Mixing Down?

Matthew Walsh

New member
Okay ... here goes. I am just starting out in multi-track audio recording with the intention of cutting an album-length project out of my home. Thanks in no small part to this site, I have squashed a great many of the bugs in getting up and running, and am getting a great and fairly hot sound from my mixed-down AIFF's (I am recording and mastering on a computer). I am almost ready to begin laying down all tracks, including vocals (my AIFF two-track tests have all so far been everything but vocals, just to see what the levels are like and to get the hang of mixing in digital).

However, I'm not automating the mix. What I do is, I go to the spots where I know the music is loudest and mix from that. The levels don't change from the mix ... that is to say, I don't move the sliders up and down through the song. They stay in that one spot.

The resulting sound is great, but am I doing something wrong? For instance, I have a song that leads off with an acoustic, and then drums, a rhythm guitar, a bass kick in over the chorus. Well, when we get to that loud part, the levels are great, the mix is dead-on, everything sounds great, no peaking over 0dB, etc. But in the softer parts, where it's just the acoustic guitar, the acoustic guitar is at that same mix level, so it's softer than the rest of the song. To give you an idea, if you mapped out the waveform of the song, the acoustic is just a straight line, and then, on the chorus, it's a nice, thick, healthy waveform that goes right up to 0dB.

However, if I had another song with no electric guitar, just an acoustic (and I do), and I wanted to stay consistent, I would have to put that acoustic at the same level I had it on the song with full instruments, therefore resulting in a softer sounding song. So now, I have no idea what to do.

I guess my question is -- and I can only tell so much from a "real" CD -- should I be riding the levels to make them consistent? I basically have the idea that I would mix everything as I did before: set the highest mix levels where the acoustic, bass, drums and electric are together. That would establish my loudest levels. Then, on the acoustic parts ... should I bring up the acoustic so that IT is at maximum loundess, then turn the fader back down to its pre-established level for the chorus? Or should I leave well enough alone and not touch the mixer once I have those maximum levels set in one place?

The recording sounds fine. The problem is, I want to make sure I'm doing as much right as possible, and I kinda thought (though I wasn't sure!) that "real" studios would ride the levels and make sure everything was always as loud as it can be at any given plac in the song. Not even this fine and informed site seems to offer this advice on mixdowns.

Thanks in advance for your help!
 
Check out Jethro Tull's "Thick as a Brick"
CD (originally vinyl) and see that the short acoustic guitar intro/ and the section that follows (a vocal/guitar duo) are way below the levels they got through most of the electric section of the piece. Is this the sort of cascading volume you're looking for, with volumes kept really low until more instruments come into play? Then leave that flat line alone. Otherwise boost it, but don't overdo it. The acoustic guitar by itself <should> be a lot quieter than the whole band, unless you <want> the mix to sound unnatural.
 
Well ... I appreciate that, and that answers a good deal of my question, but I think my question should have been more generalized. In general, are you "supposed" to boost the softer parts with the volume sliders to maximum to keep things hot all across the song? Or is it okay just to leave the fader alone (stationary) and let the soft parts stay at softer levels and let the full arrangement be what determines the levels?

For example, I have songs with just piano, bass, drums and strings, and then I have songs with heavy distroted electric guitars and drums and bass, and some with just an acoustic (and all songs with vocals, plus some instrumentals without vocals). It seems to me -- and I could be wrong! -- that on like a regular album, you don't notice a volume difference even though the base instrument might be different; we all know that a distorted guitar song and a piano ballad or a song that has both are going to be mixed differently, so how is this maintained? I hate to sound like an idiot, but I don't think I'm gonna do it right until I admit that, yes, I AM an idiot and I need help!

Don't get me wrong; I'd be pleased as punch to put all the levels in one place and leave them there, since I'm not too shabby at finding a level for everything where all instruments can be heard and heard loud. But I wanna make sure this is an acceptable way of doing it, and that I'm not eliminating an important step in mixing down.

Anyway, drstawl, what you mentioned is what I'm doing: putting each instrument at one specific level, leaving it there and letting the arrangement build a crescendo rather that boosting the faders to make it happen. It seems to sound fine to me (but I'm just a home recordist). Just want to make sure this is how it's supposed to be done! Thanks!
 
drstawl, on your advice, I nabbed my Pink Floyd THE WALL CD and used Toast Audio Extracter to pull "Hey You" off of it as an AIFF file. Imagine my surprise when I discovered the waveform was EXACTLY THE SAME as what I described my demo file being!!! You were absolutely right, and I guess I stupidly answered my own question. Thanks!

When in doubt ... turn to Pink Floyd (didn't have Jethro Tull!) for the answers!
 
if you dont like what you are hearing, try some compression on the overall mix. it is a matter of taste.
 
Back
Top