*Recording Levels*

[mentalmusic]

"TRACK AS HOT AS POSSIBLE WITHOUT GOING IN THE RED"

-it's the Golden Rule of digital recording, right?

Not anymore, I just read this thread from the TDM forum and it changes everything (for PT's at least):
http://duc.digidesign.com/cgi-bin/ub...c;f=2;t=005346

[suyenfung]

ahh yes this is something i just figured out on my own about a month ago. i'm glad to see my conclusion was correct.

[Shameless]

Yes the extra headroom we have does warrent median levels. Think back when you were doing 16 bit sessions and went to 24, wow ! mega headroom.

I translate this into my bounce as well, especially since, in many cases, I am not trying to get the squashed AOR sound. I favor dynamics to volume.

That said, the L2 is putting a new spin on home studio mastering and it will be interesting to hear how the 96k stuff will play here.

[accession]

Out of curiosity, I ran the Digi Signal Generater plug-in across a track to see where -14dBFS registers...

...and was very surprised to see it actually sits a fraction below halfway up the meters. It's quite deceptive.

I presumed it would have been around 3/4 up (eg. somewhere in the yellow). The meter scale seriously magnifies the final 15dB or so.

Thinking I'd be making the most out of the digital resolution, I've always aimed highest (without peaking -0dBFS), and pulled back the track faders to accomodate the mix bus, leaving the master fader at 0dBFS.

To be honest, I'm not entirely happy with the final results, with a slight brittle, single dimensional result. I've come to realise Sends and Aux tracks don't have much headroom, and that link posted just prior also points out the Insert paths don't either.

From today, I'll try a new plan of attack...

...aiming to bring my sounds in at -14dBFS (ie. half way up the meters), and allow the rest of the system work well within its limitations.

I expect my Digi001 converters will be thanking me too.

BTW, isn't there a difference between the LE and MIX bus summing, and the LE version actually allows for greater headroom?

Phil

[accession]

Oh, and Digidesign...

I know it's been said before, but can we please have a -dBFS scale on each meter for Pro Tools 6.0 so we'll have a much better idea of where our levels are truly sitting?

Cheers,

Phil

[accession]

Upon further investigation...

-14dBFS from signal generator = halfway up what is now appearing to be a very skewed Pro Tools meter.

-14dBFS from signal generator and through the Waves PAZ Meter (plug-in) which is scaled in even 6dB steps (0dBFS to -72dBFS) and now I see a more realistic 3/4 of the way up representation.

Um, my point was? [img]images/icons/cool.gif[/img]

Phil

[Gabriel Roth]

question for everyone who's been reviewing their levels: when you say, e.g., bring your tracks in around -14dB, i.e. about halfway up the meter ...

do you mean peaking at -14, so mostly occupying only the very bottom of the meter, or do you mean averaging at -14, and so peaking somewhere up in the yellow (but still well shy of 0)?

[accession]

Originally posted by Gabriel Roth:
question for everyone who's been reviewing their levels: when you say, e.g., bring your tracks in around -14dB, i.e. about halfway up the meter ...

do you mean peaking at -14, so mostly occupying only the very bottom of the meter, or do you mean averaging at -14, and so peaking somewhere up in the yellow (but still well shy of 0)?

<font size="2" face="Verdana, Arial">Well, using this revised train of thought, I'm working off peaking @ -14dBFS.

I just have to get into my head the fact that halfway up the Pro Tools meter is exceedingly high on any other scale, dBFS or otherwise.

And bear in mind also, -14dBFS (or thereabouts) is the equivalent of 0dBVU (er, I guess on a +4dB scale... damn these convoluted standards).

I also have to go back and review the link that started this thread to work out where things start getting messy.

I know I generally have had to back off my faders to avoid clipping the master bus, and hav had troubles overloading sends and aux busses, plus my external MPX 500 effects (over S/PDIF), and this is before we even get to the what happens in the plug-in insert path.

The main thing I'm not confident with is how and why sonic damage could occur without a red clip indication?

For now, I've recalibrated my system (Mackie 1604-VLZ PRO), allowing me to keep its levels regular, while reducing the Digi001's input sensitivity down so that I'm not slamming the converters at full strength. That's the first step.

Anyway, it's going to take another session before I can really compare the results.

What's most deceiving in all of this is the lack of dB referencing on the Pro Tools meters. This is such a simple thing which is available on any other piece of studio gear

Phil

[Shawn Parr]

There is a bit of a math difference between LE and TDM, TDM uses double precision integer math.

LE uses 32 bit floating point math.

[CCash]

Warning: the following post is probably too damn long.

Accession,

Cudos to you for trying to have a firm engineering understanding... I noticed your posts are always trying to get to the bottom of things. Hopefully this will fill in any blanks for you and others. If I tell you anything you already know don't take it personally at all, I'm only trying to help.

The "zero reference VU"... when people say, "I calibrate to 0dbVU, they're basing it on what a VU meter readout would tell you... and this is an averaging kind of meter. It's slow, so you can never feel too confident about where your peaks really are.

The meters in PT are peak meters, so they read the instantaneous highest point of the dynamic range an input sees. This is good for knowing exactly how close you are to 0dBFS... and, as you know, that's max.

In analog we often got cool results driving stuff hard, but as everyone knows, with digital, there is a mamimum -- OdBFS. Anything a little above this doesn't sound pretty.

I see you're asking for some kind of read-out for your level meters in PT. Just hold down Control and click on the volume button. It will scroll through Volume, Plug-In Delay, and Peak. I use the Peak setting all the time. Click on it to reset it to -infinity.

I've measured a lot of this stuff, and a -14dBFS sine wave is actually equal to the Digi 001 putting out -12.5 or so dB (so the headroom specks are just slightly vague, if not off).

When you say that you're shooting for levels of -14dBFS, you should really mean average levels... a VU meter type -14dB. So using the PT peak meters as your guide for determining -14dB VU on an instrument with a lot of dynamic range is pretty much useless for determing VU.

But here's the nice part: When you use a sine wave, a VU average meter and a PT peak meter will show you the same thing. There is no dynamic range in the signal, so the peak is the same as the avg. (and one reason why a 1,000 Hz sine wave is great for calibrating analog gear with digital).

Calibrating to 0VU with a sine wave means that showing 0dB on a VU meter will also show the digital average in ProTools. What number in Pro Tools? You decide. So if you pick -14dBFS to equal zero VU, use a sine wave. Use the Peak readout and a sine wave. Make your gear with VU or analog meters equal -14dBFS in Pro tools. Use Setup>Hardware>Options and set your inputs for the incoming gear. As you did, look at what a -14dB sine wave from the signal generator looks like on the PT meters.

I've read a lot of those TDM discussions and the bottom line is that the proponents are trying to record at "mix levels". In a perfect world, we wouldn't have to touch a fader for basic mix levels. And this apparently sounds good in PT, because the mix bus is getting hot enough -- but not too hot -- signal. 24 bit has changed the old rules.

So, the bottom line?? Here's the practical question: how am I going to know if I'm really at -14dBFS on average?? Well, um, you don't. Unless you're recording sine waves. [img]images/icons/smile.gif[/img] My feature request is averaging metering in PT as well as peak metering. Progams like Spectrafoo can do this.

-14dBFS = 0dBVU on the input side is a good choice for the digi 001. It claims to have a headroom of 14dB. So if 0VU = -14dBFS, then 14VU = 0dBFS, the maximum. Most +4 balanced analog equipment can easily put out +14dBVU, so you're not risking crapping that gear out. I personally use -18dBFS.

What should you choose? Let's say you're running all of your monitoring from the 001 to an analog Mackie board. I don't know the analog headroom spec for them, but let's say it's 20dB. Zero on the Mackie's meters should have 20dB to go before clipping.

Now you send your PT output at -14dBFS. You have set the mic pres on the Mackie to show that this -14dBFS equals 0 on the Mackie meters. Now send a sie wave to the Mackie at 0dBFS, or Digi001 maximum (which is the same as 0VU because it's a sine wave). The Mackie should show +14dB. Are you maxxing out the Mackie? No. It still has 6dB of "play" until it clips.

I think it's best to calibrate all gear separately, if it's reasonable. If you want to max out the Mackie, then make a -20dBFS sine wave from PT equal zero on the Mackie. Now when you peak in PT, you're peaking the Mackie too.

How do you know if an instrument is in the 0VU range? It depends on the dynamic range of the intrument. If you're recording a guitarist who has his rig compressed all to hell, you'll see very little movement from the PT meters. So turn on your peak read-out. You get -5.1, -3.0, -2.1. Better not turn him up any more... you're close to maxxing out. But you also know that the meters are moving only a little bit, so your average is probably also pretty close to the peaks. Might want to bring the whole thing down, before hitting Pro Tools. If you wanted the guitar at -14dBFS, you'd bring it down until you see a bunch of peaks just slightly above this. Again, think about final mix levels.

Last thing: You were talking about what -14 looks like on the PT meters, and you were surprised that it only showed about halfway, but on the PAZ meters it shows 3/4. It's all in the scaling, or meter resolution. Digi chose to make the "zero" at halfway, and full-on at max. There's A LOT going on below halfway then. If the total dynamic range of the 001 is 90dB (not sure offhand), then there is 90 - 14 = 76dB steps happening below the halway point of the meters.

When I finally feel like I understand an audio topic, I wanna pass it on... cause it's usually so simple... you just needed someone to explain in simple terms. Hope I did that a little. I have gotten progressively more drunk as I wrote this, but I hope it was useful.

BTW folks, I saw "Signs" tonight. There were a few things that bugged me, but overall, it's pretty cool. I have to give it a recommendation and advise you see it. I guarantee that you.... yes, even you, will jump a little at some point.

Good night!
Curt