Tracking volume shootout, theory being put to the test

[John_C]

What i think my little experiment shows, is that just as long as you don't track stupidly low, it makes pretty much NO AUDIBLE DIFFERENCE.

And yes, i know i didn't raise the absolute noise floor, I simply created a shit SNR that manifests as a raised noise floor after normalisation.

Some day when i have time i'm going to do a really exhaustive experiment on tracking levels, bit depths, sample rates and what audible effect they produce, both on single tracks clean, summed tracks clean and then the equivalent with plugins as well, and maybe some outboard. Hell, maybe i'll even try and get it published somewhere. "Digital Audio and Human Hearing"

 

[Force666]

Yeah, so in my opinion, tracking in the range of -3db / -9dB is totally ok.

Not really, it makes little sense to do so. A few tracks at -3 = clipping. Why track so hot and turn all your faders down to mix? YOu can just track at moderate levels with no effect on sound.

 

[John_C]

Not really, it makes little sense to do so. A few tracks at -3 = clipping. Why track so hot and turn all your faders down to mix? YOu can just track at moderate levels with no effect on sound.

The thing is, although the noise floor might seem really low on my samples, imagine what happens when you have 50+ tracks running, all with that noise floor

 

[Force666]

But are you talking the -18 Average? Or the -60 peak? A -18 average will usually have peaks -10 to -6 either way. Regardless, as mentioned before anyone mixing on a desk uses these reference levels. So if it's good enough for some of the biggest pros in the business it's good enough for me.

 

[John_C]

But are you talking the -18 Average? Or the -60 peak? A -18 average will usually have peaks -10 to -6 either way. Regardless, as mentioned before anyone mixing on a desk uses these reference levels. So if it's good enough for some of the biggest pros in the business it's good enough for me.

I was talking about low (but not ridiculously low) peaks. I'm not really trying to prescribe what people should do, just providing them with some concrete information to work with and some samples to illustrate what is audibly important.

 

[abt]

the left hand side corresponds to the fader control, in a measurement of decibles

meter read out is on the right

in other words, looking at the numbers next to the fader while watching the meter is gibberish, it doesn't mean anything. the numbers next to the fader are there as a label to read where in decibel addition / subtraction you are with the fader

Thanks Joey. Please ignore me, I was completely out of my brain when I wrote this. No, not on drugs, I was sick for a few days and had a temperature when I wrote this. I had completely forgetten about it until now. All I remember was that I was sitting at my desk thinking "12 - 7 is.... 12 - 7 is .... 12 - 7 is........ I know, I need a calculator. What the fuck 12 - 7! I'm going to bed"

Eventually I did workout what 12 - 7 is. (4 I believe )

Don't worry, I'm sure in time I'll be back, completely lucid, and ask an equally stupid question

 

[colonel kurtz]

I hear a lot of people talking about the 'sweet spot' of the preamp gain and for me, this thread is more about that. And I really don't thing there's any audible difference between tracking around -5 and -10dB in terms of sound.

+1000

1)keep your pres around 0dbvu peak - unless you have something that sounds especially awesome being pushed into the red

2)calibrate converters so that 0dbvu gives you a good 16-18db of digital headroom

3)stop arguing about shit!

 

[Deaf Ear]

I thought the difference was less math at the mixbus and not overloading it. Doing one track at two different levels should only show preamp coloration differences if anything. I started tracking at lower levels and noticed it was much easier to get the mix into shape, and they they improved dramatically. It would seem nearly impossible to do a real world a/b test on the basis of this theory, but it seems to work pretty well imo.

 

[Ermz]

+1000

1)keep your pres around 0dbvu peak - unless you have something that sounds especially awesome being pushed into the red

2)calibrate converters so that 0dbvu gives you a good 16-18db of digital headroom

3)stop arguing about shit!

I was under the impression that the headroom beyond 0dBVU is intended for peaks. For me the core of the signal lies around 0, and the transients overshoot. This way tends to make more sense to me, because tracking the highest peaks at -18dBFS in a digital system seems a bit wasteful and overly conservative. Not only that but pushing big transients into the red may help tame and round them off. All things conducive to an easier mix process.

 

[Metaltastic]

Not only that but pushing big transients into the red may help tame and round them off. All things conducive to an easier mix process.

You know, that's a damn good idea (well, for drums anyway) - not something I'd wanna try with cheaper stuff, but I'll keep it in mind when my ship comes in

 

[Force666]

I was under the impression that the headroom beyond 0dBVU is intended for peaks. For me the core of the signal lies around 0, and the transients overshoot. This way tends to make more sense to me, because tracking the highest peaks at -18dBFS in a digital system seems a bit wasteful and overly conservative. Not only that but pushing big transients into the red may help tame and round them off. All things conducive to an easier mix process.

Totally correct. The signal should AVERAGE at -18 dbfs (=0dbvu). An RMS meter is perfect for this.