Hey everyone,
 
I've recently come across an interesting obstacle. I usually keep my monitoring levels around 85 dB SPL. I've also noticed that I have a tendency to mix a little bright, which is perhaps due to many reasons: I am trying to mix with the reference in mind, I may be compensating for any deficiencies in my listening environment. I decided to check a couple of my mixes on different systems and at different levels and I noticed 2-6 kHz was particularly harsh, sometimes it was biting all the way up to 10-11 kHz.
 
I started to check my mixes/masters periodically at these higher levels but I realized that this wasn't sustainable. Alternatively, I've tried to train my ear to hear what a harsh mix/master sounds like at lower levels. I feel like I'm getting closer with this approach without destroying my ears.
 
My question is three-fold. How does everyone check their mixes/masters for harshness? Is there an alternative to catch this issue earlier on in the mix without having to turn the levels up so often? And lastly, how do you go about managing and balancing the dynamics of the high end in order to achieve a smooth-sounding mix at any level?
 
I've found that using de-essing and tape saturation at varying degrees tends to help a lot. I use the de-esser to make sure there's nothing in the channel that jumps out at the listener, while also, at times, ensuring a consistent presence for the sound in those frequency bands. I also tend to use a LPF and a tape saturator to darken sounds a bit, that seems to helps.
 
Anyway, I'd like to hear your approaches as this is something that I'm working hard on improving. Thanks in advance.
 
Amine

Hi Amine...one of the things which can help is see how a mix looks like using a spectrum analyser, whatever style mix you are trying achieve find another artist who is similar and see what their mix looks like...then focus on the area you are having trouble with...what does it sound like.
 
Another thing is you should try is on the master remove the very top tip...I use the linear phase EQ set to a shelving filter and remove about 6db at 18khz but it varies depending on my what I am hearing...and def tape sims and console emulators can help to.
 
Also work out what is occupying the top end and sculpt each of those instruments and sounds so they are not all bunching up in the same area...choose which one will be dominant...then EQ the rest...slightly dipping those that are secondary in the frequency band.  
 
And finally remember EQing is like sculpting...the pro channel EQ has wonderful shaping tools...learn how to use them, they will make your job easier...EQ is like colour...the pro channel is great for sculpting but a good pultec emulator will add subtle low end...the linear phase is great for surgical precision on the master bus.
 
I hope this helps.
 
Ben

BenMMusTech
one of the things which can help is see how a mix looks like using a spectrum analyser

Yeah, I have to admit that SPAN is one of the first plugins I insert now as I know I am going to be relying on SPAN more than my ears. I then rely on my chums here on the Songs forum to tell me how my mix is, for which I am eternally grateful 
 
Sorry I can't help with your questions though Amine... I really wish I could.

The first thing I would be doing is listening to really nice well balanced professional mixes in your monitoring environment.  Are you doing that?  Do you know what a great mix actually sounds like.  You might find it is much smoother and warmer than you think.
 
While you are mixing you can be dropping the reference mix in and out to keep you more in line.  When you are mastering you can do that too.
 
When you listen to something really great like a 'Steely Dan' album and then look at it on a spectrum analyser I see the top end falling slowly from about 1K upwards and a nice gentle downward slope. Also use your ears first then the analyser, not the other way around.  I find it does not help you that much at all.  They are better for showing problem areas.
 
I dont think the songs forum is the best place for accurate mix information. I think it is better for song information.
 

Jeff Evans
The first thing I would be doing is listening to really nice well balanced professional mixes in your monitoring environment.  Are you doing that?  Do you know what a great mix actually sounds like.  You might find it is much smoother and warmer than you think.

 
Big +1 to that advice! Get in the habit of listening to your favorite commercial recordings through your own reference monitors. Over time, you ears will actually train themselves (there has been scientific research that shows this to be the case) and they'll set a subconscious expectation of what a good mix sounds like. Your brain will even compensate for weaknesses in the speakers and in your room, and it'll do this all on its own without any formalized training or effort.
 
85dB is too loud for most people and causes ear fatigue too quickly. That level was chosen for the K-system because that's what movie theaters are calibrated for. At the time the K-system was invented, film exhibition was the only arena in which any universal standards existed. It doesn't mean you need to mix at 85dB, as it's only a calibration standard.
 
Try some lower loudness levels until you settle on one that feels comfortable and isn't tiring. The exact value isn't important, as long as you're consistent and your ears have been trained to it. Even very low levels can work for mixing, once you've let your ears adapt. There are some advantages to low-volume mixing, because it forces you to concentrate on the midrange, the most important part of the spectrum.
 
OTOH, monitoring at high levels tends to make you over-compensate for the flattening of the perceived spectrum, often resulting in excessive upper-mids. You can train your ears away from that prejudice over time, but that's true for any level you choose, so it makes sense to choose a level that's not fatiguing.

85dB SPL sounds awfully loud for extended exposure.  In my business that would require hearing protection for long exposure (we target < 77db SPL).  Not suggesting that this is the source of your problem, but pointing out that you might be damaging your hearing unknowingly.

Firstly and no one ever does this is to qualify SPL A or C weighting.  Big difference.  Here is Australia the OH&S spec is 85 dB A for 8 hours.  You will not damage your hearing.  So a C setting because it takes in the bottom end much more is actually louder meaning with the C setting you will be turning it down a little more.  (many people incorrectly have their meters set to A weighting instead meaning more sound pressure level in the room to get the 85 dB reading)
 
Also the best frequency response for the ear is around 85 dB C SPL so don't spend too much time at lower volumes than this at all.  You will start to make incorrect mix decisions.
 
As I have said many times too pink noise at 85 dB C sounds quite a bit louder than music at the same level so do not go on that as your reference either.
 
85 dB C is not loud at all.  I find it in fact a very nice volume.  I am in a bigger room for sure but even if you are in a small room just set up the SPL meter where you are and adjust accordingly.  80 to 83 dB SPL is also a nice volume.  When I talk about 85 dB as well the music is never really there all the time anyway. It drops down to 80 and under for short periods too. I mean the loudest parts of the music is reaching 85 dB SPL.
 
Having said all that I also love monitoring in mono through a small Auratone type speaker and yes I love doing this way down low too eg 70 dB or under.  So much can be heard that way.  It is also really good to monitor up seriously loud too eg 95 to 105 dB for short periods.  You will hear the reverbs and the bottom that way.And as the OP mentioned any harshness in a mix will rip your head off at that level too.  Steely Dan still sounds smooth and well balanced even up at 105 dB.  There is no other way to really get that insight.  But of course only for short bursts.

Jeff Evans
85 dB C is not loud at all.  I find it in fact a very nice volume.  I am in a bigger room for sure but even if you are in a small room just set up the SPL meter where you are and adjust accordingly.  80 to 83 dB SPL is also a nice volume.  When I talk about 85 dB as well the music is never really there all the time anyway. It drops down to 80 and under for short periods too. I mean the loudest parts of the music is reaching 85 dB SPL.

 
Hey, everyone.
 
I appreciate the replies. I'll try to respond to each one in separate posts.
 
I should have mentioned that I did indeed mean SPL C-Weighted, I guess I just didn't understand/wasn't aware that there was a significant difference between the different weighting systems. I do find it to be a comfortable level.
 

Jeff Evans
It is also really good to monitor up seriously loud too eg 95 to 105 dB for short periods.  You will hear the reverbs and the bottom that way.And as the OP mentioned any harshness in a mix will rip your head off at that level too.  Steely Dan still sounds smooth and well balanced even up at 105 dB.  There is no other way to really get that insight.  But of course only for short bursts.
 

 
This is exactly what I try to do, and for the reasons you mentioned. I want to get my mix to still sound relatively balanced and smooth at those levels, especially those upper mids I was talking about. But as you said, you can only do it for so long.
 
The only types of mixes I seem to have issues with that frequency band are IDM-type tracks, or any type of aggressive electronic music. Here's an example of one of the references I used:
https://www.youtube.com/watch?v=mW2M9jz-1Bo

Jeff Evans
The first thing I would be doing is listening to really nice well balanced professional mixes in your monitoring environment.  Are you doing that?  Do you know what a great mix actually sounds like.  You might find it is much smoother and warmer than you think.
 
While you are mixing you can be dropping the reference mix in and out to keep you more in line.  When you are mastering you can do that too.
 
When you listen to something really great like a 'Steely Dan' album and then look at it on a spectrum analyser I see the top end falling slowly from about 1K upwards and a nice gentle downward slope. Also use your ears first then the analyser, not the other way around.  I find it does not help you that much at all.  They are better for showing problem areas.
 

I do reference quite a bit as I don't have a treated environment back home, but you can never be reminded too many times. It's probably the most important step. I should have mentioned that I'm at my parents for the summer and I've been confined to mixing in cans, which is where I noticed some issues creeping in. I'm having to make some adjustments for that.
 
I typically have Magic AB on the master bus with 2-3 masters loaded. I listen to them before starting a mix and throughout the mixing/mastering process as well. As you say, it helps recalibrate the ears.
 
I previously posted one of the references, you'll hear that it exists in a different world from Steely Dan. The arrangements are quite different and those synthetic sounds can be tricky to control. It's a balancing act between a present, somewhat aggressive sound and a smooth one.

BenMMusTech
Hi Amine...one of the things which can help is see how a mix looks like using a spectrum analyser, whatever style mix you are trying achieve find another artist who is similar and see what their mix looks like...then focus on the area you are having trouble with...what does it sound like.
 

 
I do try to use a spectrogram when comparing my work to a reference. I find that looking at frequency level over time to be very helpful.
 

BenMMusTech
Another thing is you should try is on the master remove the very top tip...I use the linear phase EQ set to a shelving filter and remove about 6db at 18khz but it varies depending on my what I am hearing...and def tape sims and console emulators can help to.
 
Also work out what is occupying the top end and sculpt each of those instruments and sounds so they are not all bunching up in the same area...choose which one will be dominant...then EQ the rest...slightly dipping those that are secondary in the frequency band.  
 
And finally remember EQing is like sculpting...the pro channel EQ has wonderful shaping tools...learn how to use them, they will make your job easier...EQ is like colour...the pro channel is great for sculpting but a good pultec emulator will add subtle low end...the linear phase is great for surgical precision on the master bus.

I do think I just need to be more willing to darken some of the sounds in the arrangement. Maybe I can post a version of the master in the next week or so, that way you can see what I'm talking about. Thanks for your reply.