Room correction

[Yoku-San]

Dear @WiiM Team,

can you share anything about the room correction feature you are supposedly working on?

Thanks for information in advance!

 

[Brantome]

Open breakthroughs in walls differ vastly from walls. That's in conformance with acoustics, physics and math, I reckon.

So no, I can't really share that point of view.

Linn’s “space optimisation” room correction/ acoustic modelling methodology requires, inter alia, that you “accurately measure your room dimensions, including all room features, and correctly identify the construction material of each surface, and enter these into the Space Optimisation user interface.”. You can also fine tune the absorption characteristics of each surface’s material should you wish. I’ve never done so as I’d doubt I’d notice the impact, but that fact that their model includes such variables would suggest that they do matter.

 

[harkpabst]

I bet they can.

But I'd rather like to see the room correction system measuring in different spots I the room what is really happening acoustically, instead of relying on users inputting absorption coefficients.

Sorry if I sound like a broken record here, but once again doing it the way Lyngdorf does it looks like the smartest (and relatively fool proof) way to me.

 

[Brantome]

I bet they can.

But I'd rather like to see the room correction system measuring in different spots I the room what is really happening acoustically, instead of relying on users inputting absorption coefficients.

The user doesn’t enter absorption coefficients as such but rather chooses the material, after which they can use a slider to increase or decrease its absorption.

Sorry if I sound like a vrikwn record here, but once again doing it the way Lyngdorf does it looks like the smartest (and relatively fool proof) way to me.

I’ll need to look that up - got a link?

 

[Steve Woodhouse]

Open breakthroughs in walls differ vastly from walls. That's in conformance with acoustics, physics and math, I reckon.

So no, I can't really share that point of view.

Then it’s not symmetrical, which is the point of view I was expressing, which makes this a Strawman argument.

 

[Steve Woodhouse]

 

[harkpabst]

Then it’s not symmetrical, which is the point of view I was expressing, which makes this a Strawman argument.

Sorry about that.

May I add that your “point of view“ was based on a red herring then, because nobody, neither @amix nor @Haskil nor myself have been talking about imbalances due to furniture and surfaces. We don't have the same definition of symmetry, at least.

Even if we accept that new direction of argument, posting blurry images without any context doesn't help your case. Contrary to common belief numbers do not speak for themselves.

 

[harkpabst]

I’ll need to look that up - got a link?

Not right now on my phone. For a primer just search for RoomPerfect on YouTube. There are a number of videos from Lyngdorf. They also provide batbleast one reasonably good white paper that's more in-depth than the marketing materials.

 

[Steve Woodhouse]

Sorry about that.

May I add that your “point of view“ was based on a red herring then, because nobody, neither @amix nor @Haskil nor myself have been talking about imbalances due to furniture and surfaces. We don't have the same definition of symmetry, at least.

Even if we accept that new direction of argument, posting blurry images without any context doesn't help your case. Contrary to common belief numbers do not speak for themselves.

My initial post was:

I appreciate that, even in a perfectly symmetrical room/set up, you can get imbalance due to furnishings and surfaces. But, contrary to popular belief, windows, doors and walls don’t have significantly different reflection characteristics.

And the imbalances, in a reasonably symmetrically room/set up, won’t be hugely different.

You canna change the laws of physics. Or maths.

So I think I covered all that.

Important thing to remember here, back to basics. The most important sound is direct from the speaker. Second most important is first reflections. After that you’re getting diminishing returns.

So, if you have a sofa in between one of your speakers and your seating position, but not the other speaker, this may have an impact. I don’t know of anyone with a room like that.

If you have clear lines for the path of direct sound, and the sound of first reflections, then whilst there may be differences, they’re likely to be small.

I fully appreciate some may want 200 band stereo eq and to iron out every last peak and dip, or stereo imbalance caused by the waste paper basket on the left of the room, but it’s not for me. And it’s clearly not for most.

Look, I fully appreciate what you want to do and what you’d like this to be. And I hope you get what you want. 10 band PEQ? Bring it on! Excellent stuff.

But let’s remember, even 4 bands of PEQ in the Amp, Pro and Pro Plus, never mind the £90 Mini, is revolutionary for both the price, and the type of device(s) we’re discussing. If we get more, that’s great. I’m just saying, the WiiMs don’t become some sort of failure without them.

Each to their own. Best wishes. And once again, I hope you get what you want from future PEQ and room correction updates.

 

[Brantome]

Not right now on my phone. For a primer just search for RoomPerfect on YouTube. There are a number of videos from Lyngdorf. They also provide batbleast one reasonably good white paper that's more in-depth than the marketing materials.

From a quick look it looks like it shares similarities but also differences with the approach Linn take with their Space Optimisation process, and like everything else in hi fi there are those who prefer one method over the other

 

[harkpabst]

I fully appreciate some may want 200 band stereo eq and to iron out every last peak and dip, or stereo imbalance caused by the waste paper basket on the left of the room, but it’s not for me. And it’s clearly not for most.

I think, nobody has been asking for 200 band PEQ, so far. In fact, it's been WiiM themselves rising the bar to 10 band instead of the originally announced 8 band PEQ. No wonder people are asking how to make use of that, right?

But let’s remember, even 4 bands of PEQ in the Amp, Pro and Pro Plus, never mind the £90 Mini, is revolutionary for both the price, and the type of device(s) we’re discussing. If we get more, that’s great. I’m just saying, the WiiMs don’t become some sort of failure without them

I can wholeheartedly agree with that.

The ugly truth that few are talking out loud is that "traditional" room correction like it has been established mainly in the HT world (let's just EQ everything to our home curve and it will sound the same and great) simply doesn't work (and doesn't remove the influence of the room, luckily, that's not even necessary).

From a quick look it looks like it shares similarities but also differences with the approach Linn take with their Space Optimisation process, and like everything else in hi fi there are those who prefer one method over the other

I can't even compare the two. Since I only know RoomPerfect in practice (and to a limited extent the theory behind it). One of the basic ideas is behind it is that you did pick your speakers for a reason, not to have it EQed like there's no tomorrow. If you want to minimize the effects of the listening room, you need to measure the room, not just what happens at the main listening position.

 

[Steve Woodhouse]

I think, nobod has been asking for 200 band PEQ, so far. In fact, it's been WiiM themselves rising the bar to 10 band instead of the originally announced 8 band PEQ. No wonder people are asking how to make use of that, right?


I can wholeheartedly agree with that.

The ugly truth that few are talking out loud is that "traditional" room correction like it has been established mainly in the HT world (let's just EQ everything to our home curve and it will sound the same and great) simply doesn't work (and doesn't remove the influence of the room, luckily, that's not even necessary).


I can't even compare the two. Since I only know RoomPerfect in practice (and to a limited extent the theory behind it). One of the basic ideas is behind it is that you did pick your speakers for a reason, not to have it EQed like there's no tomorrow. If you want to minimize the effects of the listening room, you need to measure the room, not just what happens at the main listening position.

For many of us, and this is becoming more and more I suspect, there’s not a main listening position, there’s an only listening position.

Just look at the threads where people show pics of their set ups, single-seat desktops dominate. Form factor, desktop devices dominate, whether it’s Topping DACs, Fosi Audio amps, or WiiM streamers, these are rarely ‘full size’.

As long as the sound at my listening position (chair pushed back a little from my desk) sounds great, I really don’t care if elsewhere in the room it sounds like an empty swimming pool.

It makes positioning the sub easy (go where it’s loud, measure it, bring it down with PEQ). Room treatment, little is needed for nearfield.

Great sound for a small about of cash. What’s not to like.

 

[goodmore]

Easy to use and effective at no additional cost. It will be interesting.

 

[canard]

"auto" correction "General public" is a very hard sport, even for big brands it will break their teeth....
on the other hand, good headphone processing via an "enriched" peq and all the curves available, it will be effective.... otherwise d affordable and very useful can be a really deep sub mode perfected and a peq, see the "auto" correction of this precise register... that's reasonable... (like the modest dsp-lf way of dayton and "iwoofer pro")
"dsp" "peq" "bass", that works well and easily..
;-)

 

[Melomane]

I haven't read the whole thread. My recommendation until WiiM has fulfilled your wish: Install a DLNA server on the NAS and Convoproxy. Convoproxy can do room correction. However, I have not yet tried it out. Look here:

ConvoProxy — ConvoProxy documentation

 

[amix]

neither @amix nor @Haskil nor myself have been talking about imbalances due to furniture and surfaces.

Hmmm... English isn't my native language, so I may have been misunderstood. When I mentioned asymmetry in the room I was thinking about everything: assymetric walls, windows, furniture, whatever. So anything, that is in the way of the air coming from any speaker.

 

[harkpabst]

For many of us, and this is becoming more and more I suspect, there’s not a main listening position, there’s an only listening position.

Just look at the threads where people show pics of their set ups, single-seat desktops dominate. Form factor, desktop devices dominate, whether it’s Topping DACs, Fosi Audio amps, or WiiM streamers, these are rarely ‘full size’.

As long as the sound at my listening position (chair pushed back a little from my desk) sounds great, I really don’t care if elsewhere in the room it sounds like an empty swimming pool.

It makes positioning the sub easy (go where it’s loud, measure it, bring it down with PEQ). Room treatment, little is needed for nearfield.

Great sound for a small about of cash. What’s not to like.

Now I see where you're coming from. Fair point, I've seen all sorts of videos about desktop nearfield listening and in this scenario there's really little to worry about the room. Get your sub integrated smoothly, tame your speakers as you like and you're done.

But really and truly this scenario does not apply to me nor anybody I personally know. I tend to think that this is a very male kind of consuming music, and even with the males I know (myself included) it's not popular. So, for more traditional listeners like myself, requirements are really different. As we already agreed on: each to their own.

I haven't read the whole thread. My recommendation until WiiM has fulfilled your wish: Install a DLNA server on the NAS and Convoproxy. Convoproxy can do room correction. However, I have not yet tried it out. Look here:

ConvoProxy — ConvoProxy documentation

Personally I wouldn't recommend anything I haven't tried myself. Can a proxy be used to correct e.g. the signal fed through the analog RCA input? I don't think so. Most importantly, the WiiM Amp already comes with a pretty flexible 4 band PEQ. If you own a reasonably priced mic like the miniDSP UMIK-1 and know how to use REW, there's already a lot you can do. iOS users have been using the HomeCurve app (with varying success).

Hmmm... English isn't my native language, so I may have been misunderstood. When I mentioned asymmetry in the room I was thinking about everything: assymetric walls, windows, furniture, whatever. So anything, that is in the way of the air coming from any speaker.

Sorry, I didn't mean to cite you incorrectly. Generally I think we are after the same thing.

 

[Inertiaman]

Having 10 identical PEQs on the two channels is much less useful than having 5 per channel...

For clarity, there isn't a resource equivalence between 10 identical L/R PEQs and 5 L / different 5 R PEQs. (to me this seemed implied by the statement above, but perhaps not intended by @Haskil).

From a processor resource perspective (memory, CPU cycles) for realtime filter processing, there is no meaningful difference between implementing all the same PEQ settings on both channels versus all different PEQ settings on each channel. If Wiim can successfully implement a stereo 10-band PEQ, then the capacity for independent per-channel PEQ settings is really only "limited" by Wiim's decision to provide (or not) per-channel setting control in the App user interface.

If this isn't obvious, consider the following: A stereo 10-band PEQ needs to "mathematically" implement 20 independent PEQ filters. Since the L and R channels are independent data streams, you can't push them both through one PEQ filter. A designer/engineer may *choose* to provide the settings access only on a both-channel basis, perhaps out of user convenience for "typical" use-cases. But underneath those settings, there are still two independent PEQ filters running, one on each channel. The specific Q, f, and gain settings on each channel have no significant impact on the memory and CPU requirements.

There may be some minor advantages / resource efficiencies to keeping the settings identical on L/R, depending on how the s/w architecture is pipelined, how buffers are structured/accessed, etc. Probably more importantly, that architecture, if originally conceived for same L/R settings, could create a "hassle factor" for transitioning to independent L/R settings of PEQ (ie, you might be creating do-over work for the SW engineer(s) ). But this is essentially high level configuration SW, not the actual filters. From a realtime execution resource perspective on the filters themselves, the hardware should not, practically speaking, be a constraint to independent L/R settings.

So consider this a vote from me for Wiim to provide a "switch" in the app for setting the PEQ independently on L/R (for the geeks and tweekers) OR as it is now: same on L/R (which is certainly easier for casual users).

 

[Smartplug]

It seems long time wait to get this out on beta or public release.

 

[Brantome]

It seems long time wait to get this out on beta or public release.

Didn’t their last update say they hoped to get the beta out around the 15th March? Let’s see if that happens at the end of the week.

 

[Steve Woodhouse]

For clarity, there isn't a resource equivalence between 10 identical L/R PEQs and 5 L / different 5 R PEQs. (to me this seemed implied by the statement above, but perhaps not intended by @Haskil).

From a processor resource perspective (memory, CPU cycles) for realtime filter processing, there is no meaningful difference between implementing all the same PEQ settings on both channels versus all different PEQ settings on each channel. If Wiim can successfully implement a stereo 10-band PEQ, then the capacity for independent per-channel PEQ settings is really only "limited" by Wiim's decision to provide (or not) per-channel setting control in the App user interface.

If this isn't obvious, consider the following: A stereo 10-band PEQ needs to "mathematically" implement 20 independent PEQ filters. Since the L and R channels are independent data streams, you can't push them both through one PEQ filter. A designer/engineer may *choose* to provide the settings access only on a both-channel basis, perhaps out of user convenience for "typical" use-cases. But underneath those settings, there are still two independent PEQ filters running, one on each channel. The specific Q, f, and gain settings on each channel have no significant impact on the memory and CPU requirements.

There may be some minor advantages / resource efficiencies to keeping the settings identical on L/R, depending on how the s/w architecture is pipelined, how buffers are structured/accessed, etc. Probably more importantly, that architecture, if originally conceived for same L/R settings, could create a "hassle factor" for transitioning to independent L/R settings of PEQ (ie, you might be creating do-over work for the SW engineer(s) ). But this is essentially high level configuration SW, not the actual filters. From a realtime execution resource perspective on the filters themselves, the hardware should not, practically speaking, be a constraint to independent L/R settings.

So consider this a vote from me for Wiim to provide a "switch" in the app for setting the PEQ independently on L/R (for the geeks and tweekers) OR as it is now: same on L/R (which is certainly easier for casual users).

WORD!

Actually, for clarity, are you saying 5 band stereo is easy as 10 band mono, or as easy as 5 band mono?