Room Correction Beta firmware for WiiM Pro, Pro Plus and Amp - User Testing Experience

[Brantome]

Can someone usinf beta tell me how it initially handled any previous PEQ settings you had? I currently have a very good curve using the current four available, when the six additional are added, does it retain those previous saved settings, or is it a start from scratch type update? I ask because in the past WiiM has been good about retaining settings, but they need to be because they typically add these new features without recourse. I may want to continue to tweak, but not fully re-plow already good plowed ground. Thanks.

I did have an issue where a beta firmware update deleted my saved PEQ presets but hopefully that'll be resolved by the time it's a public release

 

[Wiimer]

Can someone usinf beta tell me how it initially handled any previous PEQ settings you had? I currently have a very good curve using the current four available, when the six additional are added, does it retain those previous saved settings, or is it a start from scratch type update? I ask because in the past WiiM has been good about retaining settings, but they need to be because they typically add these new features without recourse. I may want to continue to tweak, but not fully re-plow already good plowed ground. Thanks.

Saved custom PEQs can disappear at any time, so it's a good idea to take a screenshot of them.

 

[slartibartfast]

How big is your room and how far do you sit from your speakers?

Only about 5.2m x 4m x 2.4m and I sit about 2m from the speakers.

 

[Steve Woodhouse]

I agree to this approach during the beta development. The final product should do away with that, of course.

Generally speaking I prefer to rather feed a little too much than too little bass to the RC algorithm (for various reasons).


The current non-beta 10 band PEQ feature of the WiiM Amp still allows +/- 12 dB adjustments. See screenshot:

View attachment 6920

Mainly based in the works of Floyd E. Toole and Sean Olive, it has been found that a flat frequency response measured in a typical listening room is not perceived as natural and preferable. At the same time, a speaker that measures flat in an anechoic chamber will usually show the preferred response (slightly elevated bass and a slight downwards slope to the absolute highs), anyway.

PS:
This was "Toole in a quarter nutshell". Better look up his work incase you are interested.

More than aware already, thanks.

And I’d still try to correct ‘naturally’ as much as possible first, so as to minimise the use of EQ.

And yes, I’d rather be pulling down peaks rather than filling in dips with EQ, especially as the latter may simply not work if it’s a room cancellation node.

 

[slartibartfast]

I agree to this approach during the beta development. The final product should do away with that, of course.

Generally speaking I prefer to rather feed a little too much than too little bass to the RC algorithm (for various reasons).


The current non-beta 10 band PEQ feature of the WiiM Amp still allows +/- 12 dB adjustments. See screenshot:

View attachment 6920

Mainly based in the works of Floyd E. Toole and Sean Olive, it has been found that a flat frequency response measured in a typical listening room is not perceived as natural and preferable. At the same time, a speaker that measures flat in an anechoic chamber will usually show the preferred response (slightly elevated bass and a slight downwards slope to the absolute highs), anyway.

PS:
This was "Toole in a quarter nutshell". Better look up his work incase you are interested.

Are 10 PEQ filters available in the standard non-beta firmware?

 

[Steve Woodhouse]

Only about 5.2m x 4m x 2.4m and I sit about 2m from the speakers.

That’s fairly medium sized.

Have you tried the Harman Curve?

 

[Brantome]

Are 10 PEQ filters available in the standard non-beta firmware?

Not yet afaik

 

[Smartplug]

It’s still long winding road before RC becomes where it should be.

 

[slartibartfast]

That’s fairly medium sized.

Have you tried the Harman Curve?

No not yet. I'll have to figure out how to enter house curves into REW.

 

[Trailer]

Can someone usinf beta tell me how it initially handled any previous PEQ settings you had? I currently have a very good curve using the current four available, when the six additional are added, does it retain those previous saved settings, or is it a start from scratch type update? I ask because in the past WiiM has been good about retaining settings, but they need to be because they typically add these new features without recourse. I may want to continue to tweak, but not fully re-plow already good plowed ground. Thanks.

I received the beta firmware a few days ago and I lost my previous PEQ settings used in the official firmware - which were just the four.

 

[Steve Woodhouse]

It’s still long winding road before RC becomes where it should be.

How long and winding?

It’s only been out for around 5 weeks in beta.

And you’d have to think they’ll want to release the Ultra with everything working right from the start, rather than ‘coming soon’.

 

[harkpabst]

Are 10 PEQ filters available in the standard non-beta firmware?

Yes. I'm not on the beta program and I do have 10 band PEQ for my Amp now.

 

[harkpabst]

This is a stack of two measurements that @Trailer posted here.

View attachment 6922

Room Correction Beta firmware for WiiM Pro, Pro Plus and Amp - User Testing Experience

…and this is exactly what beta is for.

forum.wiimhome.com

Red line = new algorithm
White line = old algorithm (after rollback)

The two measurements are almost identical in peak and dip frequencies, and the only difference is the volume per frequency. What does this mean?

To me, there is nothing wrong with the iphone microphone, and it looks like the WiiM is not properly handling the data it receives from the microphone. Also, based on the 1kHz area, it seems that the error occurs as you go outward.

By the way, this is just curious, what kind of test sweep does a WiiM device play? Is it the same for all beta firmwares? What are the results of measuring the WiiM test sweep with other apps?

Maybe I didn't get the concept (because I'm not on the beta program), but I always understood the graphs like this:

Measured (grey): Initial measurement of transfer function (frequency response) prior to applying any room correction.

Equalization (magenta): Summary of all individual correction filters to apply.

Predicted (cyan): Multiplication of transfer functions Measured and Equalization.

In this posting by @Trailer ...

Room Correction Beta firmware for WiiM Pro, Pro Plus and Amp - User Testing Experience

…and this is exactly what beta is for.

forum.wiimhome.com

... the "Measured" transfer functions look totally off and they are totally different. I'm not talking about the "Predicted" transfer function, which is just calculated.

If these assumptions are wrong, someone please enlighten me.

 

[Trailer]

I’m not sure if this is the way things happen with the RC or not, and I can’t get to test this just now, but turning on the RC turns on the EQ. If you turn off the EQ it turns off the RC. So, when you do the RC the EQ is on. Do the sweeps then use the EQ that is in place or does it bypass the EQ? If they go through the EQ then you will have to have a flat EQ selected prior to getting a valid measurement.

 

[slartibartfast]

Yes. I'm not on the beta program and I do have 10 band PEQ for my Amp now.

I only have 4 with the standard firmware. When did yours change?

 

[Wiimer]

Maybe I didn't get the concept (because I'm not on the beta program), but I always understood the graphs like this:

Measured (grey): Initial measurement of transfer function (frequency response) prior to applying any room correction.

Equalization (magenta): Summary of all individual correction filters to apply.

Predicted (cyan): Multiplication of transfer functions Measured and Equalization.

In this posting by @Trailer ...

Room Correction Beta firmware for WiiM Pro, Pro Plus and Amp - User Testing Experience

…and this is exactly what beta is for.

forum.wiimhome.com

... the "Measured" transfer functions look totally off and they are totally different. I'm not talking about the "Predicted" transfer function, which is just calculated.

If these assumptions are wrong, someone please enlighten me.

I haven't participated in RC tests either, so maybe I'm just missing the point.


Yes, I think your perception is the same as mine.
Grey = Mic measurement result
Magenta = Applied EQ curve
Cyan = corrected curve

And I compared the gray curves.

 

[onlyoneme]

... the "Measured" transfer functions look totally off and they are totally different. I'm not talking about the "Predicted" transfer function, which is just calculated.

If the "predicted" graph after the correction process, and the "measured" graph after EQ changes are applied, are totally different, then the correction process is definitely broken. For sure

 

[harkpabst]

If the "predicted" graph after the correction process, and the "measured" graph after EQ changes are applied, are totally different, then the correction process is definitely broken. For sure

Thing is, I always thought that there is no measured graph after EQ ...

But yes, if it's Measured (after EQ) vs. Predicted, it's very easy.

 

[Giro413]

To be honest the corrected wiim response looks logical but again it should not correct so aggressively in order for the curve to lie exactly on the downward line. On the other hand the measured is not looking good at all. I don't know if it is your system-speakers and/or the placement or even the measurement of wiim. I have noticed that wiim makes the measurment at higher amplitude than housecurve. You have to search for the predicted in room response of your speakers and compare it to your measurement. If it is vastly different it is your space or/and wrong measurement if it matches then it is your speakers... but you should compare at the same scale.

My measurements with the sub off. Different picture, no? Not sure what that tells me, though. Will need to dig out a longer cable if I am to experiment with sub placement.

 

[Steve Woodhouse]

No not yet. I'll have to figure out how to enter house curves into REW.

You can select it on the WiiM.