TOPIC:

So miniDSP makes some of the best audio gear on the planet. With the miniSHARC you can do more with minimum phase and convolution than most other processors twice the price. Signal to noise is excellent. Everything is expandable and tweakable...

Why don't we have a loudness control? Human hearing changes sensitivity based upon loudness. We create a house curve because at any particular volume, we hear highs better than lows. So to hear all sounds evenly we boost the lows to match the highs. Now decrease the volume of the entire system. Now we hear lows even less. You need to boost the lows even more so that they sound the same at low volumes. We need loudness for the miniSHARC.

When you EQ a system, you are really just EQing for a particular volume. Change the volume and you change the perceived sound. This is the advantage of a true loudness control. I recommend implementing a tweakable loudness contour. Sort of like having a variable house curve for different volume settings. Anyone?

I suppose the workaround is to use shelving filters and set up the other 3 profiles with different amounts of gain on those filters? Just wondering why that is not acceptable?

The problem I see with a "loudness" control is that it's going to be subjective as to what it should be. And having the user specify it will be complicated. For example, if the control actually compensated for Fletcher-Munson, some will say it's too much this or that. Some will actually want low volume to have less bass (the reason for the low volume being that other people in the house are sleeping, say).

I'm not saying it's a bad idea but it sounds like a complete rethink of some kind.

John (who only listens at one volume )

john.reekie wrote:

The problem I see with a "loudness" control is that it's going to be subjective as to what it should be. And having the user specify it will be complicated. For example, if the control actually compensated for Fletcher-Munson, some will say it's too much this or that. Some will actually want low volume to have less bass (the reason for the low volume being that other people in the house are sleeping, say).
...
John (who only listens at one volume )

What I'm looking for is a user tunable and variable loudness contour. In my case, I'm wanting to implement loudness in my car stereo. At home I normally listen at one volume but in the car I have to turn it up and down. I have everything tuned to 0dB. The bass is awesome. But if I turn it down half way, it disappears completely. The drone of the exhaust overtakes the bass notes. So I would need a user programmable loudness contour. In an environment like a car, I would need even more bass boost than a Fletcher-Munson curve because I have road and engine noise to compete with.

I use a house curve on all my EQs. However this curve is only appropriate under certain conditions. What I would like to see is a way to automatically fade between 4 user programmable shelving filters based upon volume. You could even include presets to simulate known loudness curves for people not wanting to set their own.

john.reekie wrote:

I suppose the workaround is to use shelving filters and set up the other 3 profiles with different amounts of gain on those filters? Just wondering why that is not acceptable?

In some cases that might work but in my car stereo I'm already using those presets for different seating positions. I have the gain and time alignment tuned for whoever is in the car.

@ Sly

the loudness control is indeed something that has been asked for a while now.. We just never got to do it.. We'll try to see again what's required to make it happen.. Thanks for the reminder..

DevTeam

This has been a need for 6 years now! There are posts from 2010 about this topic. I'd like to see this implemented for the 2x8 DSP as I use it in my car as well.

Depending on how it is implemented, it could also be of help to those of us suffering from some hearing loss.
From what I understand, hearing aids in effect apply customised 'loudness' compression to assist with frequency dependent loss.

Loudness feature would be great to have! Somethin like the JRiver Media Center loudness feature would be a good starting point to look at. Adding a graphic display of how the boost is applied as volume increases and controls for tweaking the onset and the boost would be going further. I have a both the nanoAVR-HD and some 2x4 HD's. Adding this feature across the product line would be fantastic.

I have tried the Jriver "loudness" feature, and it works amazingly well! I would like to vote for this feature in the MiniDSP as well, the 2x4HD maybe?
Looking at reports on the HomePod, they seem to have this feature as default, which makes sense. Of course, it would also benefit some users if this feature could be turned on/off conveniently...

Dear all

While this topic keeps popping up, we agree that indeed we need to figure out a way to find a correct platform to implement (we have more than 40 products now, implementing on each (knowing it takes 7~8x processing power than a simple volume control) won't happen I'm afraid..

Maybe we could look at the 2x4HD as a test... We'll look into it.
Some of the older/locked platform running on SigmaDSP won't be possible I'm afraid as we know for a fact that each of these run at 99% DSP power (i.e. zero resources left on current implementations.. )

Devteam

Great! 2x4HD is the perfect platform from a playback point of view. No multichannel processing to complicate the issue. It does mean different EQ applied dependent on volume setting, calibration process to determine where flat is, and what if a system cannot play loud enough for the right flat response location which means you never have a flat response.
This capability along with the FIR filter to provide the basic EQ and phase linearization results in a very good minimalist playback capability.

devteam wrote:

Dear all

While this topic keeps popping up, we agree that indeed we need to figure out a way to find a correct platform to implement (we have more than 40 products now, implementing on each (knowing it takes 7~8x processing power than a simple volume control) won't happen I'm afraid..

Maybe we could look at the 2x4HD as a test... We'll look into it.
Some of the older/locked platform running on SigmaDSP won't be possible I'm afraid as we know for a fact that each of these run at 99% DSP power (i.e. zero resources left on current implementations.. )

Devteam

Nice!
There is a need in caraudio, for let’s say a combined remote/presets/dac or usb to spdif with programmable loudness.
With or without box, optional LCD, analog or/and digital in/out.
A cool device like that would be usable with any existing minidsp + all other brands.
I would certainly pay the price of an HD for that.

I agree that it is needed in car audio. That's where I need it as well. I'd be happy to buy a newer DSP for that application.

@devteam

Suggestion: Instead of trying to add loudness to an existing platform, why not create a separate loudness unit/ volume control?

Head unit > loudness > miniDSP > Amps

Leave the source unit at its maximum volume and use the loudness unit as a volume control. It can have a user adjustable loudness contour and IR receiver to integrate with existing remotes. I use an OpenDRC in my system. For me, I would set the head unit to maximum clean volume and disable the volume control in the OpenDRC. Then I could use the miniDSP remote to control both the loudness unit and the OpenDRC. Control the volume through the loudness unit and still be able to select presets on the openDRC...

You could give the loudness unit several user adjustable loudness curves and let them choose a volume threshold for each curve. Then as the user increases the volume, it could fade between the curves. So at low volume you have maximum loudness and at full volume you have minimal or no loudness.

Let's say I have 50Hz boosted 10dB at low volume. Mid volume I have it boosted 5dB and full volume it is 0dB. Between low and mid volume, the processor could adjust the boost level on a log scale so that it fades smoothly between 10dB to 5 dB boost.

A custom adjustable loudness profile will let people pick their own custom house curve according to their own hearing threshold. I don't see you needing more than 4 curves, but the more the better. The plugin would then let the user pick the volume point where each curve is fully active. Between volume points, it would be a fade between the two nearest curves.

Exactly what I had in mind.
It could also include a UMA-8 for a sort of Road EQ, automatic correction based on loudness and ambiant noise.
As in car, this loudness compensation should include the ambiant noise as well and not only the music output.
Loud and at fast speed, would call for a bass boost again, in reverse of the typical loudness curve that reduces bass at high volume.
This sure could also just be done manually if the user can define each transitional target curves:
To simplify:
1 - low noise low volume : bass boost and treble boost
2 - mid noise mid volume: none
3 - mid noise high volume: bass cut and treble cut
4 - high noise high volume: bass boost and treble cut