question for chrisharris (or anybody, but chris first)

  • Thread starter Thread starter dobro
  • Start date Start date
dobro

dobro

Well-known member
A while back in a thread Sluice started in the mp3 clinic, you talked about keeping everything on its own track rather than submixing. I didn't know what submixing was, but I think I stumbled on it tonight.

I wanted to reduce the number of tracks I've got on a particularly humongous session, so I highlighted some pairs of tracks and did a mono mixdown for them, and then saved the result back into Multitrack. That's a submix, right?

Next question: you figure doing that adds significant amounts of noise, right?

The reason I'd like to be able to this: well, for one thing, it would cut down on the size of these sessions and help my computer to play stuff without stuttering. Not only that, but I'll use, for example, the first third of vocal track 1, the middle third of vocal track 2, and the last third of vocal track 3. It would be so *tidy* if I could just bung 'em all together into one track. :D
 
Well, maybe I should let chrisharris answer first, but in case you need info tonight-

The first task is what I refer to as submixing. Mixing down lots of drumtracks or BGVs to a stereo pair is a submix. The second idea, choosing sections from different takes, is generally refered to as 'comping' (compositing) a track, and is the best approach I've found for a lead vox. I record four or five takes, then use volume automation to have the best phrase sounding at a time, then I mix it down to a new track.

Both use the same function in CEP


-Chris
 
Hmm... okay, does compositing add any noise that you've noticed? I played the composite track and the original tracks and I can't hear any difference.
 
It shouldn't create any noise unless there is noise coming from the outside.. Mics, Guitars, etc. Unless you are recording on Cool Edit Pro Tape.
 
"Why would either method add noise?"

I don't see why it would add noise either, unless there's a dithering process involved in mixing down two 24-bit waves into a single 24-bit wave. And even so, my understanding is that 32-bit floating point means you can do that sort of thing and not have to worry about adding noise to your basic 24-bit wave. Right?

Anyway, that's why I asked Chrisharris about it - he was saying something about submixing adding noise - not the first time you submix, but eventually, after you do lots of other processes.
 
The only time that submixing should really add noise is if you do an analog mixdown. In that case you would be adding the board's noise floor everytime you submixed and then again in the final mix. With digital rendering that shouldnt be an issue and copy and paste comping shouldn't add any noise either.
 
lol...HI DOBRO!! ...sheesh, I'm really on top of things around here, eh?

1. Submixes - Yes, we're on the same page. I used to use a ton of them..., mostly for drums and background vocals, b/c the workspace gets all cluttered when you've got 14 different tracks with like, one crash panned left on track 11, and a little snare roll at -2db on track 12, lol...you get the idea.

2. I don't use them if I can in any way avoid it now. It's not exactly a noise issue like analog, lol (it's more evil). I've done a lot of test mixes recently, and I've done mixdowns of single files (all in 32 bit), and the folks at Syntrillium are less than coherent (to me, anyway) on exactly HOW CEP performs its math, but the bottom line is that highs are getting attenuated, mids are getting smeared, and I finally have a decent enough sound system for the phasing on the lows to totally piss me off, lol. Small wonder that when I posted a tune in the clinic for the "dry" challenge (no EQ, no FX, no compression...no processing whatsoever before or after hitting the "mixdown" button), that the song got the most positive "sonic" feedback of anything I've ever mixed.

I don't know if it's dithering, or truncating, or my converters, or voodoo, but my crap sounds better when I don't mix anything down until I have to, and on my "mastering" now, I'm learning how to accomplish about 3 steps of broadband compression and 3 EQ adjustments TO THE POST processed (i.e., compressed) track in one step.

I honestly don't blame the software, b/c I don't think this is exclusive to CEP. Math errors in digital music (random truncating in CEP) are just part of recording in digital. Have you ever had a file that you're trying to "master," and you'll tweak it with some EQ (introducing phase), then some compression (same problem), but you maybe overdid it JUST in a couple of places, so now you wanna' surgically EQ a few places again...but then before you know it, it just sounds like ass on top of more ass? Well, it's not all due to bad mixing or mastering choices.

(oh, and there IS a dither when you're working in 32 bit, but it's not as bad).
 
You must log in or register to reply here.
Back
Top