What db should i mix at?

wavewar3

New member
I'm pretty new to the mixing world. I use fl studio. I think the meter to the right of fl's mixer is the peak meter , and I guess when people say your highest kick should be ___ or ____,they refer to that meter. I finished a dynamic processor series from lynda.com and have seen a little bit of their eq and filter series. Besides those two things , I have no clue about other mixing techniques or even where to learn. Also , what's the difference between db and dbfs? I've been given links but I was just confused while reading them. Except one that said db only measures the difference between two sounds and it's not right to say that, 'a rock concerts average db is ___ db'. and dbfs is decibels in relation to 0 db(i guess the maximum db a system can handle before clipping). Not sure if that was accurate, but would really like an explanation.
 
actually. a guy on my last thread posted a good explaination on the difference between db and dbfs. just confused on one thing. monster cat said as a tip 'the loudest sound in your track should be your kick , no louder than 8db'. do they mean -8dbfs?
 
dB is more informal language and difference between two different volumes and so on (for example boost a sound with 6 dB and it will double in perceived volume).

dBFS is how far the sound is from absolute zero in the digital domain (before it clips), such as the kick is peaking at -12 dBFS, so if you boost it more than 12 dB it will clip (except if you're in 32-bit).
Then if you go into the analog domain, you'll see the term dBVU, which refers to the analog "sweetspot"-volume (which equals to -18 dBFS).

When talking about dB's in the real life (such as your rock concert example), you use the term dBSPL, which simply means the loudness/pressure of sound (0 dBSPL = no sound).

When mixing, you should keep a good amount of headroom from the start - if your sound is too loud going into the mixerchannel, cutting the channelfader won't help, you'll need to cut the level in the first volume-control.

Then a common habit is letting the sound with the loudest peak, peak around -10 or -12 dBFS. There are a lot of different opinions in which number you should use, so pick one you're happy with.
Further on, because you're in FL, you technically even don't have to care about how loud your sounds are (since it uses 32-bit), as long as you don't go above 0 dBFS in the Fl Studio output level (where FL's 32-bit converts into the default bitdepth of your computer), buuut that would still be a bad habit, so keep some headroom :)
 
^^^^^
nice to see someone else referencing the bob katz article (the k-metering system I refer to below is his brainchild)
--------
end edit
--------

dbVU can be any one of several points in the dbfs scale (i.e. -18dbfs is not the sweet spot just one of many)

k-20 means 0dbVU = -20dbfs - orchestral or cinematic mixing
k-14 means 0dbVU = -14dbfs - pop record mixing
k-12 means 0dbVU = -12dbfs - broadcast mixing (tv/radio)

we are also more likely to use dbRMS than dbVU (although both are equivalent in some respects, they differ in others)

then we have:

all the various units referenced in this wikipedia article Decibel - Wikipedia, the free encyclopedia

also see

dB: What is a decibel?

and

dB dBu dBFS dBV to volts audio conversion digital - calculator volt to dBu and dBV dB mW SPL dB decibels 0 dBFS - convert dB volt normal decibels relatioship relation explanation analog audio absolute level true rms convertor converter decibel to dbf
 
Last edited:
0 dBSPL = no sound

The decibel is a tricky unit to tackle, as - for example - even 0db(SPL) isn't really "no sound at all" - it's just there reference limit of (theoretical) audibility. Not that this is particularly important, but I guess it just goes to show that it's one of those things that you don't want to skim over and go "oh I get it", but rather actually read about it if you want to garner some proper understanding.

And by "you" I mean everybody :)
 
Get a digital SPL meter, hold it where your head would be during mixing, play pink noise out of your speakers and turn up the volume until it reaches 85dB SPL. Mark the position of your volume knob so you know to turn it there whenever mixing. That is a standard mixing volume. This is due to the Fletcher Munson Curve.

Hope this helps! Pay no attention to dB FS as that is a digital representation of overall loudness.
 
Last edited:
Pay no attention to dB FS as that is a digital representation of overall loudness.

Or rather, pay attention but do realize that what your DAW meters show don't bear any relation to the actual sound pressure level coming out of your speakers.
 
Get a digital SPL meter, hold it where your head would be during mixing, play pink noise out of your speakers and turn up the volume until it reaches 85dB SPL. Mark the position of your volume knob so you know to turn it there whenever mixing. That is a standard mixing volume. This is due to the Fletcher Munson Curve.

Hope this helps! Pay no attention to dB FS as that is a digital representation of overall loudness.

actually it is not due to the Fletcher Munson curves or the ISO Equi-loudness contours: these are only explanations derived from experiments with thousands of participants as to the average level a pure tone at some given frequency needs to be, to be perceived as being at the same loudness as a pure tone at 1kHz.

and the level is 83dBSPL not 85dBSPL: check the Katz article referenced earlier - How to Make Better Recordings Part 2

dBfs is not a representation of overall loudness: 0dBfs is the absolute maximum that a specific daw can handle; every other measurement of dBfs is relative to that absolute maximum; -12dBfs means that the signal level is 1/15.85 of the absolute maximum the signal, ca. 0.063 x absolute maximum amplitude; -20dBfs means that the signal is 1/100 of the absolute maximum level i.e the relative levels can be determined by using the formula

level = 10[sup]dBfs-value/10[/sup] (we assume that the minus sign is kept intact in the process)

e.g.

signal is a peaking at -15dBfs

level = 10[sup]-15/10[/sup]
level = 1/31.62
level = 0.0312 x absolute maximum level

signal is averaging -40dBfs

level = 10[sup]-40/10[/sup]
level = 1/10000
level = 0.0001 x absolute maximum level

i.e. decibels in general are about a ratio, dBfs is the specific ratio of a level referenced to the maximum amplitude (level) available; dBSPL is referenced to a nominal 0 pressure level; both are logarithms of that ratio
 
Last edited:
actually it is not due to the Fletcher Munson curves or the ISO Equi-loudness contours: these are only explanations derived from experiments with thousands of participants as to the average level a pure tone at some given frequency needs to be, to be perceived as being at the same loudness as a pure tone at 1kHz.

and the level is 83dBSPL not 85dBSPL: check the Katz article referenced earlier - How to Make Better Recordings Part 2

The K-System is just one theory of mixing, and it mainly pertains to mixing certain styles of music (K12, K14, K20) and getting an appropriate dynamic range using rms style metering. 83dB SPL is Bob Katz reference level, but he also suggested that you mix at a level of your own personal taste. I stick to 85dB SPL because the Fletcher-Munson curve is most flat at around 90 dB SPL and it is what audiologists recommend as a maximum volume for extended periods of listening.

In relation to the original post, I didn't actually read your question but read the title lol. Your description of dB FS and dB SPL is accurate, and bandcoach provided some good technical information on that matter.

Hope all this helps :D
 
The ones that you feel the most comfortable with.

Start low and then give some loudness to see how the bass translates to your room, then go low again.

Mix it up but most of the times low wins (less headaches, more time for the ears to mix and less room artifacts if your room is not the ideal one)
 
I stick to 85dB SPL because the Fletcher-Munson curve is most flat at around 90 dB SPL and it is what audiologists recommend as a maximum volume for extended periods of listening.

The Fletcher-Munson curves have been deprecated by most audiologists in favour of the ISO- iso-phon curves instead, which are anything but flat at any level

consider this and the discussion of the Robinson-Dadson results vs ISO data and the older fletcher-Munson data - lindos.co.uk - Equal-Loudness Contours

I also suggest checking this one out - Moulton Laboratories :: The Equal Loudness Contours - it has some useful tips for mixing and includes your suggestion of 85dBSPL, although in the context of this is MY preferred level to mix at not this is a standard level to mix at

I do note that a large number of architects prefer to use the Fletcher-Munson data - it provides less problems than either R-D or teh ISO data
 
The Fletcher-Munson curves have been deprecated by most audiologists in favour of the ISO- iso-phon curves instead, which are anything but flat at any level

consider this and the discussion of the Robinson-Dadson results vs ISO data and the older fletcher-Munson data - lindos.co.uk - Equal-Loudness Contours

I also suggest checking this one out - Moulton Laboratories :: The Equal Loudness Contours - it has some useful tips for mixing and includes your suggestion of 85dBSPL, although in the context of this is MY preferred level to mix at not this is a standard level to mix at

I do note that a large number of architects prefer to use the Fletcher-Munson data - it provides less problems than either R-D or teh ISO data

Thanks for the information :D Never heard of iso-phon curves. Looks interesting. I guess in the end it really comes down to mixing at a level that you are used to and relating your current mix to a great reference with a spectrogram. That's mainly what I do when it comes down to it, and I just enjoy mixing at louder volumes but quiet enough that it doesn't damage my ears :P
 
Back
Top