Skip to main content

Hi,
I am interested in the allocated headroom that you guy's leave for each instrument.
Although this is really a question in regard to Digital work stations the RMS equivalent would also be of interest.

Question #1: The maximum peak level allocated to a drum kit e.g. -18 (digital) -3vu ?

Question #2: The overall Maximum peak level allocated to the band without vocals?
how much do you leave for Vocals?
e.g.( Beatles -3 vu (rms.) The band peaks at -3 before vocals.)

Question #3: The peak target before mastering?
Does everyone aI'm for the maximum dynamic range so that the peaks are at -.03 digital or should you leave the mastering engineer some dynamic leeway to apply any necessary treatment?

Regards Michael

Comments

Mixerman Fri, 08/17/2001 - 22:30

Mixhael,

Please don't take offense to this, but that shit is irrelevant. Use your ears, and make the mix work for the song.

Mixing isn't about set values. It's about the combining of instruments so they have the maximum emotional impact. Throw away all those preconceptions. We are mixing music. Try to operate on a musical level. Your mixes will come out way better than if you scientifically figure out level values within a mix.

Mixerman

Ang1970 Fri, 08/17/2001 - 23:00

I think he's looking for reference level more than "how loud should I mix my drums". In the digital domain, you can run into that digital ceiling if you aren't careful where you start off. And when you are forced to change the level of the entire mix to compensate, you run the risk of affecting the overall tone that you were so careful to sculpt for maximum emotional impact.

So, to rephrase the question, where do you start your mix levels in order not to overload the mix bus by the time you have all the elements together?

anonymous Sun, 08/19/2001 - 01:47

Hi,
Thank you for the replies,But I find the whole response quite lame except for Ang1970 (Thanks),
I apologize,I made the assumption that what I had written would convey enough info to those working engineers (who are supposed to be experienced) who would recognize why I had asked these literal type questions.The answer "I make it up as I go along" is fucking lame!!!You are Engineers? Engineers have to be fluent in dealing with the limitations & absolute boundaries of their mediums.I did not ask for A fucking mix by numbers, fill in the talent when required, analytical,audio balance holy grail, Type,help me get my mix shit together? Question!!!!I can already mix any style.(except rap & metal).I suppose when I asked the questions I was on a rock gig & in reference to that there is only so much,make everything louder than everything else headroom to play with.I know for a fact that engineers I have spoken to who worked before the mid 60's did have a very precise set of self defined & self imposed dynamic guidelines to allocate to mixes (Vinyl) after the 70's the word Audio engineer should have been dropped for Sound Mixer!I bet you all do have Guide lines? Maybe you have not rationalized them? Bullshit, you must! You just don't want to give away anything? Maybe? O K, Lets put the question another way.I suppose what I want to know is how much, if any thing,do you rationalize about your headroom limitations?You don't? "I just get a fuzzy, feely thing happening & my Dynamic headroom just takes care of Itself"!.Fucking losers! Bullshit!!!At Some point Everyone either subconsciously or not, is dealing with these issues. What I want to know is how other people approach ( rationalize)utilization of Dynamic range.I come from a analog world (JH24, 820's & Atr90's etc.) & Digital is fairly new to me (2.5 years on Logic & ProTools TDM).The leeway of analog is very forgiving, Digital is not at all.
Maybe Mixerman (Welcome as well!) you don't know what I am referring too & you couldn't give a shit cause your not into digital? that's cool too, I don't have any problem with that!.But maybe some of you other digital heads do know what I am raving about?I don't want a Digital verses Analogue debate,It would be pointless, I love both.
Anyone who has made the jump to digital would know that you are instantly confronted with these new set of limitations (with very vague standards, all brands differ)& I know there are people who are still totally analog who out of a commercial decision will be forced to go down the dreaded path in the near future.Maybe a little discussion on the subject would benefit those guys as well? I have always been in work & I have to do my research on the job (suspect eh?).I already have been making it up as I went along,What I would like to do is have a rethink, set my own reference standards in regard to digital mix headroom allocation!Fuck I must be lazy? Cause I asked for some help?Does any body really know what time it is? Does any body care?
Regards & peace Michael

audiowkstation Sun, 08/19/2001 - 04:13

Ok...I will chime in...

Now should I be mean here ....(average or a bully)

Hell no..

Here is the deal.

Ears and monitors and translation. Period.

Does it sound musical...vibing, entertaining, you clients are stoked?

I have used my (SSL SL G+8000) for analog compression running it hard on cetein tracks...yes...and will contiue...while other tracks are at -12 ON PEAKS...for the sound I want...solo, PFL or otherwise.

Gain management dictates that you do whatever you have to do to make the mix ...first of all :

Masterable.
Sellable.
Enjoyable.
Workable.

In the DAW situation...mixing can incorporate all different vibes and perceptions to just get the right sound.

When I first mixed a project in the 70's...I went for smooth...the artist wanted punchy.

I gave the artist punchy and incorporated the smooth in mastering...because I do both...and he said that it was too puchy...after my smooth incorporation.

Trust your monitors and ears. No one wants engineering induced distortion. NO ONE. Not yet anyway...Well the group deprivation asked for it....It at least gave thenm what they wanted on the final.

Do this. Bump your bottom to a great punch= smoothness ration..In loudspeaker design it is reactiance VS resistance...equalize that...lay the vox on the top like a cherry on a cocktal...and have plenty of mix. too much raw will destroy the mastering and too much refinement can leave the mix blang and masked.

One more time...Learn you complete translation.

In the 25 + years I have been doing this..I check my translation daily. It changes ...because of our perception from day to day.

Hope this helps a tad...

audiowkstation Sun, 08/19/2001 - 04:19

Sorry for the typos...another long assed session...going on...during a break...i simply don't speel check..

I do it as it comes to mind.

Remember...their are no absoltes...engineers that dial in guidelines sound like total ass IMHO.

Just do what experience teaches you...once you get nominated for a grammy or two for engineering (I have)...well at least someone appreciates your effort.

anonymous Sun, 08/19/2001 - 21:32

Thank you all,
I think I've answered my own question!When tracking I observe the normal procedure. The problem comes later when you start getting the killer sounds.(Usually the clients ego) In a digital workstation it is so easy to normalize or raise the gains of the kick & snare to an overkill level for maximum impact of -3 or -.03 that in the summing stage you run into problems.( I am talking about mixing totally within ProTools) So taking that unity at 0 = -12 or -13 or -14 or -18 ? I will peak to that & forget about tweaking the gain & leave myself the headroom & treat the medium like a tape machine!

Regards Michael

warlock Sun, 08/19/2001 - 23:01

Well I personally try to keep te levels between -6 and -3 when tracking, so the signal is hot enough, but I still have some headroom if I want to make something very loud. And you can always pull the fader down a bit...
And I keep a limiter on main mix insert. It's just a precaution because the digital distortion is horrible.

Keijo

Rog Mon, 08/20/2001 - 02:54

Mixerman is right, it's about the material you're working on, the gear you have and your own biases. Some desks can be pushed harder than others. Also, analog tape can be pushed hard, DAWs cannot be pushed at all so I'll record VERY hot with a bass drum track to get that fat sound, and back off a little on snare so I don't squash the transients before it ever reaches the compressor. I'd never dream of doing this with my DAW though.

Similarly, watch for transients with DAWs because sometimes the VU meters don't show them too well - use peak hold if you have it and aim for -6dB or so. That would be my suggestion but there are others here that know a hell of a lot more than I do.

As for mixing and overloading the 2 track bus, DAWs are great for this as you have tons of headroom and you can group faders to bring the overall level down before it hits the comp/limiter on the 2 track but levels are not set in stone, all is subjective.

Sorry if this is too simplistic but I hope it helps.

Ang1970 Mon, 08/20/2001 - 15:14

MEM,

-16, -18, -20... it's totally arbitrary. Use whatever level works for you. I'll sometimes start off a session with 12dB of artificial gain on the mix bus, so when that peak hits, you know what ball park you're in. When the artificial gain comes off, it's highly unlikely to affect the tone.

Of course, this is only neccesary for totally in-DAW production. Once you go outside to analog console, none of this is neccesary.

anonymous Sun, 08/26/2001 - 09:11

"I apologize,I made the assumption that what I had written would convey enough info to those working engineers (who are supposed to be experienced) who would recognize why I had asked these literal type questions.The answer "I make it up as I go along" is fucking lame!!!"

To you maybe...but...

Good mixing is an ART - not SCIENCE. It is as much a part of the performance as the musical performances are themselves. Once you take it into the set scientific mode you are screwed.

Just my HO.

zip >>

Ang1970 Sun, 08/26/2001 - 20:14

zip, I think everyone missed the point that what Michael was really looking for is a digital reference to unity. Selecting the right reference level will give him a starting point so he won't run out of digital headroom by the end of the mix. If you read it again, you'll see that he's not asking us "how do I mix a record?"

Yes, his response was a bit belligerent. He should have realized that these people who misread his post were actually trying to help him, and probably didn't deserve the abuse he doled out. But we must transgress this if we don't want to get stuck in the mud. I'd rather move on myself.

On second thought, maybe y'all could use a good spankin'. "Thank you Mistress! May I have another?" LOL

RNorman Mon, 08/27/2001 - 06:26

Originally posted by Michael Earth Media:
Hi,
The answer "I make it up as I go along" is fucking lame!!!You are Engineers? Engineers have to be fluent in dealing with the limitations & absolute boundaries of their mediums.I did not ask for A fucking mix by numbers, fill in the talent when required, analytical,audio balance holy grail,

Regards & peace Michael

Quit your griping and listen. Not only to what Mix says, but to your music. Make decisions based on the knowledge you've accumlated from experience. Don't clip is about the only rule. If you're running analog, find the sweet sounding point. If you're running digital at 16 bits, well, you have to be real careful. If you're running ditial at 24 bits, be a little less concerned, give yourself some space, and figure the shit out. Who knows what you're recording? You. Who has the intimate knowledge of the way you're working? You. No matter how many questions I ask of you, it's still going to be you. You have to figure it out, and that's what you've gotten as an answer. So whatsamatta? You got shit for ears or what? Want simple answers to complex questions? Don't want to experiment or learn? Shit, I'm ten years into my own studio, full fucking time and I have to learn things the hard fucking way. Mixerman might give me some pointers, but believe it or not, they might not always work until I get it through my thick fucking head that what he's saying is right.

Peg everything, bring it down in mixing. Bingo, you don't get much noise. Don't peg everything and bring it up in the mix. You get more noise, but maybe that's the fucking way to do it. Do you have U87s, 67s, 47s? Do you have Shure SM57s and 58s? Who the fuck knows? Again, it's YOU. Don't try to put the blame on people that answer your posts as if they didn't give pertinent information. They did. It's up to you to be able to recognize it when you see it.

RNorman Mon, 08/27/2001 - 06:39

OK, so I'll apologize for my last post, only in so much as I thought about the response again and almost went in and edited it and cut out all that I wrote. Then I thought about it some more, and if you have the experience you state, what's so hard about transferring it to digital? With 2.5 years in Logic, et. al., you should have plenty of your own ideas of how it works. Why ask now?

anonymous Mon, 08/27/2001 - 07:01

Michael,

For DAWs the general rules I use are to maximize the inputs, mix with the most prominent channel peaking as close to 0dBFS as possible with NO clipping. The replies about using your ears first are just a true with DAWs as with analog. As far as judging your headroom, sure, use the meters to be sure nothing is peaking, but as far as what level you should shoot for with what instrument really depends on the song/style, EQ and compression choices.

I don't mix in PT and bought another DAW because of the some of the mixing concerns there. Many people have told me about hearing audible artifacts when setting channels below unity in PT (hence the recommendations to track as hot as possible, and leave the faders at unity, mixing with plugins (kind of a bizarre workaround, IMO), or running back out to a console - making PT one dang expensive editor/harddisk playback system). For reference, Logic actually sets its audio mixer levels at unity as a default (listed as 90 in decimal scale view - 90%) - other DAWs list this as "0", but it is, I believe, around -10dB, but I haven't tested it to be sure. Since unity is merely a pass-through of the audio file as is, you have no digital computation of gain or cut, no bit depth changes, and hence no sound quality penalty (or shouldn't at least).

As far as normalizing tracks, my preference is not to unless they were tracked significantly lower than I need them, and even then I would probably only normalize the kick or snare to -1dB or less depending on whether I will add any EQ or compression (this leaves 1dB of headroom on those channels when set for unity gain, with no EQ). When you add EQ, compression, etc. you could easily boost peaking frequency bands of normalized signals into clipping, and need to drop fader levels back to compensate. Now you have multiple gain stage artifacts - gain increase from normalization, gain from EQ/comp, and cut from mixer faders - this can add up to more computation than necessary on an otherwise pristine signal. With PT and some other DAWs this can cause problems. The fader gain issues should not exist and are a design problem, not a problem with digital in general, but since they exist, you have to take them into account. Since my DAW creates no audible artifacts with fader and input trim gain and cut, I don't have to worry about leaving faders at unity, but I still follow my general guidelines as above.

RMS levels really take a little bit of practice with a DAW, but the method is identical to any other format - when you set your mix level audibly, you take into account what might be done at mastering (limiting, compression, etc) and work for whatever overall "loudness" effect the material should have. My rough estimate from looking at mix signal stats in my editors is that usually this RMS level in a DAW may range from -20dB (dynamic material) up to -3dB on highly compressed pop/rock (0dBFS reference). Just a rough reference, but by no means a "standard", since there isn't one (and shouldn't be unless we are creating cookie cutter music).

How you set the drums and vocals relative to each other is really best done by ear as has been said before - although you could work out the math of the dB difference, obviously metering wouldn't correlate, so you would have to measure the difference digitally (i.e. on a 0dBFS range, at your bit rate of choice, measure what levels vocals, drums etc are in relation to one another - if transferring analog tracks to a DAW, simply use the DAWs metering as your scale - as long as nothing clips). I hope this helps a little. Digital is only different from analog in that the ceiling is absolute - where you have the option of saturating channels/tape in analog, you don't in digital, but when the end result is a CD for either originating medium, the maximums become equal in the end.

Regards,
Dedric

anonymous Mon, 08/27/2001 - 11:04

Originally posted by Ang1970:
zip, I think everyone missed the point that what Michael was really looking for is a digital reference to unity. Selecting the right reference level will give him a starting point so he won't run out of digital headroom by the end of the mix. If you read it again, you'll see that he's not asking us "how do I mix a record?"

Agreed. I guess a more constructive answer would be to track hot then before you mix get a mental image of what you want the end result to be. Start low...achieving the mental image then push them up until you are close to unity.

It all depends on how many tracks you are mixing and what prominence you want each track to have in the mix...it's all additive so it's impossible to give numeric references for starting points.

zip >>

Ang1970 Mon, 08/27/2001 - 18:12

zip, much better. Thank you! :)

miketholen, are all digital meters determined by the same log? If so, 3/4 would be the same on every meter. But if not, 3/4 on one digital meter will be something very different on another. And that still doesn't answer the question, because the exact pont of 0VU is user definable in the digital domain. So do you calibrate to -15? -20? -18? And don't say just because PT ships 888's calibrated to -18 we all should accept that for gospel.

anonymous Wed, 09/05/2001 - 16:43

i think levels in digital are much simpler than analog.

In digital, your level should not reach 0 dBFS. Note that i don't say 'exceed' because it's not possible to exceed 0dBFS.

If you reach 0 dBFS for one sample then go back down again your a fucking genius because you've milked the full potential out of digital. If you stay at 0dBFS for 2 samples or more, than your on your way to creating a square wave, which means that for the duration of that clip, all sonic detail from other (most probably) high frequencies has been lost.

This isn't always a bad thing. Some carefully applied digital clipping can enhance some styles such as rap. I think that when mixing to stereo, you shouldn't be too anal about clipping your converters. Before scrapping the pass, ask yourself, did that clip actually sound bad? Could i even hear it?

The benefit is that you'll get a much better signal to noise ratio on your digital pre-master.

x

User login