TOPIC:

Anybody could pick up the miniDSP remote and accidentally hold the up button for a second also. I mean, c'mon, you've made your point for crying out loud.

If this is an over-riding concern for you, there's only one way to solve it 100%. You need to put hard limiters in the circuit. Maybe that means you put attenuator pads between the SHD and your power amps.......maybe you switch to much lower powered amplifiers.....or some other positive limiting action or device.

Dave.

Limiters are not the answer to broken software.

Additionally, leaning on the volume control wouldn't work, as the pain from 120+ db would have you leaving the room at a running pace before it climbed that high.

I'm not going to cripple my system's performance capability, that I occasionally use, to protect my family from a poorly implemented interface.

drdyna wrote:

Limiters are not the answer to broken software.

Additionally, leaning on the volume control wouldn't work, as the pain from 120+ db would have you leaving the room at a running pace before it climbed that high.

I'm not going to cripple my system's performance capability, that I occasionally use, to protect my family from a poorly implemented interface.

Nobody is forcing you to ever use a touchscreen to control Volumio. I never do. The need to 'protect' your family from your system is also to a large degree because you've got such an over the top system .

Yes the volume control in Volumio doesn't work very well but you're rather exaggerating how serious a problem this is.

The problem doesn't rely on the touch interface, a mouse click would work just as well.

The interface is there to be used, period.

I'd like to use it, and I do use it. Mouse, touch, whatever. I appreciate all the suggestions of "don't use it then" or "lower the limiters" but this wasn't a cheap box, and I don't think it's too much to ask that the interface operates correctly, displays correctly, and offers a couple of features that make it safe to use without jumping through hoops.

drdyna wrote:

The problem doesn't rely on the touch interface, a mouse click would work just as well.

The interface is there to be used, period.

I'd like to use it, and I do use it. Mouse, touch, whatever. I appreciate all the suggestions of "don't use it then" or "lower the limiters" but this wasn't a cheap box, and I don't think it's too much to ask that the interface operates correctly, displays correctly, and offers a couple of features that make it safe to use without jumping through hoops.

I don't believe there is any chance of you accidentally putting the volume dangerously high with a mouse click. I've just had a play using the trackpad on my laptop and it really isn't that bad.

If you didn't want hardware limiters you could set attenuation in the plugin, on inputs or outputs rather than the master volume. This would limit the maximum volume that could be achieved if you ever did end up with a 0 dB master volume setting by mistake.

The interface works. You are absolutely right it could be improved on though.

This is hardly a new issue for audio systems.
I used to "accidentally" crank up my dad's Klipschorn system fifty years ago and he'd get PISSED at me.

So, don't be coming on here whining about some horrible miniDSP safety issue that's going to damage your sons hearing.

Out.

Dave.

There is no suggestion it changes volume control behaviour but for information there is now a new version of Volumio to use:

www.minidsp.com/forum/announcements/17316-new-releases#48652

@drdyna is is completely correct, the Volumio/SHD volume control interaction is almost completely unusable so is total rubbish.
If it is any consolation to those who believe miniDSP can do no wrong, it's pretty much because they've modelled a logarithmic potentiometer too mathematically correctly

There is a common maxim, mentioned earlier in the thread, that a linear response is crap for audio so you use a logarithmic response. Unfortunately the dirty little secret of potentiometers is that although linear potentiometers are linear, logarithmic potentiometers are NOT actually logarithmic!

From here I will refer to potiometers as pots and logarithmic as log. The reason log pots are not actually log is that a pure log is almost as bad as linear. A linear pot has the top 6dB of gain in the top half of the pot travel and all the rest in the bottom half resulting in poor control resolution, the sweet spot is invariably in the bottom half of the travel where dB are crammed together tightly. A pure log pot shows exactly what @drdyna found, the bottom half of the pot travel is essentially inaudible and the sweet spot is always in the top half of the travel where dB are crammed together tightly. In both cases this cramming together of dB around the sweet spot leads to very poor control resolution.

There is one big problem with pure log, it never goes to true zero. A huge engineering compromise is required to balance going to true zero at the bottom end and having adequate resolution at the top end. In the SHD they use a rotary encoder which has equally fine resolution over the full 120dB gain range. That just does not map well to single turn rotary pot or fader.

So how do they actually build a so-called log pot? Out of as few as two linear segments which sort of approximate a pure log very crudely. They also have a small linear segment at the very lowest end of the pot to go to true zero. The dynamic range of control above this may be only 60dB, with the top 20dB having the bulk of the travel and the lower 40dB having far less travel. This naturally leads to a sweet spot around the breakpoint between the two linear segments with the travel per dB being much less below the breakpoint than above. Most of the time you are just above the breakpoint with fine resolution of dB. If you happen to need gain set below the breakpoint the resolution gets less but its much safer to have poor resolution at low levels.

Now the good new is that miniDSP should be able to introduce an indirection, a mapping of the 0-100 Volumio value to a warped range of dB. It shouldn't require any changes to Volumio. Then again, I'm not exactly sure what hoops they had to jump through to get to this stage.
I'm not sure whether faking linear segments would be state of the art, more likely a continuous function might be possible to smootly map this dB warping. It's also the case that miniDSP probably already know all this and are working on a solution.

There is one downside so it's not all win-win. On a power amplifier this faked log pot works fine if you choose your amplifier power appropriately. With SHD there could be a hugely differently range of amplifier gains which could mean everybodies sweet spot is comewhere different. My amplifiers have a gain of 29.2dB and no volume control so the concern of Volumio's poor resolution leading to extreme volumes is also real for me, though perhaps not quite to the degree it could be for @drdyna. I tend to run my SHD at -20dB gain normally, -15dB for quiet sources and -25dB to -30dB for loud sources. I suspect many folk have a similarly small range of the 120dB available from SHD that they use. So hopefully a compromise can be found that tends towards optimum for most folk.

For context I'll say my power amp has a stated gain of 29 dB but I use it in conjunction with 10 dB Rothwell inline attenuators. I started using the attenuators with a 2x4 HD where I had concerns over the use of the digital volume control to apply too much attenuation compromising quality, although having got them I realised they also provided a form of protection from accidentally going too loud. This was the first time I'd used a digital pre-amp connected directly to a power amp so I was twitchy about this! With my SHD I'm actually now contemplating removing the attenuators and applying the same effect within the plugin software, in part with Dirac headroom in mind.

I suspect speaker sensitivity is a factor here as well as amplifier gain.

Yes, speaker sensitivity is a factor too, and even the size of the room can affect audible volume. 91dB at 1m with 2.83V drive here.
I suppose the point is that the SHD Volume is almost completely unrelated to SPL which makes a general sweet spot awkward to find.

I too have eventually switched to using SHD directly into a power amp. For anyone who has excess gain after the SHD which could potentially cause damage to hearing or equipment, the output channel gain in the SHD plugin may be useful although it provides limited attenuation. It may just be enough to protect things though. Inline attenuators are wise if the internal attenuation is insufficient.

I suppose I would go further and suggest that anyone who can't run their system up to 0dB on SHD safely do have excess gain and should arguably introduce output attenuation until they can safely run at 0dB. I suspect I have excess gain but since I still run an old Volumio where the volume control is stuck at 100% I am not so susceptible to issues. I know I can run safely up to about -10dB so another 10dB isn't likely to do extreme damage and the SHD plugin outout attenuation should be sufficient for me. My amplifier sensitivity is significantly less than 1V so the 2V drive from SHD is technically too much. I

entripy wrote:

Yes, speaker sensitivity is a factor too, and even the size of the room can affect audible volume. 91dB at 1m with 2.83V drive here.
I suppose the point is that the SHD Volume is almost completely unrelated to SPL which makes a general sweet spot awkward to find.

I too have eventually switched to using SHD directly into a power amp. For anyone who has excess gain after the SHD which could potentially cause damage to hearing or equipment, the output channel gain in the SHD plugin may be useful although it provides limited attenuation. It may just be enough to protect things though. Inline attenuators are wise if the internal attenuation is insufficient.

I suppose I would go further and suggest that anyone who can't run their system up to 0dB on SHD safely do have excess gain and should arguably introduce output attenuation until they can safely run at 0dB. I suspect I have excess gain but since I still run an old Volumio where the volume control is stuck at 100% I am not so susceptible to issues. I know I can run safely up to about -10dB so another 10dB isn't likely to do extreme damage and the SHD plugin outout attenuation should be sufficient for me. My amplifier sensitivity is significantly less than 1V so the 2V drive from SHD is technically too much. I

Realistically I think pretty much all modern systems have too much gain, or arguably typical line out voltages of digital sources (2V or 4V depending is single-ended or balanced) are too high. I think you'd have to try quite hard to find someone with a conventional pre-amp that was doing anything other than attenuating such signals.

FWIW my main speakers aren't super-sensitive at 88 dB at 1m for 2.83V.

Checking the plugin it seems that output channel attenuation of up to 72dB can be applied though so I can't see this not being enough for anyone?

Ultrasonic wrote:

Checking the plugin it seems that output channel attenuation of up to 72dB can be applied though so I can't see this not being enough for anyone?

Oh wow, yes, plenty indeed, for some reason I thought it was only 12dB. Perhaps that's available gain. I haven't run the plugin for ages and its several versions out of date.

There's one decent solution for safety then, dial output channel attenuation down until you can safely run SHD at 0dB.

It won't help the insane Volumio level control sensitivity, but it will prevent any associated damage.

"There's one decent solution for safety then, dial output channel attenuation down until you can safely run SHD at 0dB."

Which is what I mentioned previously when talking about possible "hard limiting" actions. C'mon!

The control law of this "potentiometer" is not his primary issue. His complaint is that it's too easy to go to maximum.
miniDSP could alter the control law to whatever anybody would desire, but it wouldn't solve his complaint.

I think DyDyna is being sort of obtuse here, but at least I understand his issue with the operation.

Dave.

Sorry Dave, it's been a long thread which I got to late, so I was just summarising my conclusion, not claiming it was first mention.

entripy wrote:

Sorry Dave, it's been a long thread which I got to late, so I was just summarising my conclusion, not claiming it was first mention.

No problem. I just wish everybody would read/understand everybody else's posts.

My suggestion to drdyna was simply to (while waiting for devteam to possibly address this situation) implement some sort of positive action to limit the accidental maximum volume he was worried about. What I suggested will certainly work.....in the interim.

But, I'm greeted with a juvenile response "Limiters are not the answer to broken software."
So, at that point, there's really not much point in continuing with the guy.

Anyways.....
Carry on.

Dave.