The REAL Question: How loud should you mix instruments and bandwidths?


I wondered why I hadn’t put any of my finished tracks out there for a decade. It occurred to me there is only one reason. I do not know what the accepted levels for frequency occupiers. I’ve searched… Please assist me, but the agreement seems unseen and I reject totally the idea that it is up to me. Yes, it is. I want to fit in (that was a joke). Here are my burning questions, but ultimately they come under the same one. If you have a sound and its peaking within a certain range, what db average should that range be? What does the industry expect? ‘Should’ is not a trap. I want guidelines.

I’m open to making changes afterwards, of course. But lets say you have your kick drum at x, how does everything else relatively compete? And what are the standard places to ‘place’ said instruments, e.g vocals in the centre, hats wide, etc. I’m tired of trawling forums with the answer being ‘it depends’. Sure, so after what it depends on, whats the bottom line? :wink:

Sub bass
Bass mid lead
And more!

and/or talk specifically in Hz.

Share links as this seems to be an obvious question I can’t find the answer to.

Can I please ban ‘it depends’ answers unless they are resolved with a firm expectation? What are the industry standards or tendencies?

If you really need a genre bracket to help me specifically, I make dubstep, but I want this question to be applicable across the genre trap. Share best practice.


well, it sure depends on. If you look at DNB, you will (mostly) have dominant drums (right?), well there are examples talking opposite as well, but it depends of mood overall. Talking about peaks, why don’t you consider RMS value for example rather than peak itself?
From my perspective (looking at one genre, and varieties) there is dnb with stick, there is dnb with roaring snare, there is dnb with some smack, but the rule is that the rhythm itself is dominant and important…

  • well, mastering guys have job to glue it together, to make it “even”, and rich but making all exposed, but not too much sooo it doesn’t ruin balance… i don’t really see reason why you ask this? “I do not know what the accepted levels for frequency occupiers” to me it sounds a bit too deep to think while making music, at least at creative stage, the mixing and mastering afterwards are different story - and there are people who live off of it. Ask some mastering engineer, i’m sure they have answer, maybe regarding balance between different elements, and what it is more important, kick or the flute in the background (both if you ask me xD) - rather than exact value (either Hz or Db). I’ve studied music theory on Academy and the most important rule is to not follow rules, i mean they are there like example. Imagine everyone following the same rule i guess…?

I suggest that you watch this very old and very useful video about mixing, as it is very good!


(random) #3

Here are my (unfinished) Notes for Drums
Source are the Book: Drum programming a complete guide to program and think like a drummer - By Ray E Badness and the Roland SC-55 General Midi Drum Standart

Compare a real Drumset with Human (List 01-13) to Roland the Genreal Midi Pan Specs (Drum Programming.xrns - Renoise - Sampler - Sample Properties - Panning)
it looks similar

01 Kick Bass - 036 C1 (C4) Kick 1 - Foot R
02 Snare - 038 D1 (D3) Snare Drum 1 - Hand L
03 Closed + 1/4 O Hihat - 042 F#1 (F#3) Closed Hi-Hat - Foot/Hand L
04 HiHat - Pedal - 044 G#1 (G#3) Pedal Hi-Hat - Foot L
05 Open HiHat - 046 A#1 (A#3) Open Hi-Hat - Hand L
06 Ride Edge - 051 D#2 (D#4) Ride Cymbal - Hand R
07 Ride Cup - 053 F2 (D#4) Ride Bell - Hand R
08 Tom 4 - 041 F1 (F3) Low Tom2 L
09 Tom 3 - 043 G1 (G3) Low Tom1 L
10 Tom 2 - 045 A1 (A3) Mid Tom2 R
11 Tom 1 - 047 B3 (B3) Mid Tom1 R (Org Tune)
12 Crash - 049 C#2 (C#4) Crash Cymbal1
13 China - 052 E2 (E4) Chinese Cymbal

Drum Programming

Don’t use max Volume (127) for your Instruments, rather some air up (100)
because people with “real” instruments do it also (accent)

the standard tuning of a guitar starts with E-2 and on a bass guitar with E-1
The difference is an octave, that works also well with electronic instruments

(Raul (ulneiz)) #4

@krahz. A question. Is it customary for you to place each sound on separate tracks? Is there a special reason? Could you explain it to me?

For me it is coherent to use separate sounds in tracks, so you have individual control of each sound. But I want to know what people usually do with this type of batteries, or percussion in general.

Do you usually separate percussion with tracks, each track with a sound type?

(random) #5

Basically, this is only (for me) a clear Drumcomputer Layout (Hardware Displays are small) with the difference that mute Groups, Samples are on the same Track as in old Trackers

Apart from that i usually do not use many Samples, Tracks.
Rather variations, Try to make Music that is monoton as possible without sounds boring

(Zer0 Fly) #6

I don’t think any beginner would be all too happy with fixed frequency/db charts per instrument/drum hit. I think such formulas exist (as rough recommendation, per genre…) and are valid knowledge of the sound designers, composers and mixing/mastering engineers, but should not be given too much weight, rather be seen as guidelines and not fixed recipes. For example certain genres are defined by certain characteristics of the mix, and having knowledge of these (i.e. found by analysing songs or experimentation) can be helpful if you try to achieve a similar sound.

It really varies greatly, for example an accoustic drumkit is mixed very differently to a drum&bass breakbeat which will have different gravity centers because of being pitched up and processed, and also being in a very different mixing context.

I would recommend you to study music that sounds similar to what you would like to achieve. Choose some characteristic tracks/passages, load into a daw, and then listen to the sounds over and over again through different types of digital filters, so you can isolate frequency ranges and see what is going on there. Also watch analyzers, and scopes/meters of bandpass filtered music… You will see lots is stitching together, or interleaving sounds in the frequency range, or ducking some sounds in favor of others. Also you will soon notice that many sounds are very complex, going through wide frequency ranges, some more dominant, some less, and also you will see how a good mix will manage overlapping sounds in subtle ways. You can try to imitate the findings from such sessions, but you cannot always have a fixed recipe, each mix will be unique.

Other than that, one could roughly say that a master of a song will have a total frequency curve like pink noise (electronic bass music might have a bit stronger sub/lower bass) and will have a certain total “loudness” measurable in “LUFS”, so it is comparable to other songs. Ofc it will seldom be a flat pink noise curve, if the mixing and mastering engineers were talented and shaped some character into their tasks. You can also easily run a mastered song through a loudness meter and a spectrum analyzer to see how things are going. Basically - if you want your song to sound well next to certain others, try to imitate their mixing/mastering, it all can be found in analysing sessions.

I also once longed for such formulas, but now I think there are 2 things that will help much better. Thing one is good trained hearing and skill, and thing two is referencing to other songs, which you will have to analyze with care. It is really - a good mix depends on ears and feelings, in trying to craft sounds together so they sound natural. My initial experiments with trying to mix too “technical” were very frustrating - I found some training and then fitting together by ear can bring much better results.