even the the junk I can't readily hear above 17kHz... well, I usually find it annoying, so I avoid boosting the trebles like that with the free/cheap processing that's available to me. Lots of folks think digital EQs sound brittle when boosting the highs.
+100. taking a step back...... i'd add to this that trebles are normally a mess before getting to the Mastering stage (not forgetting bass also, of course). if the mix contains some harshness from one or more instruments - undetected resonances, wrongly applied EQ etc, then simply touching the top-end in order to try and make a mix sound airy can turn it into a train wreck, Maag or no Maag. far too many people bring their mixes to Mastering with excessive trebles [for a multitude of reasons] and it leaves the engineer with nowhere to go except reduction/roll-offs. i insist on stems.......if the problem is only in one stem, say, it can be treated without having to compromise on the other material. you don't want to be treating sibilance in a stereo file if everything else is ok, for example, because it will hit the cymbals [if present]. but in the forum, if newbies are mixing, newbies are mastering, so all the inherent issues get exacerbated; nobody can Master in a bedroom on a pair of 2-ways, but it doesn't stop um trying. the first problem being the fact that they will have missed problems in the mixing process because of the issues inherent in their listening environment, and these will also be present in the master. actually, because of the inherent limitations in any one pair of monitors, it's better having 2 or 3 pairs of monitors for mixing....switching shocks the ears, for one thing. a crap mix can sound great after 20 seconds, because our brains adjust to it. flicking to a different monitor is perfect for [potentially] avoiding such perils. i think many problems which continually plague people's mixes consistently over time, are partly for this reason. those who haven't had the luxury of working under this sort of environment (multiple monitors) won't understand the benefits, nor the benefits of correct acoustic treatment of a room. you gotta experience it to appreciate it. and unless we experience the benefits, we are far less likely to embrace change.....assuming the budgetary resources are willing, of course. and if they aren't, progress in our mixing skills will be limited - we will get so far then simply end up going around in circles repeating the same mistakes because the fundamental issues haven't been adequately addressed (previous discussion in a Pauli thread refers, as much as it probably does to this thread also). by all means see how a mix translates over other playback devices, but it's not a panacea for critical listening environmental issues. when we dial-in a parameter, we must hear it's consequences immediately in order to assess it's appropriateness and work the element in with the mix as a whole. but if our judgement is impaired for reasons i keep harping on about, the mix will be a reflection of our personal limitations; be they financial, academic ability, Terms of Reference......... or woteva.
Quote:+5 Side note, not really related to this production, but worth thinking about: A lot of the times it's an arrangement issue more so than a mix issue IMO which really sucks... the slice and dice associated with addressing things like that is a real drag. If the majority of the fundamentals are occurring within the same octave, the production will become effectively unmixable. When you have to highpass and/or severely attenuate the fundamentals across most of the instruments in the mix, you're gonna cut the balls off the mix as a whole and wind up with too much high mid. But then if you try to pan your way out of it, you wind up with too much non-directional frequency content on the sides... which is generally bad both artistically and scientifically, unless you have a specific reason for doing so.
When you think about it, low shelving the bottom half of the spectrum isn't much different than boosting the top half after you volume match.
One more little bit... electric guitars by and large don't produce any musical harmonics above 5000 kHz... just noise... and the ratio of analog noise to harmonics starts getting unwieldy around 4000. The potential ugliness gets a lot worse if the signal is overdriven pretty hard, too, because the distortion is adding harmonics to the noise as well as the good stuff. The takeaway is that you can get away with lowpassing electric guitars most of the time, but a gentle lowpass, 6dB per octave, is usually best, since steeper curves usually create a sharp resonance at the corner frequency and that's just going to make things worse.
You can clear out a lot of the fatiguing crap in the treble by attenuating or removing it from electric guitars... acoustic guitars are another matter altogether and the two of the same make/model acoustic guitars can sound quite different in the highs depending on string tension, technique, how often the instrument is played, how old the strings are... it's mind boggling. But electric guitars are much more predictable in the highs-- you don't need em. Save that stuff for the vocal and cymbals. Might sound a little dull after the guitars have been sizzling your ears for a few minutes, but take a break, come back and you'll see what I mean.
hahahahhahah......oooooooh yeahhhhhh.
i've since mixed this. the prior-to applied compression wasn't correct on the vocal during tracking (OR SUBSEQUENT PROCESSING PRIOR TO IT BEING PRESENTED IN THE LIBRARY?!!!!) and Matt's processing has, it seems, worsened the situation.
the fatigue is coming off the cymbals mainly, but i didn't spend much time hanging around listening, for reasons stated. in the raw tracking, the drums are good however....
....but especially note:
the vocal tracking already contains level automation - somebody forgot to disengage it prior to uploading the multi!!!!!!!!!!!!! so, if we apply compression in order to address imbalances, it will simply add more distortion on what is already present. so any recommendation regarding dynamic processing (Pauli's earlier reference refers) needs to be taken with extreme caution, therefore. the opening lines are around a typical SL of -20LUFS, give or take, while at the end it's -14LUFS. i'm using values here for the benefit of the community because they are objective and i think it's important to use objectivity to get the message across in an otherwise potentially subjective discussion, which loudness can be. further compression could turn the vocal into a dogs dinner........and the Tee's will bite, even more than they do in the raw track! i can't remember the intimate details of what i did with it now because it seems like it was 100 years ago....but i didn't compress on the vocal track at all, nor did i parallel it.
apart from my irritation at the lack of care in supplying the multi and it's inherent attributes, the material works well. but i think it would help the forum if Mike Senior insisted that all automation/processing is removed before supply. it makes the process of mixing a complete crapshoot otherwise and a total waste of time in my opinion; Skelpolu's Entwine was in my opinion, a complete disaster, by way of example. there are way too many others!!