Originally Posted by arnyk
There is another source of spectral shaping in recordings when played back, and that is the recording itself.
Most microphones used for recording have non-flat frequency response and the departures from flat appear to be somewhat intentional.
There are several physical processes that add a high end roll-off when the microphone is pulled back from the source, which is often done.
not "somewhat intentional" but carefully chosen, to the point that a studio with many differnent old mics with different kinds of sweet sounding departures from flat will use their mic arsenal as a marketing tool. For good reason.
It's a bit like photography. Being a large to pretty doggone large guy depending on when we're talking about, I notice how the two dimensionality of photography changes one's appearance. The old "adds ten pounds" thing is, IMO, all about taking a three dimensional person and putting those bits that are toward the back in a rounded shape flat on the screen/paper.
Similarly, a flat mono recording of even a simple acoustic guitar flattens out aspects of the real sound. Then you get into where you place the mic. The guitar sounds differently to an audience (at least in the good old days of playing in the dorm basement with all your friends purely acoustically) than it does to the player just because it's different from different locations. The same mic at the "point it at the 12th fret 6 inches away" location sounds different than if its nearer ths aoudnhole, over the bridge, oreven at the 12th fret but nearer the fretboard. Typical stereo acoustic guitar micing is not an attempt to capture the sound of the guitar in space, it's an attempt to capture two different perspectives from closer than anybody ever actually listens to a guitar, including the player that add up to something at least interesting.
All the other instruments are the same. A long winded example. To me, the last really good-sounding Eagles album was Glyn Johns produced Desperado. when they went to Bill Szymzyk (whose work with the James Gang I oddly love) they got away from simple micing on the drums, for xcample, and close miced every drum. That's a cool thing to do for a big stadium or large room, because there's not another way to get appropriate gain before feedback, but NOBODY listens to the high tom tom from half an inch away then moves their head with Superman like speed to hear the snare at half an inch then moves again to hear the rest of the roll down the toms from half an inch away from each drum. IN the end,to me they took the soul out of Henley's drumming simply by divorcing it from the real sound of a drummer playing the drums. Only adding to the losses was the abandoment of the "ridiculous" requiremnt that Johns imposed that the band ACUTALLY BE ABLE TO PLAY THE SONG, all the way through, before they recorded it. Once you abandon the idea of a single take (at least mostly) covering the whole song, you're quickly on the slippery slope to each drum hit or vocal note or guitar phrase being individually selected from 20 different takes of the same thing. That kills the reality of the performance (and Yes, I've comped a couple of guitar solos where it made musical sense, and Ive even tossed a great solo because the sound wasn't quite right and attempted to replace it with a note for note recreation. The recreation never had the same "soul," for lack of a better word, as the original. The magic lives in miliseconds . . . .
Anyway, enough rant from me except
If you hang out at mixonline or anywhere else that prefessional sound engineers air their opinions, you'll see that a major issue in mixing and especially in mastering (at least before the single minded focus on getting every fricking sound to within 4 dB of full scale) is portability. You make a mix on your however good they are main monitors, and check it with car speakers mounted on the bridge. And if you don't like what you hear through the crappy speakers you change the mix on the great speakers, at least potentially worsening it in that context in order to assure that the thing sounds pretty dang good wherever it gets played, not just on carefully dialed in (at least hopefully) studio system.
Then you send the album to the mastering engineer. They dotimportant stuff, not least of which at least used to be to take mixes that have different average loudnesses and making them work together in a sequence that pulls the listener in and doesn't necessarily require constant volume knob adjustment. Either way, the mastering engineer for pop or rock or rap almost certainly turns out a product that does not sond to the mixer like the music he mixed. And probably (especially in loudness war affected stuff) sounds significantly crippled to the mixer.
IMO outside of probably classical, some jazz and the intentionally audiophile, minimalist recordings, the whole idea of the process is to create a product that may NOT be great on a great system (at least not as great as it could be) but that translates acceptably to the car, to the boombox, to the MP3 player, etc etc etc. Complicates the whole thing.