I've read in several places that it is best to mix at low levels because:
1) our ears don't hype the lows and highs as much, resulting in better sounding mixes
2) our ears don't fatigue as quickly so we can mix longer and the the natural compression of the ear protecting itself doesn't kick in, messing with the dynamics we hear
3) weird room resonances aren't excited nearly as much, so room coloration plays a smaller role in what we hear
4) after mixing at low levels when the song is turned up, the sound just ROCKS! It sounds even better than if we mixed it loud to begin with.
 
Great, I'm in! But what is "low level"? I read that the ear has the flattest response at 85dB SPL, so that should be the target loudness to use. That seems a little loud to me. Not sure I'd call that low level. I also heard that you should mix at a level such that you can still have a conversation with a person sitting right there without shouting.
 
Thoughts?
 
At what level do y'all mix?

MIXING LOW...
 
will only tell you how the mix works at low volume.
 
to properly mix, you must at least check at about 85 db.
 
your mixes will never translate properly if mixed at only low volume.
word.
 

85dB is as close to a standard as there is, but it has been suggested (by Bob Katz if I recall correctly) that for smaller setups with nearfields 79dB might be more appropriate.

At low volumes the Fletcher-Munson effect kicks in, and our ears/brain emphasise the mids, especially in the vocal frquency bands because that's what our hearing has evolved to regard as the most important fequency zone to hear clearly. Wiki has a good article about it here - https://en.wikipedia.org/../Equal-loudness_contour

Attempting to mix with the Fletcher-Munson effect warping our hearing is frustrating and pretty pointless because the mix will end up sounding quite different depending on playback volume.

In the real world, unless mixing entirely for club PA systems we don't know how loudly someone will be listening to music, so checking what's going on and if there are imbalances across the likely range of volumes is a good idea. I'd suggest not worrying too much about very low volumes though, because a mix catering for them won't work outside that situation. All we can do is hope the listener has switched on the "loudness" filter in their playback system - which adds treble and bass (or subtracts mids) in an attempt to overcome the changed perception at low volumes.

Incidentally, 85dB (time weighted) is loud enough to cause potential hearing damage if listened to for long enough periods. A search for industrial/music/entertainment industry noise level guidance will find you lots of information about that. It's quite common nowadays for engineers to track with quite low control room volumes, 80dB or less, partly to reduce potential hearing damage (deaf engineers generally don't turn in good results which upsets their customers and employers) and partly so they can hear themselves think and so they can talk to other people in the control room while tracking.

On my main speakers I might mix at 83-85 dB SPL but when it comes to checking a mono mix on a signle mono small speaker I like to do it at quite a lower volume. Yes you can monitor your mix at 85 dB SPL coming from a single small speaker but I like to drop that down well below that.
 
I love to sit right in front of it and be close to looking at it.  Then turn up only to quite soft levels.  65 dB SPL is a super nice low level.  Yes 20 dB down from 85 dB SPL.
 
70 dB SPL max though as I like to keep this sort of listening in the softer area.  These are all C weighted measurements of course. I hear lots of things down at this volume.  If snares are too loud they jump out here.  Vocals too loud as well really stands out.  Most things out of balance will reveal themselves.
 
I practice listening to reference CD's down at that level in front of the mono speaker from time to time.  It reminds you about when things are very good down there.  To translate well I think you have to get back to your main speakers and I do that at 85 dB SPL.  I also check at 95 dB SPL and 105 dB SPL for a short burst. That tells you a few other things too.  (Bass and reverbs)

Should note that the concept of mixing at low volume to reduce the room's influence is mostly a myth. The ratio of direct to reflected sound remains constant regardless of their absolute SPL values. You still need absorption to raise that ratio. The benefit is that you can't hear low frequencies as well, so those resonances are less distracting. But they're still there.
 
The real value of low-volume mixing is that you hear the midrange more clearly, which is where the real meat of your mix lives. Toward that end, some go as far as to reduce the volume until it's barely audible. It's actually a good technique for determining if you vocals are popping out like they should, especially for pop music.
 
By contrast, always listening at high volume (and I consider 85dB "high") is that you get an equally unrealistic picture that's only valid if everyone listens to your mix at that volume.
 
Since you're after a specific number, I'll offer 73 dBSPL. 

I like to edit at fairly high volumes. That for me means setting basic levels and a good portion of automation. I always check at low levels after leaving a project alone for a few days for actual mix checks because I know my memory has cleared a bit and my ears aren't in compression mode. Low level checks are very valuable for overall balance and includes walking into a room adjacent just to see what track or instrument or is predominant. Mixing is mostly removing the ugly stuff, if something catching your attention on playback - then it probably needs some attention. It can be an arrangement quirk, or tempo shift, or ugly resonance or transient that sticks out and draws my attention.

This is the third or fourth time I've recommended reading the following to answer a question on levels:
 
https://www.soundonsound.com/sos/may14/articles/reference-monitoring.htm
 
This is a detailed article and requires a bit of work, but it's well worth it.

Thanks David very good article and interesting while I had not read it previously I do exactly everything that the article recommends.
 
A few things to note that I have found.  85 dB of pink noise sounds louder to me than music at the same level. Band limited pink noise is also great and Bob Katz has that file on his website as well as full range pink noise too.  It is important that these noise levels are at the right level exactly.  So be careful making them yourself.
 
The only thing I find slightly confusing is 85 dB should be the same in a small room or a big room don't you think. OK in a smaller room the speakers may be working less hard of course to achieve it but isn't 85 dB arriving at your SPL meter no matter the size of the room?  I do understand the differences though.  For me I find I can work very nicely at 83-85 dB in a smaller room and it does not sound too loud to me at all.  But then my threshold may be slightly different after playing a lot of music over the years.  My ears are still very well balanced frequency wise but I may have to push the level 2-3 dB in order for it to sound right to me.  This is where they don't factor that in.  Everyone's threshold may be slightly different.  I think the key is to experiment here.  Start at 85 dB SPL for sure but then modify from there.
 
As to monitoring down low I still love doing that and I find 65-70dB being perfect for these types of operations. And yes low level monitoring really reveals many things too.  The low level monitoring not only reveals ugly stuff but it mainly reveals when something is just too loud/soft eg the snare drum.  At very low levels if your snare is say too loud it will jump out on a small speaker way down soft.  What is cool about this is when you nudge it back down so it sounds right and then power up again at 85 dB SPL on your main monitors the snare will still sound great and present.  It is just that you have now satisfied two speaker systems instead of just one.
 
 

I would refer you to Bobby Owsinski's  Mixing Engineer's Handbook, which includes interviews with various producers. I gather that all of the suggestions above are useful, but not any one of them. Mix at different levels. Some guys like to turn it up and stand just outside the door as one method. I'll listen to different levels on different systems. What sounds good on my nearfield monitors sounds different on my living room system, sounds different in my truck, etc.

mcourter
I gather that all of the suggestions above are useful, but not any one of them.

 
I don't agree. That above statement is saying all of the suggestions are useful but none of them are useful. Which is it?
Mine are very useful. The article that dmbaer linked to is very useful.
 
Listening at low level on a small mono speaker can actually sound like standing outside the door of a loud mix blasting in another room would you believe.  I pretend while listening down low on a small speaker (in mono) is like not being let into a concert until the song is finished and they are making you wait outside the doors to the venue until a suitable time.  It works for me.  (My small speaker has a tiny little tweeter in it and I have put a variable control on the tweeter so I can get it off completely.  That helps the standing outside the door concept.
 
You can get a mix to sound really good on a small speaker and it will sound great on every system you listen on.  The reason some may not be getting that is because you still have not mixed it well.  Simple as that.  With earbuds the main thing there is to control the bottom end a little more and perhaps tighten it up because earbuds cannot handle the bottom so well.  Also with stereo imaging you may have to bring things in a little more rather than real wide.  The earbuds can make hard panned L and R sounds sound a little disconnected.  But you need your bigger speakers though for good earbud compatibility.  The small speaker won't help you so much.