Should you use Fletcher-Munson loudness compensation?

[ebslo]

I tend to do the opposite of loudness compensation. If I'm listening at low volume it's because I want it to sound quiet, either just because I feel like it, or because I don't want to disturb/wake others. So for quiet listening I use an eq profile with -6dB low shelf at 100Hz.

Then when I want to crank the volume way up, it's because I want to feel it in my chest. But with the regular room curve, the highs get a bit oppressive before it starts to sound really powerful. So I use a profile that gives <40Hz a couple extra dB, and a -3dB high shelf at 1kHz.

Just wanted to throw a different perspective into this, as it's a case where my actual preference is opposite what the science indicates my preference "should" be.

 

[Soundmixer]

The recording engineer has should have already accounted for human hearing when he creates the recording master. If he is older, he sometimes does not trust his own ears and will depend on a visual representation of levels.

Bingo!!!!! The unfortunate thing here is once you create the master( and encode it) and it is out in the field, there is no guarantee what you will hear. Everyone's room is different, their sound system is different, their ears are different, and their audio taste is different. The sheer variability here is staggering.

 

[dlaloum]

I tend to do the opposite of loudness compensation. If I'm listening at low volume it's because I want it to sound quiet, either just because I feel like it, or because I don't want to disturb/wake others. So for quiet listening I use an eq profile with -6dB low shelf at 100Hz.

Then when I want to crank the volume way up, it's because I want to feel it in my chest. But with the regular room curve, the highs get a bit oppressive before it starts to sound really powerful. So I use a profile that gives <40Hz a couple extra dB, and a -3dB high shelf at 1kHz.

Just wanted to throw a different perspective into this, as it's a case where my actual preference is opposite what the science indicates my preference "should" be.

Some members of our household, find things "too loud" - when patently (verified by SPL meter) they are not... Turn down the bass, and suddenly things are no longer "too loud".

The words are not communicating what they think they are communicating.... but the result is clear - the bass needs to be turned down. (and this is in a system tuned to my preferences, which is relatively flat - with little or no Harman style bass boost).

Some people's preferences are effectively an inverse Harman bass ... -10db up to 100Hz....

 

[Sal1950]

Some members of our household, find things "too loud" - when patently (verified by SPL meter) they are not... Turn down the bass, and suddenly things are no longer "too loud".

Many do not wish to "feel" the music, it is only a background noise in their life.
We wish for the music to become a part of us and the beat of the rhythm enters our soul.

 

[dasdoing]

for some people the concept seams to be hard to grasp.
take the following recordings. we all more or less know how loud they are in reality. adjust the volume until you feel it is the most realistic. THAN put the volume knob way down. does it still sound like the real event? no it doesn't. it sounds like a synth or some other synthetic sound generator:

Freesound - Thunder and Rain 2 by recordinghopkins

Freesound: collaborative database of creative-commons licensed sound for musicians and sound lovers. Have you freed your sound today?

freesound.org

Freesound - small plane w_return passby and ambience.wav by patchen

Freesound: collaborative database of creative-commons licensed sound for musicians and sound lovers. Have you freed your sound today?

freesound.org

loudness compensation will bring back the realness of these events at low volume. that's what it's all about. no matter if it is a real event, or what "the producer intended"

 

[krabapple]

Equal loudness contours aren't 'preference' curves. That's not what subjects in those tests were asked to state. They were asked to state when tone A sounded the same loudness as tone B.

But of course EL compensation is for making the perceived balance of bass and treble remain the same throughout a range of output levels. It's for maintaining a similar listening experience as much as possible within that range. It's not for preventing neighbor disturbance, or helping your spouse stay asleep. AVRs have night listening and subwoofer control modes for that.

 

[Sal1950]

But of course EL compensation is for making the perceived balance of bass and treble remain the same throughout a range of output levels. It's for maintaining a similar listening experience as much as possible within that range. It's not for preventing neighbor disturbance, or helping your spouse stay asleep. AVRs have night listening and subwoofer control modes for that.

Exactly, the requirements are the opposite.
Bass especially seems to lose it's impact when volume levels go down so a good F-M curve tends to continuously increase bass level as the volume is lowered.
If you want to avoid neighbor disturabances, you wouldn't want the bass to increase, you would rather want it lowered.

 

[dlaloum]

Exactly, the requirements are the opposite.
Bass especially seems to lose it's impact when volume levels go down so a good F-M curve tends to continuously increase bass level as the volume is lowered.
If you want to avoid neighbor disturabances, you wouldn't want the bass to increase, you would rather want it lowered.

Quite....

But some people want things to be at "background" volumes - that's how they listen to things.

The last thing they want is for things to sound "real" - as that would disturb their perspective on the home....

This is not about neighbour disturbance....

 

[ebslo]

We wish for the music to become a part of us and the beat of the rhythm enters our soul.

Equal loudness contours aren't 'preference' curves. That's not what subjects in those tests were asked to state. They were asked to state when tone A sounded the same loudness as tone B.

But of course EL compensation is for making the perceived balance of bass and treble remain the same throughout a range of output levels. It's for maintaining a similar listening experience as much as possible within that range. It's not for preventing neighbor disturbance, or helping your spouse stay asleep. AVRs have night listening and subwoofer control modes for that.

Exactly, the requirements are the opposite.
Bass especially seems to lose it's impact when volume levels go down so a good F-M curve tends to continuously increase bass level as the volume is lowered.
If you want to avoid neighbor disturabances, you wouldn't want the bass to increase, you would rather want it lowered.

Sure, but why wouldn't I just turn the volume up?

 

[GalZohar]

On one hand not increasing the bass back might cause less of a disturbance at a given volume level. However, increasing the bass can allow you to hear things nicely at even lower volume levels. For example, if you listen at -30db with Dynamic EQ on you will get some balance and your lowest bass will be about -15db. If you listen at -15db without Dynamic EQ you will probably cause much more disturbance than -30 with Dynamic EQ on, and have less balanced sound. Also, I find that the bass itself isn't always what is the biggest issue in terms of disturbance.

Basically, if you have specific problems with bass being a disturbance, you might want to just turn that down to be able to play other frequencies loud, turn everything down to maintain some balance, or turn everything down even more with loudness compensation to keep things balanced.
If your noise problems aren't specific to bass, you need to turn down everything anyway, so might as well keep the balance with loudness compensation enabled.

Personally I prefer less loud but balanced rather than louder but out of balance.

 

[dlaloum]

On one hand not increasing the bass back might cause less of a disturbance at a given volume level. However, increasing the bass can allow you to hear things nicely at even lower volume levels. For example, if you listen at -30db with Dynamic EQ on you will get some balance and your lowest bass will be about -15db. If you listen at -15db without Dynamic EQ you will probably cause much more disturbance than -30 with Dynamic EQ on, and have less balanced sound. Also, I find that the bass itself isn't always what is the biggest issue in terms of disturbance.

Basically, if you have specific problems with bass being a disturbance, you might want to just turn that down to be able to play other frequencies loud, turn everything down to maintain some balance, or turn everything down even more with loudness compensation to keep things balanced.
If your noise problems aren't specific to bass, you need to turn down everything anyway, so might as well keep the balance with loudness compensation enabled.

Personally I prefer less loud but balanced rather than louder but out of balance.

What I was trying to say is that perceived loudness has a lot to do with spectral balance....

And the studies on the "average preferences" and "average responses" are just a majority perspective.... they are neither wrong nor right, just a particular perspective based on more people preferring or perceiving in a particular way.

Just like the "Harman Target Curves" - which are the preference of around 60% of the population..... but that also means that they are NOT preferred by around 40% of the population... which is a sizeable minority!

To add to this around 20% of people have "misophonia" - which means they perceive sound substantially differently from the other 80% (!!) - these would undoubtedly be among the 40% who do NOT prefer the Harman Target.
And the response of this minority to the audible spectrum is very different. (this is an area which has had very little study...)

 

[ZolaIII]

I have my own built preference regarding presentation disregarding of it everything goes trogh two point of equal loudness normalization; on material source I use EBU R128 and on reproduction part its ISO 226 control on my amplifier or taken in consideration (two 15 dB loudness apart) EQ's with hedaphones.
Why? Because I want to hear all there is in material in any given circumstances as much as possible.

 

[DSJR]

My speakers have a subtle kind of 'F-M' loudness curve designed into them (the maker made a big thing of this ten or more years back, but seems to have changed his view a little in recent times and new models). My ears now work against this curve due to age, damage and acute Rhinitis, so I'm increasingly going back to livelier, more neutral if not upper mid forward speaker considerations again.

 

[ZolaIII]

@DSJR I do subtle adjustment (±2.5~3 dB all together or two loudness level's I can distinguish) to uper mids - female uper/over vocal frequency range (3 - 4 KHz) separation and boost because purely subjectively I like it like that (a lot and on speakers only).

 

[DSJR]

@DSJR I do subtle adjustment (±2.5~3 dB all together or two loudness level's I can distinguish) to uper mids - female uper/over vocal frequency range (3 - 4 KHz) separation and boost because purely subjectively I like it like that (a lot and on speakers only).

I don't have the room for it, but I have a trusted old five band graphic equaliser I once used (and was all but ridiculed for). I don't use a streaming option primarily so can't invest in DSP solutions and the room boosts bass far too much anyway. As said above, I'd rather get the basic sound balance as 'right' as possible before tinkering further and currently, it isn't even before my ears are taken into account (our minds can accommodate a lot - up to a point). I can still tell if a 'hyper detailed' high end system screeches the detail out or not

 

[ZolaIII]

@DSJR I use what I like to call open box system, PC really in a form of deticated music laptop able to work fanless and mostly do a network streaming from there to amplifier. How ever for mentioned vocal adjustments I use simulation of old passive tube stage EQ from 50's.

 

[Sal1950]

I don't have the room for it, but I have a trusted old five band graphic equaliser I once used (and was all but ridiculed for).

EWW, not an equalizer, they just destroy the sound of a good system.

 

[DSJR]

EWW, not an equalizer, they just destroy the sound of a good system.

Yeah, I know - and I'm assuming you're chuckling here rather than actually meaning it? It's a basic two channel five band JVC midi sized box which I bought for relative pennies. When I first used it a long time ago, I 'imagined' the sound was lightened slightly, but I was still dealing with audiophiles back then and now I don't care really

 

[Sal1950]

Yeah, I know - and I'm assuming you're chuckling here rather than actually meaning it?

Basically yes, just sarcasm.
But there is some truth to the way cheap equalisers handled the analog signals and the phase distortions they can induce. I'm not qualifed to add more detail and without measurement of particular units can any concrete comments be made..

 

[DSJR]

Basically yes, just sarcasm.
But there is some truth to the way cheap equalisers handled the analog signals and the phase distortions they can induce. I'm not qualifed to add more detail and without measurement of particular units can any concrete comments be made..

I should maker space for it, give it a go and if promising, maybe look for a classic used Klark Teknik.

They still make 'em though and this price isn't bad at all. Scares the bejesus out of me though as it's totally opposite to my brainwashing/conditioning over nearly half a century that says these things are to be totally avoided...

B Stock : Klark Teknik DN360 Dual 30 Band 1/3 Octave Analogue Graphic Equaliser - Andertons Music Co.

Get the B Stock : Klark Teknik DN360 Dual 30 Band 1/3 Octave Analogue Graphic Equaliser at Andertons Music Company today!

www.andertons.co.uk