JRiver can sound better than Roon, it's scientific ;)

Fletcher and Munson, Robinson and Dadson are right!

Equal-loudness contour sounds to my ears (and should for most homo-sapiens) far superior and much more fun if/when listening at <75dB volume levels (at the listening position).

JRiver with “Loudness” (the common name for equal-loudness contours implementation) just sounds much better to me. Unmistakably.

Otherwise - I prefer Roon, but not sure I could pass an ABX test all settings were equal. With Loudness enabled and calibrated - JRiver wins. Or should I say - Equal-loudness contour wins.

I assert that “Equal Loudness” should be implemented by Roon under DSP Volume* so that people who already use DSP Volume, or don’t mind using it, would enjoy the substantial humanly detectable SQ improvement obtained from Equal-loudness contour curves.

*In JRiver it is also available only under “Internal Volume” which is the same as “DSP Volume” in Roon:

Other volume modes (no loudness):
image

Internal (aka DSP) Volume (with loudness):
image

As for the simple SPL calibration required at the listening position -
Same goes for Room Correction right? Loudness calibration is among the simplest and quickest to do.
Those of us who care about SQ already have a calibration mic or at least an SPL meter.
A $24 Dayton Audio iMM-6 Calibrated Measurement Microphone for iPhone, iPad Tablet and Android will do a splendid job.

“An equal-loudness contour is a measure of sound pressure (dB SPL), over the frequency spectrum, for which a listener perceives a constant loudness when presented with pure steady tones. The unit of measurement for loudness levels is the phon, and is arrived at by reference to equal-loudness contours.”
(https://en.m.wikipedia.org/wiki/Equal-loudness_contour)

5 Likes

It’s usable even without calibration.

  1. Adjust it by ear until you are happy at one specific volume level
  2. Tell Roon to auto adjust by volume

Step 2 works only if you use Roon’s volume control, otherwise it doesn’t know. But that leads to my previous suggestion:

Copy the old Quad volume controls:

  1. A loudness knob, which makes it easy to adjust the Fletcher-Munson curve in the moment
  2. A tilt knob, which makes it easy to adjust a house curve.
1 Like

I believe this is how RME ADI-2 DAC is suggesting to do Loudness calibration (page 14 in its guide).

Perhaps clarify your post a bit so that it won’t confuse folks new to this issue that it is indeed possible with Roon. It isn’t at the moment.

I too would like to see a ‘loudness’ contour option in DSP. Sometimes I add a simple EQ curve to boost the low end when listening at modest volumes but it’s certainly not the same as the Fletcher-Munson curve(s)

2 Likes

Indeed not even close!

btw - here is an open source loudness implementation in CPP -
https://sourceforge.net/p/equalizerapo/code/HEAD/tree/trunk/filters/loudnessCorrection/

(From the wonderful Equalizer APO project)

I’m not familiar with CPP, can this be used in Roon? I see the EQ program is Windows-based, I run ROCK.

No, sorry… Just wanted to point Roonies to a bit of reference to implement :stuck_out_tongue:

Equalizer APO on it’s own can be used (even with Roon) to achieve loudness and much more (a lot of redundancies with Roon’s DSP) but it is APO based… meaning - will work only in Windows with directly connected devices and only in shared mode (i.e. not bypassing the “OS Mixer”…). Not ideal.

I haven’t noticed this option before - Thanks :slight_smile:

By the way - it’s very interesting why all other producers gladly avoids native “Equal-loudness contour” option in their software players and boxes, focusing on upsampling and filtering - which actually improves practically nothing…

"Equal-loudness contour" (ISO 226:2003) it’s obviously pure physics.
Simply physiology of human hearing, widely used in the past.
When you turn it on, your equipment may suddenly turn to dream set, worth 1 Billion$ :slight_smile:

If we do not know what the deal is, then the deal is money, big money.

P.S.
Producers sarcastic comments are very welcome… :wink:

What is this, leveling?

No no — as @Ron_Kuper said, I should be clearer, I meant that Roon could do a simple feature for loudness that doesn’t require a calibration mic. It doesn’t exist today.

Loudness compensation curves were originally based on an attempt to find average equal-loudness levels.
In practice, this ideal will not be approached by using a one-size-fits-all solution.

Roon might consider implementing a tilt control, as in certain earlier Quad or Sonab amplifiers, Luxman C-1000 or a few currently produced preamps from the likes of d’Agostino.
Your subjective preferences should rule!

No built-in standard loudness compensation, chosen without knowledge of the music playing and of sound pressure at your eardrums and your tastes, could be reasonably successful.

Some reasons why:

  1. We are all different in this regard, and our preferences vagy over time.
  2. Bass level curves and bass quality/quantity perception vary between rooms, loudspeaker placement and speaker properties. Headphone listeners discover the importance of sealing. Music and sounds of course are highly variable.
  3. To come near the theoretical ideal of loudness compensation, one needs to measure sound pressure and frequency response at listening position (for headphone listeners: in the auditory canal) and choose the theoretically applicable compensation curve.

So use a “tilt control” to shape a warm sound, if that’s to your liking when listening at low volume. “Set pleasant control”, as Vanilla Fudge once told listeners!

1 Like

Im glad that you found something that you feel sounds better than Roon… feel free to use it. As for me, I feel Roon is best. By the way, it is impossible to prove either opinion scientifically since ‘best sound’ is subjective.

I am curious however… if you prefer jriver, why aren’t you on their forums instead of ours?
The impression I get from reading your comment is that you are sent here by jriver to drum up business. While I’m sure that’s not true, everyone is entitled to their subjective opinion.

My friend - I believe you have missed my point!

Roon is best to me too, I just want to help make it better!

Pls read about equal-loudness contour studies, it is as scientific as it gets.
Many industry standards and professional recording practices are based on these studies, for better or for worst.

In any case I did wrote “can sound better”.

I don’t!
I use Roon but actually combine it with a free open-source project called Equalizer APO to get the loudness function. I even wrote about it here in our community -

If you’re keen in the technical details you’ll notice that -

  1. It is a theoretical compromise in SQ, since this is only supported with the “System Output”, i.e. no exclusive mode, but it really sound much better at lower volumes so “Exclusive Mode” vs. “Loudness” - loudness wins almost every time. Just to communicate the level of improvement and enjoyment with this feature.
  2. Worse - without native support this workaround is only achievable for folks connecting their DAC/AVRs directly to Roon core.

In light of all the amazing things Roon has accomplished, the open source reference and the awesome suggestions made over several threads on this subject - it should be fairly simple for the Roon devs to implement.

The whole purpose of the Feature Request category, right? Help make Roon better.
It might have been a harsh angle to take, naming a competitor like this. I want Roon to be the best player possible, especially SQ wise. No compromises.

Hope your heart is settled.

1 Like

An Equal Loudness curve or a Tilt control sounds like something that could be implemented as a Roon Extension using the API atm. I’ll point the folks in Tinkering/API to this thread.

2 Likes

Yes, and I would add the following, that loudness (and even dynamic compression) are for ambiance music rather than for active listening purposes:
And when we listen to say a distant double bass, do we think we should apply frequency-dependent loudness compensation - or just move closer if interesting ? I see loudness requirement as needed more for some of the audio systems we use than for our direct perception, because our brain “knows” that we do not equally hear all sounds. Otherwise people would use audio protheses with loudness compensation all the time for “augmented” hearing, right ?
I think that the loudness benefits lay more in bringing back some energy from both ends of the spectrum in systems that are chronically deficient in energy in the lower end and somewhat balanced in the high frequencies to avoid sounding too high pitched. Furthermore in small systems already struggling to reproduce bass transients, asking them to move more air increases distorsions and gives a slow and indistinct bass line. Of course, at 75 dB in a say 45 dB-C norm noise background, that can still be better than hearing no bass at all.
But if this is good for ambiance music reproduction, that is not what we usually call “high fidelity”, that involves a reasonable dynamic range. Likewise, dynamic compression might be good for ambiance sound reproduction (level up low sounds above noise level while keeping louder sounds quieter), but compression should be avoided for high-fidelity purposes : it dims music intonations, interpretation subtleties and that sense of intensity of music that makes it so precious.
Hope this helps - maybe some of us are more often in “ambiance music” and other mostly interested in “active listening”, then have diverging opinions as a result. For sure Roon, or JRiver can be used for both purposes.

My understanding is that it should be flat @ >75dB levels and increases gradually as the volume is lowered to compensate for our inherent difference in perception of different frequencies (“phon”) .

So yes - critical listening at above 75dB at the listening position would inherently be without “loudness” curves and that’s OK since we could perceive all frequencies clearly.

Below those levels and certain music becomes less “fun” since some frequencies fall below our perception and are “lost” as the studies shows and the loudness curves aim to fix.

It is our ears that needs compensation at lower volumes, not the system. Definitely not the system.

Try it yourself at low listening volumes and report.

Where are opinios diverge?
It is a feature. Like PEQ. You don’t have to use it.

Opinions obviously diverge on “is it useful or not”, hence the very existence of this thread. I simply say that for ambiance music it makes more sense and is more used, than in active listening, for high fidelity music reproduction.

Loudness is by definition a dynamically variable frequency correction that should also not be confused with a static frequency correction for flat average response, performed most of the time in the intention of transducers and room response correction. I had anonymously been allowed to publish about digital room correction in 2004 (Tact RCS2.2 system) and have since then tried to limit its usage to the bare minimum, better work out the room and system.

In hifi, not only you don’t (or should not) have to use it but in fact many people that love good sound never use a loudness correction (they rather use some static frequency response adjustments). I have seen loudness corrections only used in conjunction with small domestic systems, including by myself very occasionally but not in a “high fidelity” context. Otherwise, I have a limited experience but have never seen anyone use actual loudness in a large system, in 40 years of music loving and hearing tens of them.

Perhaps this is because of an aspect of loudness not as a measurement but as a correction principle that contradicts the idea of natural reproduction, even at moderate level: loudness has to be fundamentally a level-dependent correction, as human ear sensitivity as a function of frequency was measured to significantly differ, at various constant levels. But how do you apply it on many styles of music which is essentially made of transients with constantly changing level ? Especially when it would a priori be most useful, to boost small signals out of our best sensitivity band like bass that is rarely at constant level (or very boring then…). That inherently is not a feature for true high-fidelity.

Finally, not only we don’t listen to sine waves, neither do we have a Bruel and Kjaer frequency analyser between our ears, it is a much more sophisticated signal analysis that we are unconsciously doing when processing music… Our brains can be trained at accommodating biases… or fooling ourselves until we “recalibrate” with actual live music (preferably acoustic).

Yes, you don’t have to use it, and if you use it all the time with benefits, for the most part you might be doing rather room and system correction.

On a well tuned high efficiency system (not an easy one), not need to add any loudness. It has high dynamic range both micro and macrodynamics. Even at moderate level, the low end is articulate, defined and comes through with realistic impact.
As the tonal balance gets perfected it becomes very sensitive and can be fine-tuned using Roon DSP with minute corrections down to 0.1 dB, without exceeding 2 dB if possible.
Adding huge corrections alike loudness curves would risk creating a caricature of accurate reproduction. With less than half a dB incorrectly added, bass already appears fatty and looses its speed, it ability to be both dense and fast, irrespective of level.

1 Like