What DB should I be peaking at before mixing?

Robbb

New member
Right now I'm peaking at 0DB before adding any effects. All I've changed is the volume of the tracks. It sounds way quieter than other examples that ive heard of peoples unmixed beats if I bring the volume down any more. What DB should i be peaking at before mixing? And, what am I doing wrong here?
 
"Before mixing" as in individual channels or just putting all the faders up and you're clipping the main buss?

Either way -- Most would probably suggest any individual track peak no higher than around -10dBFS. There are plenty who suggest your main buss not peak above -6dBFS, but there's really nothing wrong with being closer to full-scale. "Somewhere below -0.0dBFS" is fine. My personal rule of thumb is to keep as much headroom as I can within reason at every possible step in the entire process up until the final adjustments during the mastering phase.

The confusion here -- You aren't peaking anywhere before mixing. Individual tracks at unity will have a particular peak level, but you haven't set it yet until you start mixing. And if you're listening to unmixed beats, are you saying that you're listening to the individual tracks that would make up the mixed beat? Are you confusing the difference between "mixing" and "mastering" by any chance?
 
Sorry. Im pretty new to this. Let me try to break it down. I have a project in logic with guitar, drums, and bass in seperate tracks. I haven't put any EQ, Compression, or effects on anything yet. All I have done is change the volumes of the individual tracks to get the whole project to sound as good as possible. So I guess you could say that Ive only completed the first step of mixing. But right now, when I play the project as a whole, the master volume is showing that the volume level of the project is peaking at just under 0DB. I know that the volume should be lower than that, in order to leave space for compression, EQ, etc. But if I turn the volume of the individual tracks within the project down any more, it will be far quieter than unmixed projects that ive heard from other people who know what they're doing.
 
Volume perception is relative the meters give you the accurate info. I mix without ever looking at or adjusting the actual volume control so it’s normal to find that some songs will actually have lit up the red warning in Cubase, but others when I get them nearly done are way down on level, but sound fine. This is where the newer LUFS meters really help. Average volumes and the peak volumes all contributing. My own way of working is to use commercial releases of similar material on Spotify or iTunes and run those through a LUFS meter and use these as what a particular style should run at. With so much headroom nowadays, I don’t find any real problem simply adjusting the final mix up or down a bit. On a purely maximum level, I find that my mixes usually are a tiny bit, or a lot lower than a -3dB peak, rarely louder. Classical style music is always for me quieter in production, sometimes quite a bit lower, which I then adjust up.
 
I have a feeling these "unmixed" projects you're comparing to had a limiter or something thrown on the 2-buss. Apparent playback volume is far and away the least of your worries during the mixing phase. Will that volume compare to "commercial" stuff? Certainly not - nor should it.

Now if you want - and this just came up recently - go ahead and throw a limiter on the main buss to see how your mix handles the "abuse" of the freakishly excessive volume levels that everyone seems to think of as "normal" these days. Just know that it's not really normal. It's basically a pissing contest between artists to be louder than the last guy and it's doing insane amounts of damage to the recordings. And I'd argue that if the average listener knew how much damage there was, there would be an uprising from the listening public. But that's really for another thread...
 
Surely it does not matter what the individual tracks peak to withing reason? If the mix results in a lower final level than someone else's mix, no matter, you can increase the final mix level*.

What is often lost sight of is the volume level 'in the real world'. If your monitors are calibrated, -18dB could equate to around 80dB SPL at the listening position and that is quite loud. Peaking close to 0dB fs is going to be very loud.

*I shall not use the term 'mastering' for I am not in the least worthy!

Dave.
 
its been of interest to me too recently. the tracking, tracks volume doesnt need to be anything specific and doesnt need to be near zero and better to be far away from zero to insure no clipping. that makes sense.

but then when it all comes together, and the mix level is going on before cutting a Wave-Stereo version..Ive been messing with the LUFS too recently, a new thing to me. Im hearing -12 to -15LUFS for the streaming etc...and in my opinion its sounding better with more dynamics...my output VU meters seem to be -6db to 0db with some swings...when the LUFS is around -12 to -15 LUFS short & int... when I put on McCartneys RAM in my player the needles are also 0 to -6 db but it sounds a lot louder and really polished and just better all around in everything...but the dynamics is much more pleasing to my ears than the loud flat muddy blare with the clip leds of the loudness wars...these days if it goes there I turn off the tune, especially home grown blare-loudness wars...the commercial stuff is bad enough, the home grown weed loudness wars is just mud.

**correct actually the RAM was showing -17 LUFS...but it sounds louder and cleaner than my HR stuff..? weird
 
Last edited:
LUFS and dynamic range are two different things, though the latter is related to peaks somewhat. Most loudness measuring tools will tell you about dynamic range.

While LUFS is designed to keep things sounding "the same loudness" there is no doubt in my mind that other factors impact how you hear something, not the least of which is how we hear music, its dynamic range, the things reproducing the sound, and if speakers (vs earbuds/headphone), the space being heard in, as well as the content and where the measured "loudness" is coming from in the music.

I like to start with dynamic range and keep an eye on that, because if you lose that, and want/need it, getting it back will take some unwinding and can cause revisiting a lot of decisions you thought you'd made already. LUFS only matters when you go to the "mastering" phase, really, but it's good to make sure you've not already shot past it in the mixing step, because then you're going to have a different problem, and probably have to go back to visit dynamic range again. The peak is, I suppose, less important, but comparing where you're at with that along with LUFS will really tell you whether you've left much tweaking room after the mix is done. And, maybe you don't need it, but a little bit might be appreciated.

If I said -6dB anytime, it was probably something I read :). But, it's not a horrible place to be if you've got a lot of dynamic range and at least that many dB of LUFS to spare, IMO.
 
Sorry. Im pretty new to this. Let me try to break it down. I have a project in logic with guitar, drums, and bass in seperate tracks. I haven't put any EQ, Compression, or effects on anything yet. All I have done is change the volumes of the individual tracks to get the whole project to sound as good as possible. So I guess you could say that Ive only completed the first step of mixing. But right now, when I play the project as a whole, the master volume is showing that the volume level of the project is peaking at just under 0DB. I know that the volume should be lower than that, in order to leave space for compression, EQ, etc. But if I turn the volume of the individual tracks within the project down any more, it will be far quieter than unmixed projects that ive heard from other people who know what they're doing.

My guess is your meters are registering too hot because you HAVENT EQ'd. EQing should lower the meters not boost them. . I mix I as I go. I never save all the mixing for the end. If I record an Acoustic guitar it gets a HPF and rough EQ and Compression setting. Same for Vocals etc. My pre-master mixes are very low in volume highest mix peaks are -8db. Volume comes during mastering.
 
Last edited:
so if you track and do a "quick" HPF, EQ, comp...almost like a standard template in the DAW or hardware approach. you keep the highest peak -8db on the master-buss ..that sounds easy to do.
but if I get Robbb's question, thats what the question was.


so do most engineers leave the Main Bus FX empty during this stage? or do they add a "mix buss" limiter/no. 6 etc on the main buss...and have reverbs on the main buss, post tracking, but to start the mix to 2 track?

then Mix to 2 track then apply stuff on the Main Buss like Mix Comp, Reverbs etc...?
Like Limiter No. 6 has "MIX BUSS"....Klanghelm Comp/Limiters also has a MIX BUSS offering. Ive never used those but does that work better for the first Mix stage?
 
All great stuff, however no-one seems to have mentioned one very important (IME) point; frequencies below ~500 hz are going to push peak and lufs numbers up with the result that the whole mix and individual tracks will seem quiet to the human ear even when meters show red.

So try this: put a low shelf on one track. Set it with the breakpoint in the 200-500 hz and watch the meter as you slowly bring down the low end. You should see the peak and/or lufs drop significantly. Don't leave it like that, it's just a tool to find out how much dynamic range is being taken up by the low frequency content of that track(the difference between eq'd reading and non eq'd reading). You will find you can increase the volume of the track now without creating overs on the main mix.

This is one of the reasons that a commercial release with a lot of dynamic range can sound both louder and clearer. The low frequencies can both add "mud" or loss of clarity and reduce the total apparent volume by taking up much of the dynamic range with unneeded low frequency content.

So start just as you have been doing-get a relative mix. Then start getting rid of un needed low end and compensating accordingly with the relative volumes as you go. You will end up having more volume and clarity on the main mix.
 
All great stuff, however no-one seems to have mentioned one very important (IME) point; frequencies below ~500 hz are going to push peak and lufs numbers up with the result that the whole mix and individual tracks will seem quiet to the human ear even when meters show red.

So try this: put a low shelf on one track. Set it with the breakpoint in the 200-500 hz and watch the meter as you slowly bring down the low end. You should see the peak and/or lufs drop significantly. Don't leave it like that, it's just a tool to find out how much dynamic range is being taken up by the low frequency content of that track(the difference between eq'd reading and non eq'd reading). You will find you can increase the volume of the track now without creating overs on the main mix.

This is one of the reasons that a commercial release with a lot of dynamic range can sound both louder and clearer. The low frequencies can both add "mud" or loss of clarity and reduce the total apparent volume by taking up much of the dynamic range with unneeded low frequency content.

So start just as you have been doing-get a relative mix. Then start getting rid of un needed low end and compensating accordingly with the relative volumes as you go. You will end up having more volume and clarity on the main mix.

Exactly, My point was that if the OP is not doing any subtractive EQing in the low end and low mids the meters are going to be booming. Putting a limiter on the mix buss is not the solution IMHO. I can't speak for other engineers I only put a console simulator on the mix buss.
 
You are confusing 'volume' with 'headroom'. Volume is a subjective effect of human hearing, and is produced in audio equipment by selecting higher or lower voltage levels, typically through (mostly) passive resistive channel 'faders'. Just to confuse matters still further 'volume' is not 'gain': the latter usually refers to the first stage of your mixer or chosen recording interface, it uses 'active' electronics and is the amount of electronic 'boost' you have to dial in to bring your mic (or whatever) into the voltage range your interface needs to operate properly. 'Headroom' is a somewhat ambiguous reference either to the 'blank space' you have in hand above your maximum recorded level before your mixer / gear hits distortion OR (and IMO more usefully) the difference between your wanted recorded sound and the unwanted background noise that is always present in any electronic equipment, including audio kit. There shouldn't be a problem when you get around to applying compression, EQ, etc. With analogue kit you just turn down the level ('volume') of the mix or discrete track you're sending to your compressor (or whatever), and with digital equipment you just 'de-amplify' i.e. fade that source as much as is necessary so that it comes back from your fx processor without hitting 'clip'. Generally it is better to start with a 'hot' signal at the first stage of the recording process, i.e. what you're sending to he mixer. You can always turn down the volume (voltage) of a 'hot' source so as not to 'bust' the next stage of processing. But if you record with an unnecessarily 'cold' i.e. weak signal, you are going to get proportionately more electronic or natural background noise in your finished product, which you will never be able to get rid of. True, all decent recording software now includes remarkably good 'noise reduction' options - as long as the noise is an unvarying constant, unlike human chit-chat or passing traffic - but it is still better to have 'swamped' the noise in the first place. Finally, in most analogue mixers at least, audible 'clipping' will not occur if your red light winks occasionally; like car speedometers, the manufacturer is 'playing safe'. In professional equipment, the nominal 'full deflection' signal is +4dBu, that's about 1.2 volts, (.775 volts for consumer devices) but in practice can peak much higher before anything nasty happens; in fact in the days of reel to reel tape we deliberately ran as 'hot' as we could get to feed the tape: the result sounded 'juicier', although you'd probably NOT want that in recording Classical music
 
Back
Top