does trk normalizing lose headroom in mix

[Lexicondonn]

Hi Dave, a nother fast question...

If you record with your digital levels @ -12
and then process the individual tracks with say the UAD-1 1176 (comp) to whrer the track levels will now read -2 db... in your opinion do you think that this takes away over all headroom in the mix or even on the individual tracks as well?


Thank you

[travista00]

if you're mixing all in the box, you're gonna need a lot of headroom on the individual channels so as not to overload the master fader (digital summing.) though, you can always lower the master fader to regain that headroom. again, this is specific to "in the box" digital summing. when breaking out to individual channels on an analog console, the answer changes as gain staging is pretty different and more about hitting "analog sweet spots." but, in the end, it all gets crushed by eddie shredder, i mean schreyer, anyways!!

[Lexicondonn]

Ok thanks Travis,

I'm also talking about.....if I record into Nuendo a bass guitar track and I set the input meter to read say -2db using nice compresson and a good player of course,
it seems to feel warmer when I record the same bass track but at say -12 -14db.

Is it just me because I have been a tape/analogue guy until a while ago.

However, when I do record hot like -2db
it does have balls and I do like the it. I just need a few other opinions.

My concern if we are recording all tracks in digital which are all really files in the end, as hot as 0db do these files end up picking up extra bits of junk noise information and become brash sounding ?

Perhapes a dumb question , just a thought.

thanks

[Boby Stoica]

Wich converters? It can be the analog part of your converters that behaves in this specific way. Or just the output of your pre. Or the gain on the pre.

[Jules]

IMHO - The trick here I feel is to be less preoccupied with the 'level before clipping' while recording and be more aware of the optimum level your equipment PRE CONVERTER likes best to operate at.

Make your converter conform to your outboard (via calibration / adjustment) and not the other way round.

Note: There is a theory that back when digital was new it was a good idea to feed converters 'near digital zero' for a better conversion quality - (so mixers diligently and painstakingly set levels to DAT players to be a hair away from "overs") and that nowadays, digital 20 years forward, with better converters and SN ratios - this precautionary procedure is now redundant.

Personally I have adjusted my methods to fit in with this theory, and dont worry too much about extra hot levels to my converters.

[Bloodz]

Hate to clutter the forum with this question, hopfully sombody else out there needs the reference...What exactly is digital summing or summing in general for that matter?

cheers,

bloodz

[TonyBelmont]

Originally posted by Bloodz
Hate to clutter the forum with this question, hopfully sombody else out there needs the reference...What exactly is digital summing or summing in general for that matter?

cheers,

bloodz

Digital summing in this context is basically when you mix completely in the computer without the use of an analog console.

[LTA]

Originally posted by Jules
IMHO - The trick here I feel is to be less preoccupied with the 'level before clipping' while recording and be more aware of the optimum level your equipment PRE CONVERTER likes best to operate at.

I concur, especially when running 24 bits.

Digital is funny, in that headroom is based on a fictional nominal level. Because of that, you can swap headroom for for a higher s/n ration. But the noise floor of some analog gear is signifigantly higher than the "noise floor" on the converter. Normalizing just takes the blank bits at the the top and inserts them at the bottom, raising the signal and noise floor similarly. You normalize, you end up dropping the fader back down to compensate or you end up with mixes with no headroom to fit in a kick or even snare sometimes.

Depending on what i'm doing, i'll use either the vu meters on my pre's, or the sound coming over the monitors (ala mp-2nv) while adjusting the master output to set levels with the RMS values my RME box spits out. After switching over to using rms (vs peak), i don't have issues with low levels, and 99% of the time it won't peak. Even when the drummer gets a little crazy during a take. No need to normalize it, as the faders usually will get dropped below 0 vu anyway. A distorted chunk guitar that has been normalized might even need to be cut 30 dB (or more) below the 0 fader position.

I'd almost suggest you try to limit the concept of "peak metering" from your workflow. About the only thing it is good for is when the little red light comes on saying you had 3 full scale samples in a row, suggesting you back off a bit. It will get you past some of the complaints of working with digital. Mainly, the misconception that 0 dBFS is your friend.

[adam_w]

It's a funny one - like the last guy said (in more eloquent words) 24 bits is all about more dynamic range that you can potentially capture, and if you're doing pop or rock..you don't use much and can afford to not necessarily hit just below zero
all the time..wheras in 16 bits, I reckon yes, recording hotter sounded better than not.

Normalising is probably a pointless excercise and not improving your audio quality, whereas using a compressor and getting an overall level increase is different..I like to fantasize and think that when I use something like a plugin and crank an input and/or output and get level gain I do so cos it sounds better.

summing; well..I reckon if you are applying old school gain structure through your input chain, convertors & plugins you'll be ok. Yes, plugins need proper gain structuring too ! As far as coming out, don't clip your master fader. Build your mix properly, and away you go.

There's lots of debate and some real scientific sh1t
regarding summing math, but at the end of the day..I'm not convinced it makes tremendous difference.

[DigitalAnalog]

Normalize doesn't improve digital audio quality, it simply raises the overall level of the file. [I find normalize is best used in digital mastering when you want to tweak the last bit of dynamic range out of your masters before you burn your CD.]


When tracking, it's best to record high digital levels to get the best resolution encoded to your digital media. At lower levels, fewer bits are being used to describe the audio wave form generally degrading the signal and introducing granulation noise. Normalizing a low level digital recording just brings up both noise floor and signal level, not completely unlike the effect of introducing amplifier noise when bringing up a low level analog recording.

Inorder to get the record levels high, proper gainstaging thru quality gear is the way to go.

When playing back, summing amps certainly may come into play but at least you know you've got the best possible recorded signal under the circumstances and don't have to introduce any extra noise by having to crank up your "tape" returns.

Greg

[Chrisac]

I tend to record at mid levels (I would also never normalise) . In all my tests recordings I have found that recording less hot gives plugins a lot more headroom to work particularly on the stereo master buss which results on in amore dynamic, less static sounding mix

Whe i first started recording digitally I was recording everything as hot as possible and all plugins sounded crap to my ears. Now I get better results. I dont know why, Im not scientist.

I always use an L2 on the mix buss now. During my hot years I could never get a decent result from that plugins, now I do. The same goes for Compressors, well everthing really.

I realise there is techinical details regarding bit information and many say recording hot as possibble is the way to go, but if I do that I have to pull all the faders down de to verloadung of the master buss which I dont like doing, or the master fader which I like doing even less.

[Lexicondonn]

Thank you for your input . It has been interesting and helpful. Deifferent opinions are important.


Hey happy thanksgiving if you're Canadian and if not....come on up for a slice of turkey eh!


Peace

[stealthbalance]

for me whenever i mix all in the box in PT , i make sure that all the levels are low, especially on the stereo buss - or 4 me , the aux faders i have pre stereo buss.
its the only way for me to get pt to sound big and fat.

whenever i get a project that has been recorded to the wall on everything , it just makes me nuts. i also must put a trim plug across every channel so my faders are sitting somewhere where i can ride them easily.

ive actually become acustomed now to using the trim plugins on the whole mix , as i love having my basic levels of the mix to have all the faders pretty much in a straight line. it helps me to experiment much more.
mho

s

[Bob Olhsson]

Originally posted by DigitalAnalog
...Normalize doesn't improve digital audio quality, it simply raises the overall level of the file. ...

In fact "normalizing" degrades the audio quality as does every unnecessary additional bit of signal processing. The fewer times the numbers get crunched, the better the end result in my experience. It's very much like patching around unneeded stages in a console.

[norman_nomad]

Originally posted by Chrisac
I realise there is techinical details regarding bit information and many say recording hot as possibble is the way to go, but if I do that I have to pull all the faders down de to verloadung of the master buss which I dont like doing, or the master fader which I like doing even less.

Actually the technical data suggests that you should record at conservative levels in a 24bit system.. just make sure the lowest signal recorded is above the noise floor of the gear and ambient noise of the recording environment.

Using up more bit depth does not yield a high fidelity recording.

I've also found my mixes sound better when giving plug-ins a bit of room to breath, so to speak. Took me awhile though!

[Lexicondonn]

Ok! so now you've given me another question.

I was unaware that processing caused audio degradation of any kind. So if I use my UAD-1, with the LA2A....it's going to degrade the signal? Hum!

How much ? And how many times is allowed before you notice it?

[kittonian]

Processing a signal with a high quality plug-in doesn't "degrade" the audio signal like normalizing does. You aren't actually changing the audio file itself (this is what non-destructive editing and mixing is all about). The audio files shown are simply a reference to the actual audio file (unless you choose to work in destructive editing mode). When using a plug-in the sound is effected because it is played through that plug-in which in turn has an effect on what you hear.

[norman_nomad]

Originally posted by Lexicondonn
Ok! so now you've given me another question.

I was unaware that processing caused audio degradation of any kind. So if I use my UAD-1, with the LA2A....it's going to degrade the signal? Hum!

How much ? And how many times is allowed before you notice it?

This link will help!

Bob Katz is neat