Device for Equal Loudness/Fletcher Munson Curve? Do Any Speakers Adapt to This?

[stevenswall]

Is this of any interest?

Adaptive loudness compensation in audio reproduction
Leonardo Fierro

This work involves the study of the psychoacoustic phenomenon of nonlinear and frequency dependent loudness perception, its modeling, and the use of digital filters to introduce an adaptive compensation based on the reproduction level.

Music and sound are mixed and mastered at a particular loudness level, which is usually louder than the level they are commonly played at. This implies a change in the perceived spectral balance of the audio source, which is largest in the bass and sub-bass ranges. As the volume setting in music reproduction is decreased, a loudness compensation filter can be designed to introduce an appropriate boost, so that the low frequencies are still heard well and the perceived spectral balance is preserved.

This thesis describes a loudness compensation function derived from the standard equal-loudness level contours and its implementation via a digital first-order shelving filter, and it documents a formal listening test, designed and conducted to validate the accuracy of such a method.

A bit. Moreso if there were a corresponding product.

 

[A800]

DEQ 2496

DCX 2496 also.

 

[DonH56]

I've always suspect that one reason most rock/pop recordings sound so bright (whether midrangy or tinny) is because the engineers mastering it do so at a zillion decibels and I'm playing it back at lower levels.

IME mixing and mastering is fairly loud, 75 ~ 80 dB, but not really insanely loud.

 

[Hayabusa]

Is this of any interest?

Adaptive loudness compensation in audio reproduction
Leonardo Fierro

Very interesting!
Nice to see there is also an ISO standard describing the equal loudness curve.

Leonardo's experiments only try to compensate for low frequencies, but the equal loudness curves also have effects at high frequencies to a lesser extend.

Would be nice to see for instance how well his findings match the Dynamic EQ on Marantz/Denon receivers.
If I find the time I will measure my Marantz AV8805 curves and compare

 

[Dj7675]

Here is what Audyssey DEQ does at various Reference Level Offsets.. DEQ works well. I would prefer it not to boost highs but as long as you have a downward in room response it sounds fine to me. The other attachment is what the variable loudness from a McIntosh does based on how far you rotate the knob. I had a few McIntosh preamps/integrated amps that had this and loved it before using room correction.

 

[Willem]

Some speaker manufacturers tune for somewhat lower listening levels.
The dynamic loudness function on the RME ADI-2 is a quite fancy implementation because it is adjustable. That is important because room size and speaker efficiency also come into this if the system does not measure spl.

 

[MRC01]

IME mixing and mastering is fairly loud, 75 ~ 80 dB, but not really insanely loud.

Well, that eliminates one excuse for making recordings that sound so bad. They must do it on purpose, perhaps to make it sound as loud as possible on radios & car stereos.

 

[mitchco]

I've always suspect that one reason most rock/pop recordings sound so bright (whether midrangy or tinny) is because the engineers mastering it do so at a zillion decibels and I'm playing it back at lower levels.

Yes, this happens more often than not for folks that don't get that our ears/brain frequency response changes with SPL. Anecdotally, when I was still recording and mixing sound, we had a well known Producer working with an up and coming band and I was the house engineer. When it came time to mix, this guy wanted to mix at "concert" level of 95 up to 103 dB SPL C weighting (that's really loud!). I left the room and put my ear plugs in (lol!). I kept mentioning that at that level, it is going to sound thin and weak at regular listening levels (i.e. 77 to 83 dB SPL). What he (and may others in the industry) did not understand that our ears have the flattest response (relatively speaking) around 83 dB SPL and as the level is turned up, so is our ears sensitivity to bass response. Conversely, below that "reference" level, our ears sensitivity goes down in level relative to other frequencies and hence the reason for consumer loudness controls.

So mixing at "convert level" sounded great as that is how the Producer thought it should sound and did sound like the band at a concert. But at that level, he turned down the bass considerably in the mix in order to have it sound balanced with the rest of the mix. Of course, that resulted with the band ending up with a thin sounding album. While I tried to explain to both the band and Producer that this was going to be the result, they thought I was crazy as they said listen for yourself, it sounds fantastic (at concert level with my ear plugs in lol!). To the band it sounded like their live concert and having seen the band live, I would agree. Now had it been mixed at "reference" level and then turned up to convert level, they would have been really blown away

There were a ton of this going on in the 70's and 80's where rock bands would "rock out" in the control room as if it were a live concert. Did not help that most studios had huge large format soffit mounted monitors with a ton o watts that could pound it out. Nowadays with more education in the field and folks like Bob Katz and others showing the way on how to make better recordings in the 21st century this tends to be less of a problem today. And the move to small two-way "bridge" monitors also helped as they simply don't sound good at "concert" levels...

 

[ReaderZ]

View attachment 46320

Is there a device that will change the frequency response of a system to match this at normal listening levels?

Thinking about using a MiniDSP to make 4 presets for various volume levels, but I'm curious if there is device made for this. Recently Joe N Tell on Youtube uploaded a video with a Yamaha amp that could do something like this but instead of a button it was continuously variable. That would be a good second solution if there isn't a device that does this.

This should be taken care of when the recording is made, not when sound is reproduced.

 

[mhardy6647]

How does this analog variable loudness from Yamaha work?

YamahaCR-2020loudnesscurves by Mark Hardy, on Flickr

The recommended procedure: The "volume" control is set to the loudest level to which one intends (ever) to listen with the Loudness control set to FLAT. The Loudness control is then used to reduce the listening level in lieu of the Volume control. The only problem: those eras of Yamaha components generally had large (sometimes quite large) Volume knobs and much, much smaller Loudness knobs.

DSC_7247 (2) by Mark Hardy, on Flickr
P1020541 by Mark Hardy, on Flickr

This two-control approach to level and loudness was fairly common in the 1950s and 60s, used, e.g., by EICO and Sherwood (among many others).

 

[stevenswall]

YamahaCR-2020loudnesscurves by Mark Hardy, on Flickr

The recommended procedure: The "volume" control is set to the loudest level to which one intends (ever) to listen with the Loudness control set to FLAT. The Loudness control is then used to reduce the listening level in lieu of the Volume control. The only problem: those eras of Yamaha components generally had large (sometimes quite large) Volume knobs and much, much smaller Loudness knobs.

DSC_7247 (2) by Mark Hardy, on Flickr
P1020541 by Mark Hardy, on Flickr

This two-control approach to level and loudness was fairly common in the 1950s and 60s, used, e.g., by EICO and Sherwood (among many others).

Dang, just need a digital device that does this now, that would be excellent to control the max volume I'm going to listen at and have that be flat, and go from there.

 

[JustAnandaDourEyedDude]

Member IVX on his website, describes one of his products E1DA PowerDAC V2, in which he mentions it comes with an iOS and Android app which provides a volume slider based on classical Fletcher–Munson loudness curves. Perhaps his E1DA#9038S and E1DA#9038D products also come with this feature in their app? Although it seems from the ASR review thread on the #9038S that the Gen 3 product will come with a hardware volume control. Could the PowerDAC V2 be used as a pre-amp attached to your phone or Android-based DAP (the phone/DAP acting as a transport only) to feed your power amp or active speakers? Presumably, per some of the previous posts in this thread, use of the Fletcher-Munson curves maintains the tonal balance of the music independently of the volume level for the "average" listener? As opposed to a normal volume control on an analog amp which only knows the signal instant-by-instant? It seems a Fourier decomposition to frequency space is required to use the Fletcher-Munson curves, which requires sampling the input signal over a (moving) time window and not just instantaneously, followed by inversion of the frequency representation back to the time domain. Therefore, it seems this type of volume change is better done in the digital domain before conversion to analog. Also, the reference SPL at which the tonal EQ was originally done needs to be known, and of course the desired gain or volume knob/slider increment (at the current instant) relative to the reference, in order to identify the two Fletcher-Munson curves to use. This is all somewhat confusing to me, so please excuse any mistakes in my expressed thoughts. Also, I am just speculating here, because I understand very little regarding electrical signal compatibility between components.

 

[flaviowolff]

Dang, just need a digital device that does this now, that would be excellent to control the max volume I'm going to listen at and have that be flat, and go from there.

A PC with EqualizerAPO can do it. It has a loudness filter. Then you use the Windows Volume Control to engage it's dynamic compensation.
Though I'm not sure how it would work with those USB dacs, whose drivers make the Windows volume slider directly control the dac itself, such as the Dragonflies.
I'm still astonished that MiniDSP hasnt implemented such a feature in their devices. There are lots of threads begging for that in their forums.

 

[dasdoing]

for Alsa@Linux there is this: https://github.com/dpapavas/alsaloudness
for Jack@Linux there is something coming soon: https://github.com/sadko4u/lsp-plugins/issues/81

 

[snapsc]

While it may be possible to do this...it may not be the best solution....especially for smaller woofers or music played loud. My thinking is that often these drivers are covering up to 600hz and beyond...and when they are driven harder for the purpose of getting more bass, they likely also exhibit more distortion farther up into the more audible bandwidth.

Of course this does not happen to all speakers and not all the time. Generally speaking, my own experience is that you get a better sounding result by adding a subwoofer.....run the mains full range and then blend the sub in where it sounds the best...both the frequency of blend in and the volume.

 

[dasdoing]

While it may be possible to do this...it may not be the best solution....especially for smaller woofers or music played loud. My thinking is that often these drivers are covering up to 600hz and beyond...and when they are driven harder for the purpose of getting more bass, they likely also exhibit more distortion farther up into the more audible bandwidth.

Of course this does not happen to all speakers and not all the time. Generally speaking, my own experience is that you get a better sounding result by adding a subwoofer.....run the mains full range and then blend the sub in where it sounds the best...both the frequency of blend in and the volume.

this is a none-issue
at 83dB you have no EQ
at, let's say, 73dB you will EQ DOWN the mids

see this graph:

 

[snapsc]

Normally the loudness control adds bass to better balance what you hear... the lows with the highs... and the louder you go, the less the boost.

An 83 dB listening level which is safely loud normally sounds a little anemic in the low end without bass boost in many rooms.

My only point is that sonically it is often better to boost bass with a sub and not a tone control. The nice thing about using sap with a sub is to better smooth the bass as well.

 

[RayDunzl]

I don't feel a need to adjust the frequency response with changes in volume.

I guess, if something is playing at a lower level, I expect to hear it with the psychoacoustic changes that come along with the lower level.

If "critically" listening, the volume will be in the flat zone anyway. So...


The "old" Fletcher Munson equal loudness curves:

I tried the ISO curve for my listening levels, didn't like the result with the recordings I played.

100dB low bass when listening at 45dB (a pretty low level) in the midrange? No thanks.

 

[dasdoing]

If "critically" listening, the volume will be in the flat zone anyway. So...

actualy yes/no
you can only hear critically at reference level
if you want to hear critically 10dB below it you need to compansate. that's what this is all about.
If the piano sounds natural at 83dB it wont at 73dB; it wont sound like the piano is played less loud

about the following curve, let me explain:
at 83dB OUR curve is flat. So to get a compensation curve for 73dB we need to substract the original 73 curve from the original 83 curve. the result is displayed.

if we are talking about a device (hardware/software) there will be no boosting. you leave the volume knob at what gives you 83dB > then you throw one of the curves. let's take -30dB level: at 10Hz it would attaunate ~13dB, at 90Hz ~20dB, at 1000Hz 30dB


the effect can be easily tried out for "-10" and "-20" using theses files in a convolver: https://hydrogenaud.io/index.php?topic=83607.msg722690#msg722690

 

[RayDunzl]

If the piano sounds natural at 83dB it wont at 73dB; it wont sound like the piano is played less loud

Will a piano being played at a distance farther from me (lower SPL) have adjusted its frequency response for my relative location?

If I turn down the volume, I imagine the sound to be more remote, and don't miss whatever it is I'm supposed to be missing,.

I've played with the curves, didn't find them personally useful, and don't worry about it.

If you do, that's fine.