Recording Signals, - 18dbfs??

[mister-newb]

Not sure if this is the right forum, and I hope I don't get attacked as I did in another forum. I am really new to recording, as you can tell by my name. I don't understand the talk about recording signals. Can someone please explain it to me?

I've heard that it's best to track at - 18dbfs. Does this mean that in my recording software, which is Sonar, I bring down the fader to - 18?

[Methlab]

No it means that your levels should peak in that area. You keep the fader at zero. I try and peak at -12 db when i track in.

[Farview]

The fastest, easiest way to get a handle on this is to follow this link and start reading.

Tweak's Guide to the Home and Project Studio

It's pretty complete so it will help you get started with all of this.

[Jack Ruston]

No it does not mean your signals should peak in that area. Basically it's what we call a reference level and it refers to the calibration of your analogue to digital converters...Simply put...we have various ways of measuring signals in the analogue domain, the most common in music production being VU or volume units. Most preamps are designed to operate with the signal averaging at 0VU. Now the VU meter is an averaging meter which means that it will not react to very fast peaks. Typically therefore you might get peaks measuring some ten dbs higher for example. Now when we plug that preamp into an AD converter, clearly if we matched 0VU to digital full scale...0dbfs (which can not be exceeded under any circumstances) our peaks would clip. So we adjust the sensitivity of the AD convertor in order to leave 'headroom'. Typically this headroom will be 18dbs. So 0dbVU = -18dbfs. Thus if the preamps are operating how they should be technically speaking, the readings in digital will peak around -10 or something like that.

There are a couple of different headroom conventions, and various units calibrated slightly differently. You dont need to concern yourself too much with that. A good rule is this 'Get the signal peaking around 2/3 to 3/4 of the way up the digital meter in your software and all will be ok'

J

[Sigma]

Originally Posted by Farview

The fastest, easiest way to get a handle on this is to follow this link and start reading.

Tweak's Guide to the Home and Project Studio

It's pretty complete so it will help you get started with all of this.

hahhaha guys test has mistakes

the " hook" is a chorus..not only an intro to a song...and there are others too

[Farview]

Originally Posted by Sigma

hahhaha guys test has mistakes

the " hook" is a chorus..not only an intro to a song...and there are others too

What are you talking about?

[mister-newb]

Originally Posted by Farview

The fastest, easiest way to get a handle on this is to follow this link and start reading.

Tweak's Guide to the Home and Project Studio

It's pretty complete so it will help you get started with all of this.

Thanks. Although that guide seems very thorough, after reading through it a couple times, I still did not see anything specific addressing recording signals, and how hot it should be.

If you know where I may have missed it, I'd appreciate it, or if you someone could offer me an informative rundown on this, I'd be grateful.

Jack Ruston, your response was helpful. I'm guessing that I just need to keep my fader at 0, and when tracking, just make sure the red doesn't get too close to it, leaving room for compression, so I can add to the signal?

[cavern]

o.k. i can see where your at, download this free plug in/ rnd | inspector / install it and run it as an insert on every track you record, then adjust your external source so that the plug in peaks at -18 dbfs and your good to go, then everytime you add some e.q., compression ,ect. check the inspector plug in to see where your at..dont go to close to 0dbfs, leave some headroom for your final mix or mastering... personaly i prefer to record around -12 or 14 depends on the source. note..if your rig only has external monitoring.., you may have to record a short piece and then check the plug in on playback to see where you are,..this will get you recording at proper levels and you can work on understanding why your doing it as you go...hope this helps.. good luck..

[Kiwiburger]

Average & Peaks gets a bit confusing. Another way of looking at this is if you used a continuous sinewave signal generator. That way peak and average are the same.

Imagine you ran such a signal - say 1kHz sinewave - through your preamp and set the output so the VU meter read 0dbVU. Being an analog device, your preamp should have a lot of headroom above this level - 0dBVU is not a maximum ceiling like 0dbFS n digital land.

Now calibrate your A/D converter so that this signal at 0dbVU peaks your Peak meter at -18dbFS. (Or whatever is the design 0dBFU target for your converter - read the manual).

Now, when recording your wildly varying real world signals - you now know that when your VU meter reads 0dBVU, the signal is in the right area. Obviously real world recording will have peaks and transients, unlike a test tone. That's why you want headroom, both in the analog gear, but also leaving headroom in your DAW.

[narcoman]

oh lordy...


AES -18 is about average signals level - its not about peak. Use it this way - when you're playing forte (not ff or more) you should be at -18. That way, the biggest rock hit or timp crash in the world wont clip your AD. Forte in rock would be your chugging along verse etc etc. Then any big drums fills or (in other genres) sub hits will get you peaking up at -6 or so.

[peeder]

My advice to newbies is the tried and true logic of gain staging: make things as hot as possible without going "over".

In an Analog-to-digital Converter (ADC), "over", going over "full scale" (FS) i.e. 0dbFS, generally results in an instant transition to a square wave, which is quite an unpleasant variety of distortion in most cases. But in analog hardware, or digital plugins, "over" generally just means "too much distortion" which could be, in theory, at any level. Plugins might get too much distortion at -18dbFS, 0dbFS, or +18dbFS...you will have to test each plugin and see, and some plugins (e.g. apEQ) let you choose whether and how it will distort. Analog hardware the same.

You want to go as hot as you can without going over to minimize noise and maximize dynamic range.

You will have to turn up each gain stage in your chain, perhaps sending in a sine wave at a nominal level, or just using your signal source, evaluating when it's gone over or not. Work backwards from the monitors when gain staging. In a floating point DAW, you will never go "over" on the bus so you can just settle for unity gain internally unless you want to drive or lower something.

Going into an ADC peaking at -0.3dbFS is not "over" despite the jihad launched by many people screaming that it is. Your ADC may or may not sound good as it approaches full scale. It might even have a soft clipper that you decide you like pushing it above full scale for. You might even like going square on the transients (generates tons of harmonics for e.g. drums). And you might have awful ears and awful monitors and you might generally be a dope in which case you should let someone else do the engineering.

But you can be safe in knowing that most of these rules of thumb are pretty much wrong and you can determine what works best for you in each specific instance...and the optimal use of your equipment will require you to abandon these rules of thumb and do so.

[Kiwiburger]

Originally Posted by narcoman

oh lordy...


AES -18 is about average signals level - its not about peak. Use it this way - when you're playing forte (not ff or more) you should be at -18. That way, the biggest rock hit or timp crash in the world wont clip your AD. Forte in rock would be your chugging along verse etc etc. Then any big drums fills or (in other genres) sub hits will get you peaking up at -6 or so.

This seems like a classic generation gap thing.

Older experience engineers have plenty of boxes with VU meters.
Younger engineers have DAWS, calibrated in dbFS Peak, and digital toys with Peak meters, and no VU meters in sight.

So when somebody says "you should be at -18" I have to figure out - are you old or young? Are you rabbiting on about Average dBVU or Peak dbFS?

I'm very convinced about leaving plenty of digital headroom. But if you really set your gain so that FF material peaked at -18dBFS, you would be recording well under the target 0dBVU range of your converters.

Converters make noise, so you still don't want to under-record and then have to boost the noise.

I think the word "dB" should be banned like **** and **** and ro******* ... far too contraversial.

[Methlab]

Lol..all i was simply trying to tell him was to peak at around -12 or so. People get so fired up on here about stuff that is not worth arguing about.

[Farview]

Originally Posted by peeder

My advice to newbies is the tried and true logic of gain staging: make things as hot as possible without going "over".

In an Analog-to-digital Converter (ADC), "over", going over "full scale" (FS) i.e. 0dbFS, generally results in an instant transition to a square wave, which is quite an unpleasant variety of distortion in most cases. But in analog hardware, or digital plugins, "over" generally just means "too much distortion" which could be, in theory, at any level. Plugins might get too much distortion at -18dbFS, 0dbFS, or +18dbFS...you will have to test each plugin and see, and some plugins (e.g. apEQ) let you choose whether and how it will distort. Analog hardware the same.

You want to go as hot as you can without going over to minimize noise and maximize dynamic range.

You will have to turn up each gain stage in your chain, perhaps sending in a sine wave at a nominal level, or just using your signal source, evaluating when it's gone over or not. Work backwards from the monitors when gain staging. In a floating point DAW, you will never go "over" on the bus so you can just settle for unity gain internally unless you want to drive or lower something.

Going into an ADC peaking at -0.3dbFS is not "over" despite the jihad launched by many people screaming that it is. Your ADC may or may not sound good as it approaches full scale. It might even have a soft clipper that you decide you like pushing it above full scale for. You might even like going square on the transients (generates tons of harmonics for e.g. drums). And you might have awful ears and awful monitors and you might generally be a dope in which case you should let someone else do the engineering.

But you can be safe in knowing that most of these rules of thumb are pretty much wrong and you can determine what works best for you in each specific instance...and the optimal use of your equipment will require you to abandon these rules of thumb and do so.

Correct, the ADC is fine up until clipping. Of course, the analog chain leading up to it will not fair so well when trying to boost signal will little transients 18db above its sweet spot.

Percussive things like drums are fine when peaking up around 0dbfs, synth leads and other softer types of signals will overload the analog path when set that hot.

You have to consider the entire chain, not just the converter. If you overload the preamp on the way in, you're screwed before you start.

[Jack Ruston]

Oh my GOD there's a lot of misunderstanding in this thread.

-18dbfs is NOT a reference for peaks!

Recording as hot as possible into digital systems stopped being a good idea when 24 bit was introduced.

Just trust me on this one: operate the preamp at the level it's supposed to be worked at, and all will be well. If there are times when you want to slam the preamp for creative reasons use some sort of attenuator after it to bring the signal back to normal levels.

If you're interested in finding out more about this I think there's some good stuff in Bob Katz's book as well as many others. There was an article in a recent sound on sound which set out all the common standards and how they relate to each other etc etc.

J

[peeder]

I'm sorry but I paid for my gawddamn preamps and I will decide what level of operation sounds good to me. As should the rest of you.

Again,

dfegadthumbsup

[Jack Ruston]

Well I don't want to be pedantic, but you've just said 'record as hot as possible' which indicates that you're not really concerned about the level the pre is operating at. If you were making a conscious decision to run your pres hot (which was a point I addressed in my post) then I could understand it. But you're not...you're just saying that a digital system should be used as hot as possible without clipping: Which is bad advice for a number of practical reasons.

I'm not trying to be confrontational here, it's just that the 'birth' of digital as it is today has really suffered from some old working methods that we've carried over from early digital and analogue. We have a responsibility to new engineers not to propagate these myths.

J

[peeder]

Originally Posted by Jack Ruston

Well I don't want to be pedantic, but you've just said 'record as hot as possible' which indicates that you're not really concerned about the level the pre is operating at. If you were making a conscious decision to run your pres hot (which was a point I addressed in my post) then I could understand it. But you're not...you're just saying that a digital system should be used as hot as possible without clipping: Which is bad advice for a number of practical reasons.

I'm not trying to be confrontational here, it's just that the 'birth' of digital as it is today has really suffered from some old working methods that we've carried over from early digital and analogue. We have a responsibility to new engineers not to propagate these myths.

And I don't want to give you basic tutelage in reading comprehension. And so I will just repeat the post which you have selectively read and taken quotes out of context in a vain effort to continue the silly jihad.

I have gone well past your larval stage of rules of thumb--I've done the math and I've done the bench work--and I suggest it's instead our responsibility to blast new engineers (and whatever you, for example, are best described as) out of that way of thinking.

Originally Posted by peeder

My advice to newbies is the tried and true logic of gain staging: make things as hot as possible without going "over".

In an Analog-to-digital Converter (ADC), "over", going over "full scale" (FS) i.e. 0dbFS, generally results in an instant transition to a square wave, which is quite an unpleasant variety of distortion in most cases. But in analog hardware, or digital plugins, "over" generally just means "too much distortion" which could be, in theory, at any level. Plugins might get too much distortion at -18dbFS, 0dbFS, or +18dbFS...you will have to test each plugin and see, and some plugins (e.g. apEQ) let you choose whether and how it will distort. Analog hardware the same.

You want to go as hot as you can without going over to minimize noise and maximize dynamic range.

You will have to turn up each gain stage in your chain, perhaps sending in a sine wave at a nominal level, or just using your signal source, evaluating when it's gone over or not. Work backwards from the monitors when gain staging. In a floating point DAW, you will never go "over" on the bus so you can just settle for unity gain internally unless you want to drive or lower something.

Going into an ADC peaking at -0.3dbFS is not "over" despite the jihad launched by many people screaming that it is. Your ADC may or may not sound good as it approaches full scale. It might even have a soft clipper that you decide you like pushing it above full scale for. You might even like going square on the transients (generates tons of harmonics for e.g. drums). And you might have awful ears and awful monitors and you might generally be a dope in which case you should let someone else do the engineering.

But you can be safe in knowing that most of these rules of thumb are pretty much wrong and you can determine what works best for you in each specific instance...and the optimal use of your equipment will require you to abandon these rules of thumb and do so.

And let me just add that this is why digital sucks. Because very few people know what the hell is going on with digital, there are a million myths yes, but the "cure" for these myths is often as bad as the disease. You get all these "experts" proclaiming The One True Way to digital, and the results are....drumroll please..."Sterile and lifeless". "No mojo." "Nothing beats analog, dude!" **** the experts and their fad diets. Truth in digital lies in custom-fitting everything as you go, and knowing everything you can get away with. Which means you will have to go into therapy for a while deprogramming yourself from this cult that has frightened the life out of your music.

Follow this "keep levels low" televangelist religion and you will win the battle against excessive distortion...perhaps...but you are certain to lose the war of creating a truly compelling artifact.

[Jack Ruston]

You're talking about analogue distortion...And it's a point on which I totally agree with you. But that doesn't change the fact that a mix full of hot digital signals leads to problems. You just end up having to pull everything down...and lets not even get into arguements about moving faders and all that stuff. Basically there's no benefit to recording hot and a lot of niggling problems, not least of all the risk of losing a take due to unwanted overloads.

J

[peeder]

Originally Posted by Jack Ruston

You're talking about analogue distortion...And it's a point on which I totally agree with you. But that doesn't change the fact that a mix full of hot digital signals leads to problems. You just end up having to pull everything down...and lets not even get into arguements about moving faders and all that stuff. Basically there's no benefit to recording hot and a lot of niggling problems, not least of all the risk of losing a take due to unwanted overloads.

Actually you might be stunned to learn that you can't overdrive the bus of a floating point DAW (no really), and you can just trim off any excessive level going into your bus via a master fader or trim/gain plugin...no problem, no loss of quality, effortless automation, no more faders to move than that. Or just neglect to apply makeup gain in your bus compressor if it likes the input level. Doing so lets you get as much distortion out of your plugins as you want...yes and you might also be stunned to discover that virtually all plugins that people think sound good offer increasing (and often pleasant) distortion going up to 0dbFS (and often past 0dbFS without going full square)...both on individual tracks and on the bus, without losing much in the way of dynamic range. If you print too low you will run into dynamic range problems that will tie your hands. No, 24 bit did not magically solve that.

And the occasional unpleasant over? They can be fixed nowadays in seconds with declippers. Quite well actually as long as the damage isn't too severe.

When you evaluate the entire chain start to finish, and you pick and choose exactly where you want to get your distortion, where you want to avoid a loss of dynamic range, where you have headroom and where you don't, digital actually becomes very easy, fast, and absolutely pleasurable. And your analog gear becomes your best friend...use it strategically, and never be afraid to squeeze every inch out of it!

But these "secrets" to getting there are studiously kept from people. I suspect when people discover these things they think it's their competitive advantage or something and they don't sing.

[Silvertone]

Originally Posted by peeder

And I don't want to give you basic tutelage in reading comprehension. And so I will just repeat the post which you have selectively read and taken quotes out of context in a vain effort to continue the silly jihad.

I have gone well past your larval stage of rules of thumb--I've done the math and I've done the bench work--and I suggest it's instead our responsibility to blast new engineers (and whatever you, for example, are best described as) out of that way of thinking.




And let me just add that this is why digital sucks. Because very few people know what the hell is going on with digital, there are a million myths yes, but the "cure" for these myths is often as bad as the disease. You get all these "experts" proclaiming The One True Way to digital, and the results are....drumroll please..."Sterile and lifeless". "No mojo." "Nothing beats analog, dude!" **** the experts and their fad diets. Truth in digital lies in custom-fitting everything as you go, and knowing everything you can get away with. Which means you will have to go into therapy for a while deprogramming yourself from this cult that has frightened the life out of your music.

Follow this "keep levels low" televangelist religion and you will win the battle against excessive distortion...perhaps...but you are certain to lose the war of creating a truly compelling artifact.

I usually stay out of these things but this morning I'm loaded for bear...

If you record at -18 and your recordings are "sterile and lifeless" I suggest you use different recording techniques as level has NOTHING to do with this. More likely it's your mic's, pre-amps, instruments and room... oh yeah and bad engineering techniques.

Now getting onto the rest of this posted statement...

Are you high?

No really, are you smoking weed or crack or something?

Do you know anything about proper gain staging? Really dude there actually is such a thing. That's all that Jack was trying to teach the person that posted this question. Then you post with your own home made up logic and insult someone like Jack, give me a break.

Using your analogy: You know, cause your car has a gas peddle you should push it all the way down and go as fast as possible, don't pay any attention to the speed limit they were designed by old f*cks, it's a brave new world... yada... yada...

Pretzel logic at it's best.

mister- newb - listen to Jack. -16 to -18 is the average level each individual track should be. When they combine at the 2 bus your recordings will sound a lot better for it! Good luck and keep reading and trying and only listen to all suggestions (my own included) with a grain of salt.

You can't hurry great engineering skills they are learned over decades.

[peeder]

Originally Posted by Silvertone

I usually stay out of these things but this morning I'm loaded for bear...

If you record at -18 and your recordings are "sterile and lifeless" I suggest you use different recording techniques as level has NOTHING to do with this. More likely it's your mic's, pre-amps, instruments and room... oh yeah and bad engineering techniques.

Now getting onto the rest of this posted statement...

Are you high?

No really, are you smoking weed or crack or something?

Thank you! I get to use this pic! fuuck

Do you know anything about proper gain staging? Really dude there actually is such a thing. That's all that Jack was trying to teach the person that posted this question. Then you post with your own home made up logic and insult someone like Jack, give me a break.

Using your analogy: You know, cause your car has a gas peddle you should push it all the way down and go as fast as possible, don't pay any attention to the speed limit they were designed by old f*cks, it's a brave new world... yada... yada...

Pretzel logic at it's best.

mister- newb - listen to Jack. -16 to -18 is the average level each individual track should be. When they combine at the 2 bus your recordings will sound a lot better for it! Good luck and keep reading and trying and only listen to all suggestions (my own included) with a grain of salt.

You can't hurry great engineering skills they are learned over decades.

And you have joined Jack in not reading or being able to understand what I am saying. In fact you probably know only a fraction of what I have already posted here, and you're one of those idiots who set up shop as a "mastering engineer" like many audiofools do because you just couldn't cut it anywhere else. (Hey you came at me...)

My approach to gain staging is way better than some moronic "aim for -18 to -16db, always." You aim for a sound while maximizing dynamic range. That is how you gain stage, that is what I stated, and if you are aiming for a meter reading rather than a sound then I was actually being charitable to you above.

[Jules]

This is a subject that would be good to put a 'how to" explanation into the Tips & Techniques section for constant reference..

If one of you wants to write it (or copy and past your text) I will approve its addition..

It should be titled something like 'Optimum level calibration for DAW recording - a guide"



It needs to be real easy to understand for a newbie..

[Doublehelix]

As a general rule of thumb, I try to set the channel meters in my DAW (after I calibrate my A/D) to somewhere between -6 and -9 dB on the peaks. This leaves me some headroom so that when summing applying things like EQ I do not get any nasty digital overs which can never be repaired. You might lose a couple of bits or resolution, but with 24-bit systems, you have lots of room to play.

Best of luck with your search for answers.

[DeadPoet]

I think you all have the same idea about this but explain it in a different way.

My explanation:

*) (one of ) the weak(est) point(s) in a DAW is the way audio is brought in. Try to be conservative with your analog levels going in the A/D's. Every A/D has a different sweet spot and some will sound better at -18dBFS than others at -6 or -0.00003. If you don't know the sweat spot of your own A/D's, try to stay on the safe side. There is no real consensus about this, though the number -18, -12 and -6 pop up here and there.

These numbers are for average level: the audio bounces up and down around this number but stays away from the dreaded 0dBFS.


*) Once inside the box your DAW software has unlimited headroom (yes it has. I have done null tests testing the summing process of C4.1 and PTMP7.4 clipping its internall busses up to +100dB).

*) All plugins are not created equal. There are plugins that either a) distort b) clip or c) do not distort or clip if you push them hard.

I have done some simple tests with some plugins and for instance the Waves SSL, API and V-series belong to category a) while the URS CSP belongs to category c).
It could be handy to make a list of plugins to know which one belongs where so you can make mix/creative decisions based on plugins behavio(u)r.


Someone correct or flame me when I'm wrong

Herwig

[petys]

I'm not really understanding any of this. If your meters are not going into the red in either the preamp, ad converters or software, you shouldn't be getting any digital overs. Aren't the a/d and software meters pretty accurate?

When I record, I make sure my pre isn't clipping. Then I make sure the peaks in the a/d meters aren't clipping and I think I'm good to go. Am I missing something here?

As far as the levels once theyve been recorded, and they are on my hardrive, does it make a difference where my faders are in the software, unless the meters in the software are clipping?

And as far plugins clipping, are you saying that they won't show up on your software meters as clipping?

[JEGG]

Yes, a generation gap seems to be part of this (myself included in the old generation).

At 24 bit depth, there is simply no reason whatsoever to record "hot."

If you want to allow, or want to make some sort of mild or obvious distorted artifact somewhere in the analogue chain (or somewhere in a badly designed digital buss-and there aren't that many of them), or you like stressing over levels, than go ahead and record hot.

[Andreas G]

Originally Posted by mister-newb

I've heard that it's best to track at - 18dbfs. Does this mean that in my recording software, which is Sonar, I bring down the fader to - 18?

In the analog domain the pro reference level is + 4dB.
AD converters might have their reference level at + 20dB (for good reasons!)

So if you track with an average level around 0dB on an analog VU of your pre, compressor, console or whatever, it will be around - 16dB in your DAW. If you have no VU on your preamp and your ADDA reference level is at 20dB, -16dB average relates to your 0db VU – meaning you won't drive your analog chain to hot.

BTW this belongs also for DA conversion which gives designers of analog gear often big headaches, because people drive their DAs so hot that they actually overdrive the analog device. On the other hand designers have to take care of those guys working in the analog world with a reference level at 4dB. So my advice for DA: Don't drive it to hot! what comes in (-16DB) goes out the same way and will end at 0dB VU.

This is gainstaging btw. So first I'd check which reference level your ADDA is calibrated to.

Hope it helps abit :-)

Andreas

[gollumsluvslave]

I think you have a basic misunderstanding of gain staging in the digital realm, and have this way of acting like you know it all and that anyone that disagrees with you is an idiot.

Especially when some of the people that advocate lower levels in digital (Paul Frindle for example) really know their stuff.

But of course, I didn't just take Paul's word for it - I tried some of his advice out - it was really quite educational.

Terry Manning:-

My firm belief is that if users of Protools, and the other DAW systems, would do the following, then MANY of the "digital" or "in the box" audio problems would vanish:

•STOP RUNNING YOUR SIGNAL SO HOT! Do not use the built in peak meters as you would use a VU meter. If you will keep your input levels lower, your sound will improve. You are not really gaining anything by trying to squeeze out that last little bit.

•If you can, find a way to also meter every input with a good old fashioned analogue VU meter...whether it's with an analogue console, a tape machine, a dedicated "meter box," whatever. Let these meter indications rule, while of course cross referencing the peak DAW meters as well. And use MUSIC as your general guide for I/O levels. Setting up with a 1k tone from inside Protools for your reference level to an analogue console following will not give accurate results on all types of program material.

Paul Frindle:-

A wholse host of posts and observations here:-
PSW Recording Forums: Reason In Audio => DAW & Desks: Is ANYBODY actually still mixing on their desk?

The simplest practical advice right now (with the kit you are currently compelled to use - and if your paymasters will let you) is to think of sample value levels in the green section of the meter as always legitimate (i.e. repeatable at destination). Those in the yellow area (-6dBr and -3dBr) are most probably ok, but caution should be taken as they're definitely big enough and may just cause reconstruction errors if clipped or intentionally distorted in the digital domain or digitally recorded from an artificial source. Levels between -3dBr and 0dBr are dangerous and those that actually reach the red light are almost certainly broken signals and are very likely to degrade in various ways at the destination DAC - in both your's and the end user's!! And above all - don't assume that any meter anywhere within the system indicates a legitimate signal by dint of it not hitting a red light

In a nutshell digital meters don't show signal, they show sample value - Paul recommends that to be certain of avoiding illegal reconstruction values (even when there is no 'red light' on the digital meter) we should be wary of peaking above -6.

Sure, if you are like Peeder and know your plugins and gear inside out, or actually like the risk of clipping at the DAC then you can ignore this advice.

I've followed this (helped in no small part by implementing K-System, which reduced my RMS mix level dramatically) - both have really helped really improve the sound of my mixes.

[mister-newb]

Ok, although I don't quite understand the technical stuff, I think I at least have a better understanding of this. Thanks everyone for the comments. I'm glad my question, as newb as it was, generated some discussion.