Proper levels when mixing in a small studio

Since I had a few issues recently with the volume levels of my finished mix once bounced from Sonar compared to what I was hearing while mixing I decided to look into the best way to get a final master bus level...both for mix down and for final master.
 
I usually mix with my master at around -6 or lower to save my ears, and depending on material I have been using K-14 for most of my final mix exports on acoustic material and higher on pop/rock mixes...might even throw in a limiter.
 
K-14 usually ends up being somewhere around -6 db average anyways with peaks getting a little above the 0db mark on my master or 11/12 RMS. K-20 is a little less forgiving unless you mix orchestral and not cinematic orchestral as that's  a different set of rules.
 
But my main concern is the levels to calibrate monitors  so that when you hear how loud it is you get an accurate and consistent way to know. One thing I found out that can really throw you off if you use ARC is that ARC is defaulted at -6...so if you're listening to a louder mix...say running it though the paces you need to remember to put the volume on it to the same if you toggle it on. If you don't it will sound like the nuts dropped right out of your mix and it's simply that volume control.
 
It seems the preferred method is to get a pink noise generator in your master and have it generate pink noise at -20 db and your master to unity gain. Apparently Sonar doesn't include a tone generator, but they can be had for free online as a plug-in.
 
Set up a DB monitor which can be an app in your phone. Set it to C weighted and slow response( make sure the app you get has this ability) and put it in your listening position between your monitors....then pan to each monitor and run the following test. Slowly turn up the volume on the back of your monitor until the Db meter reads at a 75/85 db level. Rule of thumb seems to be 75/80db for a smaller room and 85db for a larger space. Do this for both left and right.
 
A couple of things strike me as maybe not quite right.....would there be a cumulative increase in both monitors together? If so, then shouldn't this be taken into account? To me it seems there would be the cumulative effect of both monitors.
 
Another thing- In the last issue of SOS there was an article and the author( sorry I don't remember) mentioned that you might want to tweak your pink noise for your space if necessary...ok then why have pink noise at all if it isn't an accurate measure? Now the author was talking about something slightly different because he covered getting a balanced mix and I'm talking mainly about Monitor levels. His idea was to pull up the faders so they are just below a measured pink noise level. IOW mix below the pink noise with a few exceptions.
 
My goal is to have my setup such that when I move the volume on my master fader, it corresponds to the same way everyone else is going to hear it in the final export.
 
 
 

you never move the master fader.
 
you move all the other faders going into the master fader, to affect gain changes.
 
the soundcard output, i set to give me a volume of 85 db at the point where i sit between my nearfield monitors.
---that is for final mixing.
i always control the VOLUME i am monitoring at, thru my soundcard control.
my monitors are fixed for a certain volume ceiling, and my headphone amp is set the same.
my 85 db volume setting, corresponds with about -5db on my soundcard output.
in other words, i have HEADROOM on my monitors to go louder than 85db at my seat, but only rarely go there.
 
mastering requires a different scenario and setup, it is similar to this, but different, and i wont go into that here.
 
 
i will mix much lower, and louder, at one point or another.
i will also mix for fine tuning in the headphones, AFTER my basic eq and levels are discovered.
that's how i do it.
 

YOU KNOW,
I hate windows media player...
it does something terrible to the sound, and i can't figure out what it is.
bypass everything, still sounds bad..
 
when i want to listen to stuff on a computer, with a computer speaker setup, that kind of thing, i use WINAMP.
 
works very well.
 
my sound card (pci card, Maudio Audiophile 192) has a control panel that lets me save setups,
so i have a setup for analog tracking, midi tracking, basic mixdown, heavy sample mixdown, soft, loud, and another setup for mastering.
 

Since you quoted something from SoS, maybe you already read this last May:
 
http://www.soundonsound.com/sos/may14/articles/reference-monitoring.htm
 
Very good write-up.  This is the first time I saw the advice about moderating the reference level for a small listening space, but it makes perfect sense.  Furthermore, had I not done so, my calibration would have produced an unpleasantly loud result.  Using the 75 to 78 level recommended (my room size is borderline between small and moderate), the level seems comfortable and exactly right.

One thing I don't recall reading in link dmbaer posted. Most recommend using a sound level meter with C-weighted measuring capability. Here is a description of it.
https://www.noisemeters.com/help/faq/frequency-weighting.asp
 
In a nutshell C weighted is supposed to account for the response of the human ear at higher volume levels making it excellent for a measurement such as this.
 
Although K-20 seems to be mentioned more than K-14 in the Bob Katz system of measurement. I like K-14...K-20 seems to push above the boundaries typical of most of my mixes, the exception being orchestral music.

Thanks for this info. I'll check it out as soon as I can!

A good test signal for setting up correct monitor levels is band limited pink noise.  You remove everything sharply below 500 Hz and above 2 KHz with some steep filters.  As a result there will be a level change and it needs gain to be added to compensate.  The quality of this test signal avoids all problems in the low end of your room.  It is not feeding anything into the room where the room may be unstable or inaccurate.
 
eg if you are using a K-14 ref level a sinewave with its peaks reaching -14 dB FS will produce 0 dB VU on the VU meter.  Full range pink noise at K-14 reaches about -1 dB on the VU meter which is to be expected due to the randomness of the signal.  If you create the band limited pink noise filter, the signal needs close to 8 dB of extra gain to make up the level.  This depends on the steepness of the slopes around 500 Hz and also 2 kHz.  Switch the filter in and out and note the VU meter changes.  And also the SPL changes in the room.  There should be little difference.  I adjust each monitor for 83 dB SPL individually and that leaves me with a nice 85 dB SPL when both are on.  You should check the volume settings on the rear of your active monitors as well for fine tuning level there.  They are not always exactly the same even if set in the same position. 
 
Full range pink noise at the ref level and at 85 dB SPL is quite loud compared to music at that same level.  Music sounds quieter to me at that same  SPL level.  Your ears are hearing at 85 dB SPL and are in a good place to balance EQ and levels well.  I find having a permanent SPL meter in front of me keeps me honest and in control.  If you are listening at quite precise levels while you mix I feel it helps you mix better too.
 

Just to let people know the band limited pink noise (500-2K) that has been posted on Bob Katz's website above by Paul  is already perfectly gain adjusted so there is nothing to do with those signals. They have been corrected to read the K-20 level.
 
When comparing to a sinewave BTW at exactly the same level you will notice both pink noise test levels are about -1dB down on the VU meter. This is quite normal.  Probably due to the random nature of the pink noise signal.  I tend to use the sinewave test tones for exact level line up and any voltage measurements eg from your interface. The noise is great for monitor calibration and setting SPL levels etc in your control room.  Avoid using sinewave test signals for these types of applications.  With sinewaves you only have to move the SPL meter a small distance for wildly varying levels etc..Pink noise is much more consistent across the spectrum.  Band limited pink noise is even better as it excludes the bottom end of your room.
 
It is handy to add 6dB and 8 dB of gain to the pink noise signals for the purposes of making a K-14 and K-12 versions of those test signals.
 

The C weighting listens to more (bass) of the spectrum hence it takes in the low end of the music much more.  With the band limited pink noise however the C weighting is not so important but I would still use it on that setting none the less.  As you say Tim it is just leaving out the bottom end to not involve the low end of your room.  The gain however of the band limited pink noise is adjusted to compensate.
 
Tim I work at K-14 a lot and K-20 at times. (for very high quality projects) It makes no difference. The only difference is at K-20 everything is just 6 dB lower down right throughout your project that is all, hence more headroom.  The VU meters need to be recalibrated for K-20 though, hence they still show the same readings eg everything around 0 dB VU.
 
And of course the SPL level in the room also drops by 6 dB so you have to turn up your monitoring level by 6 dB so you back to 85 dB SPL etc..
 
The only thing is you end up with a mix that is sitting around -20 dB FS instaed of -14 dB FS.  But you have got 20 dB of headroom for transients and things hence it will sound a tad nicer.  But in mastering though you have got further up to go to get the track loud.  eg if you want to end up with a loud punchy master around say -10 db FS then from K-14 you only have to add 4 dB of rms gain. But at K-20 you have to add 10 dB of rms gain. So there is just a little more work to do that is all.
 
Another reason why I like K-14 a lot too is my Yamaha digital mixer is calibrated for K-14 too. ie I get +4 dBu at the output exactly when the I am at K -14.  Meaning the output of my digital mixer can produce a whopping +18 dBu at 0 dBFS. Something that many average audio interface cannot even approach. There is a big hefty power supply inside the digital mixer which has no problems producing very high output levels with no distortion or clipping in sight.

I try mixing at a -4.5 db level on the master track.
That way i can provide enough head room to boost up when needed.
Usually with the Boost11 plugin.