mixing too loud or too quiet

[kundevelop]

Hello,

I know when mixing is good switching often monitor volume level from soft settings to loud setting to calibrate overall mix on Equal-loudness contour curve. I would like to ask to MEs which issue is common when mixing track always at the same monitoring level: too quite or at contrary too loud level.
For example when mixing all the session at very quite level which frequencies are percieved more? bass or high ones? resulting in a mix with too much bass or too much highs?

Faust

[fondone]

great question.....looking forward for answers

[chironomidae]

Yes

If you mix at a low level, your mix will end up having too much bass and highs. Your ear is most sensitive towards mids, presumably for vocal communication, and the lower the volume the more and more pronounced these differences become. That's why most ME's recommend 85dBs RMS for monitoring - that's the sweet spot, where equal loudness is flattest while your risk of hearing damage is minimal.

[kundevelop]

thank you very much for your replies.

Originally Posted by chironomidae

Yes

Your ear is most sensitive towards mids, presumably for vocal communication, and the lower the volume the more and more pronounced these differences become.

which range identify average human ( male and female ) voice? when you say "towards mids" around which frequencies are you thinking about?

regards

faust

[kundevelop]

According to Fletcher–Munson curves if I should find a starting point to work on balancing a mix done at very quite level, which approach do you think it is more appropriate:

1) gentle wide bell boost around 1500hz.

2) bell boost around 550hz + bell bost around 5000hz littlebit stronger ( according to the graphic of the Fletcher–Munson curves at low levels ).

or something else? I'm not a pro, just a producer of electronic music trying to get the best out of my sound.
thanks in advance.

Faust

[chironomidae]

I just meant that presumably the dip in Fletcher–Munson is related to typical vocal ranges as an evolutionary necessity, in order for people to understand each other better.

You can try to compensate for F-M by using eqs, but it would be much more preferable to work at a higher spl. The problem is that as mentioned earlier, F-M is based on a average among a large number of results, and each person will be a bit different. I think the only way you could hope to compensate for F-M at lower levels would be through careful experimentation using both low and high spl tests.

[riotchild]

Keep in mind the Fletcher–Munson curves apply to pure tones via headphones only.

The updates curves done by Robinson and Dadson are better as they involve a single loudspeaker, but the current standard is different to both of these.

See here...
http://www.sengpielaudio.com/Acoustics226-2003.pdf

This of course doesn't negate any previous comments on perceived loudness.

[kundevelop]

thanks for replies and PDF,didn't know about it...saved!

[Lifted]

Originally Posted by SafeandSound

When mixing too quietly, it would be typical to mix bass heavy due to the inability of the ear to perceive low frequencies as well as the mid band at such levels (you may overemphasis bass instruments with fader level or by LF EQ), the same is true of high frequencies, but to a lesser extent.
The mix would probably be bass heavy with a touch too much top end.

The Fletcher Munson curves were based on an average of many listeners so not everyone will be exactly the same.

If you mix at a louder level 80-90dB your ears response flattens off so a
better balanced mix will result when played back at similar levels.

Some amps (older amps, non professional) had Loudness switches to compensate for the reduced percieved levels at low and high frequencies when playback level was low, or of course tone controls.

Dead on target. Just wanted to add one thing:

When you listen excessively at high volume your ears get tired, and your hearing becomes not as sensitive to an actual sound coming out of the speakers. (I believe high frequencies become dull, especially mixing on the headphones at high volumes).

After a long mixing session, take a break, relax, taking a long hot shower tends to restore your hearing back.

Plus this is always good to get a different view on your work, as over time your brain starts perceiving the mix better than it actually is.

[JPeters86]

I think it's very crucial to mix at different levels. To me, this is especially true during the EQing-phase of the mix (as explained earlier in the thread). For balancing levels, I generally prefer relatively low volumes, this seems to work best for me.

[loji]

The other concern to be aware of is how your listening level accefts your dynamic processess (outside of EQ).

If you listen at a very high level, you will use less compression since it's already loud enough, this may result in a 'quiet' print master when played at regular volume.

However if you listen too quietly, you will push the comp harder to make up the percieved volume, resulting in loss of dynamics and 'squashed' waveforms. This mix may clip when printed.

It would depend on where your monitor point was in the path though.


ie: the reason to use K-system.

[t_d]

where's a good place to get an RMS meter for the studio?

thanks

[kundevelop]

hardware or software?

[t_d]

how would software work? don't you need to measure the actual sound coming out of the speakers? (unless you use the software on another machine with a microphone, etc)

but, yeah, hardware....

Originally Posted by kundevelop

hardware or software?

[t_d]

would this work?

high quality 30-130dB digital sound decibel level meter - eBay (item 220450803371 end time Aug-10-09 17:02:08 PDT)

[Bob Olhsson]

Our goal is for the mix and master to sound great at ANY volume level.

[imixrecords]

Check out Bob Katz' K-System. He's got a great way to solve this issue. There are many posts on this site about the K-System and many PDF's to be found thru google.

[loji]

You want a SPL meter not an RMS meter.

You can get one at radio shack for $50 or a nice one (like NTI Minilyzer) ~$400.


generally use A-weighting for pink noice and C-weighting for music listening. C weight lets 'in' more bass for the measurement.

[t_d]

perfect info, thanks!!!

was wondering about A vs C

i'll start with the radio shack one...

Originally Posted by loji

You want a SPL meter not an RMS meter.

You can get one at radio shack for $50 or a nice one (like NTI Minilyzer) ~$400.


generally use A-weighting for pink noice and C-weighting for music listening. C weight lets 'in' more bass for the measurement.

[dakine]

There's a great iPhone sound meter app that's very accurate

[t_d]

excellent..! i'll start with one of those then... 99 cents seems like a good entry point!'

Originally Posted by dakine

There's a great iPhone sound meter app that's very accurate

[The MPCist]

I start low.. and check mixes from time to time at high levels to see if they 'feel' right.... But 95% of the time, it's pretty low.

[Paul Russell]

I have an SPL meter in the studio and generally mix at 75-80dB RMS, with occasional forays into 80-85dB to see how it sounds.

Most low-mid range nearfields distort at 85dB, dropping the bass and getting piercing in the mid range. If you're going to mix at that level it's important to have a pair of monitors (and a room) that can accurately handle that SPL.

IMHO, YMMV etc..

[t_d]

ok, this was interesting...

i grabbed the iPhone SPL app. figure it's got to be somewhat accurate.. at least within a db or three.

now, i compose and master primarily experimental ambient music. generally low on dynamics, acoustic and electronic.. minimal and gentle.

i held the meter at mix position, pointing at speakers and played the music at my average comfortable level.. and it was reading about 45dB... i cranked the level knob on the Central Station to a level WAY louder than i'd ever want to listen or experience this music, and managed to get the meter to read round 60. i could not imagine working with this kind of music anywhere near 85dB.

i can see 55db as a a "loud" level and 45 being more average.

so, this brings up some interesting thoughts. first of all, it generally confirms what i already know about GS.. which is that most of the information gleaned here is not necessarily relevant to quieter, more avant garde styles of music... ie: it just proves that these "rules" (be it a particular compressor setting, particular piece of gear, etc) really have to be relevant to what you actually create in order to mean anything. when music sheds dynamics in favor of texture or space, a whole lot of things change.

secondly... i'd like to hear from any other MEs or artists in general who deal with quiet forms of music and hear what levels you read if running this kind of test with an SPL meter... see where you come in in relation to my own 45-60dB levels.

[Bob Olhsson]

I think a lots better way to think of this is to imagine a two band limiter with a crossover at 500 Hz. and the band below 500 bypassed.

If you vary the volume, you'll hear the compression above 500 very clearly. The exact crossover frequency will also vary between individuals and everybody has an octave of maximum sensitivity beginning somewhere between 2k. and 5k. depending on the individual. Once you learn to listen for this effect, the implications for mixing and mastering become pretty obvious.

The test for a mix or a master is how well the music holds together as you vary the volume. Volume is a moving target as are individual hearing differences. This is why checking at a variety of volume levels is critical and another reason excessive signal processing can lead to trouble.

[lucey]

Mixing at a high level is mostly just tiring, so why bother for long periods?

If you really know your monitoring you can work at a medium low level most of the time, but you'll want to crank it on occasion to see how it blows up, or holds nice and steady. There is some hard physics here and super low levels are a bad idea, but the more important issue is always the monitoring and your familiarity. My clients that use Avantones for example, always seem to mix well on them. It's not a speaker that's too sensitive to room, placement, volume. But you will need a full range speaker to do a complimentary listen, at low and high levels, and to make fine adjustments once the mix is roughed in.

[Adam Dempsey]

Originally Posted by chironomidae

That's why most ME's recommend 85dBs RMS for monitoring - that's the sweet spot, where equal loudness is flattest while your risk of hearing damage is minimal.

I think you mean "85dBSPL" (though lower level is common, then with a Dim switch on the console, too).

My 'old' Nokia phone has a built-in SPL meter with A, C or no weighting. I can't say it's accurate to industry specs, but it's consistent so it's its own reference, just out of interest for gigs, etc. From what's been said above, seems like it may be more accurate than the iPhone app. Did you hold the phone at ear level? How directional is the phone's mic?

[loji]

i can see 55db as a a "loud" level and 45 being more average.

No .. something seems wrong ... maybe check the weighting, or calibrate against a known working meter. 'A weighting' will read lower (maybe 10dB or more).

~60 C weight is the sound of a mid suburb house with windows open. ie: ambient noise floor. A quiet room would be around 30~40 I'd say.

By contrast 'Loud' is around ~90dB (movie explosion, 'legal' noise ordinance in some states)


the comfy level should be in the mid 70~80 range. I'd double check the Iphone isn't set to something wierd, otherwise at least it's just a buck.

[Mr.HOLMES]

Originally Posted by Bob Olhsson

Our goal is for the mix and master to sound great at ANY volume level.

Bob interesting it never worked for me listening at 80-85 db the whole mix long.
I just start normally mixing very low and in between I go up and make some adjustments than I go back and so on.

this also gives me a feeling how much base is to much when i go back to crank the volume.

[joeydego]

how loud is 80-90 dB? How about mixing at lower volumes but checking your work every so often at this volume?