Send a tone through the system adjusted to 0dBV
(.775 Vrms) - use a voltmeter to verify this and then connect the voltmeter to the output of each channel and adjust each meter so that all the 0VU levels correspond to .775 Vrms on the output side. That way, when you're at 0db on the meters, your channels are at unity gain.

I hope this is all correct - someone who's an actual EE should jump in and correct anything I've made mistakes on.

Dan Roth
Otitis Media
Audio - Video - Film
dan@otitis-media.net

Here is the concept:

What you do is send a 1kHz tone, at unity gain, from the output of your analog device, to the input of your DAC. The DAC can be calibrated so that "0" from analog land is anywhere between -12 and -18 dBBFS. By sampling at -18, I have 24dB of headroom at the converter. You can go lower to -12 and still get a good sample. Apogee guys say -16. It's up to you.

All converters receive the same voltage, say 5V. So hitting the converter with excessive voltage, more than 5V, does not increase resolution, make it louder, etc.

The concept is to the get the sample over the threshold of the noise of the converter, while maximizing dynamic range. If your music has little dynamic range, then you may get away with increasing the sensitivity on the DAC, but it won't improve the quality much, if at all.

This analogy is not technically correct, but it helps some novice customers of mine: Imagine a sine wave and a doorway. The idea is to get the sine wave centered in the doorway opening, so it all fits, and the peaks/valleys do not get chopped by the jam or the header. That's as basic as I know to tell it.

Some guys compress the crap out of the signal, reducing the dynamic range. 1-bit = 6dB of dynamic range. Over compressing, normalizing, etc will just add noise, decrease the dynamic range in the process, and cause more bits to be used, which is a waste. If something has 72dB of dynamic range, then it only requires 12 bits, so why use 24?

Let it breathe and have life.

[ August 27, 2003, 08:49 AM: Message edited by: sheet ]

Once you have calibrated your system, try to keep the levels at a max of -2 to -3dBFS...this give you a bit of headroom.

Thanks guys,

Mike

I like to shoot for -6dB levels.. This leaves a little extra headroom for processing like eq and compression at mix time. At mix, go for -2dB on the mix bus.. Kurt

mchimes thanks for bringing this up, I've been scratching me head on this.
ok, when you guys say keep the level at say -14dB, are those figures for RMS value or peaks?

So.. Kurt do you shoot for-6 peak ?

I normally try to get around -9 to -5 on the peaks depending on what the source is, and I still have to fight keeping the master from clipping. I do have to lower individual tracks- once I find them on playback. But some tracks need
to be brought down pretty far. Am I losing resolution there?

Thanks

No you won't lose enough resolution to make any difference.. the whole record hot to preserve resolution is a hang over from the 16 bit days.. it's not too much of an issue with 24 bits...

I agree with Kurt. I also try to be around -6dB for tracking and the main stereo out under -0.8dB.

I also agree with Kurt. I generally track around -14 to -6. The extra headroom on each track is making my mixes sound better, due to the fact most likely cause I'm not running my gear as hot to try to aciheve a signal closer to full scale and the fact that myplug-inshave more headroom.

Originally posted by sheet:
Here is the concept:

By sampling at -18, I have 24dB of headroom at the converter. You can go lower to -12 and still get a good sample. Apogee guys say -16. It's up to you.

.

I am confused, if you calibrate at -18, why do you have 24 db of headroom. How do the affect one another?

-Mark

ok, when you guys say keep the level at say -14dB, are those figures for RMS value or peaks?

You're calibrating so 0 VU on your preamp...or .775Vrms...registers at -14dBFS (or wherever you want it to be...there's no"right" or "wrong" way to do it, although with a 24-bit converter I'd lean towards the conservative side unless you're actually recording a symphony orchestra or something else with a huge dynamic range) on your converters. It'll be kind of hard to do with Grace 101 since it doesn't have any sort of indicator that tells you exactly what your level is, but if you have access to a voltmeter you can do it that way.

-Duardo

Send a tone through the system adjusted to 0dBV
(.775 Vrms)

I hope this is all correct - someone who's an actual EE should jump in and correct anything I've made mistakes on.

Small nitpick(sp?):

0 dBV is 1.0 Vrms
0 dBu is 0.775 Vrms

FWIW...

:) Ok, I have a question, where are you placing the 2-bus fader when going for these levels? Mine seems to favor a very small travel around 0, to a tick or 2 under, anything less is quite noticable.

IOW, are you metering with the buses full up, or near full on your DAW? I still shoot for the recommended range, but my master bus fader(s) must be near full to get those readings, unlike analog.

--Rick

I always leave the master fader in PT at 0. If I get too hot it's time to re-think the mix. If the mix is good except for a few stubborn peaks I'll usually get lazy, and squash 'em with some really fast limiting. Dropping the master seems to cloud the mix to my ears.

But then I'm a hack! :s:

:p:

Originally posted by M Brane:
I always leave the master fader in PT at 0. If I get too hot it's time to re-think the mix. If the mix is good except for a few stubborn peaks I'll usually get lazy, and squash 'em with some really fast limiting. Dropping the master seems to cloud the mix to my ears.

But then I'm a hack! :) No, I really needed to hear that, I am not used to the method of summing in digital with respect to an analog board. Running at 0 always seemed strange because there is no real travel. On mine, 0 sounds the best, and I too get a cloudy mix if I reduce too much. At least I am not alone in what I'm hearing, Thanks,

--Rick

Alright, I'm going to chime in and really show my recording ignorance.

sheet: You say to send the 1khz tone from an analog device at "unity" gain. I've read the definition: "No increase or decrease in signal strength at the output of an amplifier or device compared to the signal strength at the input." But it really doesn't hit home with me.

Let me offer real world examples and maybe someone can fill in the blanks in a way that I can understand. I have a Toft Atc-2 channel strip and a Crane Song HEDD 192 (got it just a couple of days ago). I run Pro Tools LE 5.3.1.

I should send a 1khz tone that is registering 0 on the Toft VU to the HEDD and observe the input level there? On the HEDD meter, I should be shooting for around -14?

This may be relevant: In the manual for the HEDD (which is very spare), under "specs, analog input," it says: "Digital zero is +16dBm. This allows 12db of headroom above a +4dBm 0 VU." Now there's some information, for engineers maybe! It also says gain adjust trims are on the back pannel. They must be deep inside a couple of 1/8 holes, because those are the only things on the back panel that I can't identify with an obvious purpose.

I have a CD I just bought called Recoding Studio Test CD. I listened to it once and figured I just got ripped off ($15), but it does have a 1khz tone. So should I mike the computer speakers playing 1khz with the Toft and adjust the level to 0 VU and then watch the HEDD meters? Will that work?

I hope someone is willing to really get down to the basics for a novice with some rather expensive equipment (it's just not right, is it?).

Originally posted by Steve Butler:
...I have a CD I just bought called Recoding Studio Test CD. I listened to it once and figured I just got ripped off ($15), but it does have a 1khz tone. So should I mike the computer speakers playing 1khz with the Toft and adjust the level to 0 VU and then watch the HEDD meters? Will that work?

:) Actually, and you can laugh if you want to, you don't need a tone. Just a mic and a steady sound source to get an even reading on the meters of your pre. You can use a vacuum cleaner. Or some other steady sound source. You raise the input level so the meters show 0, +4db, from there you set the input level of your DAW to read ( -16 DBFS ) on your digital input meter. With the steady noise, meter at 0 on the pre, and (-16DBFS )on the input should line you up. BTW, if your meter is post sub fader (control), that post sub/master fader (control) should be full up at 0 DBFS, then the input level control is set to show (-16 DBFS) on that sub/master meter. This depends on your software as to where you can view levels, best bet is at the input.

I hope I did not confuse you,
--Rick

P.S A test tone from a speaker into the room may not present a steady enough reading do to acoustic reasons. If you prefer to use your speakers, use pink or white noise. Your test CD should have included this, if not, you can download a small test set from
http://www.hitsquad.com
go to music and recording software, audio generators. Make your own test CD with the noise, then play IT through your system.

Thanks for the reply. Well, I understand the vacumm cleaner and sending the signal into Pro Tools which is my DAW. But sheet and others were talking about calibrating the analog to digital converter, rather than the DAW. In your explanation you make no mention of the converter but suggest shooting for -16 in Pro Tools (I guess you mean on a record enabled track?)

I don't mean to sound like I'm not thankful for the help, but what happened to calibrating the converter?

Sorry, I realize PT is not the DAW but the software running inside it, or maybe it's not, I don't actually know now that I think about it. The reason I used the realworld examples with the name of products that I have is so that someone could fill in the blanks in a manner like Toft 0 VU, HEDD -14db, PT -14db. To help avoid confusion.
Thanks again.

Originally posted by Steve Butler:

I don't mean to sound like I'm not thankful for the help, but what happened to calibrating the converter?

:) Sorry Steve, what I mentioned was to establish unity on the primary source, your Toft, which apparently does not have a line in allowing the use of a tone.

ANY device downstream, including your metered converters must have an established reference to make the fine adjustments. The converters being one more device in the signal path. You must start somewhere with a known level. Your converters, and it's trims, seem to offer adjustment to allow a variable headroom range. Chances are, the factory settings ARE set for unity.

But suppose you wanted to drive that pre harder for a particular sound. If the pre does not provide a metered and controlled output, the only adjustment left would be the trims in the converter. This part in the chain cannot exceed it's maximum. So depending on the dynamics of the source, you have to determine how much headroom you will need on the converter.

--Rick

Originally posted by M Brane:
I always leave the master fader in PT at 0. If I get too hot it's time to re-think the mix. If the mix is good except for a few stubborn peaks I'll usually get lazy, and squash 'em with some really fast limiting. Dropping the master seems to cloud the mix to my ears.

:p:

I can't argue what you think your hearing...but so you know on Protools, lowering the master fader is mathmatically the same as lowering the channel faders.