You Need Loudness Correction

[HalSF]

I use loudness correction daily by cranking up the volume.

 

[jkim]

Indeed, not very clear what deltas, they are also off it seems.

These are the deltas measured from Audyssey wrt to the 85dB reference level for different volume offsets:

View attachment 461648

@Hayabusa What is the source of these measurements?

 

[dasdoing]

How did you derive these curves? According to the differences (deltas) in equal loudness contours, the compensation should be mainly on the bass response---also see the study I posted above as well as @solderdude's compensation curves (also Audyssey 's delta curves posted by @Hayabusa).

source: https://github.com/dpapavas/alsaloudness

For one, the equal loudness contours, that are the basis for compensation, are specified only up to 12.5kHz and need to be estimated above that.

imo it is better to estimate the missing data. It is pretty obvious that above 10 kHz, most people need very high or even extremely high SPLs to hear those frequencies.

 

[dasdoing]

This script creates custom curves: https://github.com/grisys83/FIR-Filter-Maker-for-Equal-Loudness

DEV made a thread here: https://www.audiosciencereview.com/forum/index.php?threads/my-custom-loudness-control-tools.50021/

 

[ZolaIII]

It's not derivate, it's evolution of ISO 226. ISO 226 2003 differ from revisions after 2010 (2012, 2018) by inclusion of compensation for sub bass in later one's. As still there are no analog transcevers capable of following it there is no rush in adopting them. You can eventually use long room fundamental (8.5~9m) in controled manner for the purpose.

 

[Quinton595]

Perhaps I am preaching to the choir here, but calibrated loudness correction has been one of the easiest and most worthwhile upgrades to my listening setup in years.

It is widely known the human ear is not even close to "flat," but further complicating this is the degree to which it is not flat changes with volume. An implication of this is that all content is mixed for a certain volume level, and listening above or below this level will change the tonality of the content. The low-bass region is the most impacted.
View attachment 461213
But if we know this, and know how loud our headphones (or speakers) are playing, we can correct for it.

My tools of choice are a UMIK-1 (with miniDSP provided calibration file), REW, and EqualizerAPO.
EqualizerAPO provides "Advanced filter"->"Loudness correction" which relies on leaving the hardware volume knob to a set point and then using the operating system volume control after calibration. During calibration, I used REW's SPL meter and maxed out my Fosi DS2 gain to achieve about 70dBC on the calibration tone.

When active, this is how it looks during typical playback.
View attachment 461214

If you are already familiar with EQ software like EqualizerAPO, you might already have an EQ profile loaded. I did too, and my immediate reaction was too much bass. I have suspected for some time the subjective preference for a bass boost is to serve as pseudo-loudness correction - thus with actual loudness correction applied, you will likely want to adjust your previous EQ profile. I turned my bass down 3 dB, on a profile I loaded from AutoEQ.

I am not very familiar with other EQ suites and platforms, and how to achieve loudness correction with them. My hope is this thread can become a good resource for knowledge sharing around this.
What is definitely clear is that this is a marked improvement over listening without loudness correction. In a reasonably quiet home environment I had no reason to listen to content at the 80dB or more it is often mixed to, and this allows me to enjoy the intended tonality at a more comfortable and considerate volume.

Do these features work with the popular "PEACE" Ui for the Equalizer APO?

 

[birb]

i use Loudness Compensation via Linux Studio Plugins through EasyEffects on my Linux system works great and has 4 curves, ISO 226:2003 (most common for modern), ISO 226:2023 (newest), Fletcher-Munson and Robinson-Dadson (original curves made way back)
did it by ear for setting my reference level

 

[Caldera]

Why does the Eversolo DMP-A8 loudness FIR filter of the DSP look so strange?

It appears that the Eversolo loudness FIR filter mimics the equal-loudness contour of the human hearing system. However, this is not something that should be directly “corrected.” What actually requires correction are the differences between loudness contour curves—for example, when a music recording is mixed at 80 dB SPL and later listened to at home at 60 dB SPL. In this case, the difference between the 80-phon and 60-phon equal-loudness contours defines the appropriate shape of a usable FIR filter (Toole et al., 2026).

Reference
Toole, F. E., Olive, S., & Welti, T. (2026). Sound Reproduction: The Acoustics and Psychoacoustics of Loudspeakers, Rooms and Headphones (4th ed.). Routledge, Taylor & Francis Group, New York and London.

See also my message:

Eversolo DMP-A8 Balanced Streamer/DAC Review

I am a new owner of A8 and I am very happy with the performance. It replaces my WiiM Ultra or Ultra + Topping E30 II (which is nice and cheap combo). I use it as streamer and DAC -> Integrated amplifier. During my selection I tested number of DACs and streamers in my home environment. To my...

www.audiosciencereview.com

 

[JohnnyAudio]

ZROCK by Decware does something like this.

 

[dasdoing]

It appears that the Eversolo loudness FIR filter mimics the equal-loudness contour of the human hearing system.

it does. cringy beginner mistake