Room Correction - Confused by Result

[mikerick39]

I have a WiiM Amp Pro connected to two bookshelf speakers and two subs. I’m redoing room correction for the first time since doing it initially after buying the Amp Pro in September.

The bookshelves are just under six feet apart on top of a vintage stereo console and about 18 inches in front of large windows (see pics). The subs are hidden inside the stereo console directly beneath each bookshelf.

My listening position is about 10 feet away at the end of a wooden table typically covered in miscellaneous stuff. To my left is a large bank of floor to ceiling windows and to my right is open to my living room, and behind me is open to the kitchen area that is ten feet back. My floor is hardwood with a large rug under the table. Walls are plaster, and ceiling height is ten feet (see pics).

The problem:

The initial room correction measurements for both Left and Right above 1kHz look pretty good (see pic). However after WiiM’s equalization is applied (see pics) the measured result shows large dips between 4-7kHz and just over 16kHz (see pics).

Those large dips don’t correlate at all with the Eq applied by RC so where did they come from???

This problem is occurring when doing either L/R individual correction or the stereo correction. In fact it gets much worse when doing the Stereo correction.

My equipment:

- Calibrated UMIK-1 (horizontal, at seating position and height, centered between bookshelves and pointing straight ahead). Cal file uploaded into WiiM Home App.

- iPhone 16 Pro Max

- WiiM Amp Pro with latest 5.2.717556 firmware update

- Audio First Fidelia Bookshelf speakers (reference level anechoic frequency response and dispersion)

- Two Dayton Audio DIY 8” Ultimax 2 ported subs powered by one Aiyima A70 Stereo Amplifier

Anyone experienced in using WiiM’s room correction have any ideas or suggestions why this is happening and what may fix this? Thanks in advance for whatever advice you may have.

Note…I am only interested in help with the WiiM RC itself. I fully understand room treatments will help overall (not possible due to WAF), and I also understand that RC is often not recommended from around 300-500Hz and below due to bass nodes and nulls being uncorrectable via Eq.

 

[johim]

Hey Mike, interesting setup, and thanks for laying it out so clearly. I’ve seen similar anomalies with RC results when there are reflective surfaces (like large windows) and asymmetrical rooms. Those dips in 4–7 kHz and above 16 kHz might be artifacts from reflections or mic positioning inconsistencies, even with a UMIK-1. Have you tried running the sweep multiple times to see if the dips stay consistent?

 

[canard]

Hey Mike, interesting setup, and thanks for laying it out so clearly. I’ve seen similar anomalies with RC results when there are reflective surfaces (like large windows) and asymmetrical rooms. Those dips in 4–7 kHz and above 16 kHz might be artifacts from reflections or mic positioning inconsistencies, even with a UMIK-1. Have you tried running the sweep multiple times to see if the dips stay consistent?

Fréquence et longueur d'onde

Convertissez fréquence et longueur d'onde

www.rt60.net

;-)

 

[mikerick39]

Hey Mike, interesting setup, and thanks for laying it out so clearly. I’ve seen similar anomalies with RC results when there are reflective surfaces (like large windows) and asymmetrical rooms. Those dips in 4–7 kHz and above 16 kHz might be artifacts from reflections or mic positioning inconsistencies, even with a UMIK-1. Have you tried running the sweep multiple times to see if the dips stay consistent?

Johim:

Thanks for your response. Yes I get the same results or very similar every time. I’m really confused as to why the major dips in the upper frequency ranges I’m seeing post RC Eq aren’t showing up before the application of RC EQ.

 

[mikerick39]

Fréquence et longueur d'onde

Convertissez fréquence et longueur d'onde

www.rt60.net

;-)

Umm…merci?

 

[jiaxin.li (WiiM)]

I have a WiiM Amp Pro connected to two bookshelf speakers and two subs. I’m redoing room correction for the first time since doing it initially after buying the Amp Pro in September.

The bookshelves are just under six feet apart on top of a vintage stereo console and about 18 inches in front of large windows (see pics). The subs are hidden inside the stereo console directly beneath each bookshelf.

My listening position is about 10 feet away at the end of a wooden table typically covered in miscellaneous stuff. To my left is a large bank of floor to ceiling windows and to my right is open to my living room, and behind me is open to the kitchen area that is ten feet back. My floor is hardwood with a large rug under the table. Walls are plaster, and ceiling height is ten feet (see pics).

The problem:

The initial room correction measurements for both Left and Right above 1kHz look pretty good (see pic). However after WiiM’s equalization is applied (see pics) the measured result shows large dips between 4-7kHz and just over 16kHz (see pics).

Those large dips don’t correlate at all with the Eq applied by RC so where did they come from???

This problem is occurring when doing either L/R individual correction or the stereo correction. In fact it gets much worse when doing the Stereo correction.

My equipment:

- Calibrated UMIK-1 (horizontal, at seating position and height, centered between bookshelves and pointing straight ahead). Cal file uploaded into WiiM Home App.

- iPhone 16 Pro Max

- WiiM Amp Pro with latest 5.2.717556 firmware update

- Audio First Fidelia Bookshelf speakers (reference level anechoic frequency response and dispersion)

- Two Dayton Audio DIY 8” Ultimax 2 ported subs powered by one Aiyima A70 Stereo Amplifier

Anyone experienced in using WiiM’s room correction have any ideas or suggestions why this is happening and what may fix this? Thanks in advance for whatever advice you may have.

Note…I am only interested in help with the WiiM RC itself. I fully understand room treatments will help overall (not possible due to WAF), and I also understand that RC is often not recommended from around 300-500Hz and below due to bass nodes and nulls being uncorrectable via Eq.

The noticeable dips around 4–7 kHz and 16 kHz are not caused by the EQ.
Instead, when playing in stereo, the sounds from the left and right speakers reach your ears via different paths,
which leads to phase interference at certain frequencies, resulting in what is called “comb filtering.”

Your room’s structure worsens this issue: with glass windows on the left and an open space on the right,
this acoustic asymmetry makes high-frequency reflections more complex and interference more pronounced.

During single-channel measurement and correction, the system measures the left and right channels separately, so “comb filtering” is not obvious;
however, during actual validation or listening, the two sound sources combine and can cause interference at certain frequencies, appearing as sharp dips.

We suggest you try:
Slightly adjusting the toe-in angle of the speakers;
Or adding curtains on the glass window side to see if the dips improve.

 

[canard]

out of curiosity...
How is your microphone held?
nearby surfaces?
Have you tried it at 90 degrees?
;-)

 

[mikerick39]

out of curiosity...
How is your microphone held?
nearby surfaces?
Have you tried it at 90 degrees?
;-)

I used the stock UMIK-1 mic stand. Surfaces are just like in the pictures and did not change between initial

The noticeable dips around 4–7 kHz and 16 kHz are not caused by the EQ.
Instead, when playing in stereo, the sounds from the left and right speakers reach your ears via different paths,
which leads to phase interference at certain frequencies, resulting in what is called “comb filtering.”

Your room’s structure worsens this issue: with glass windows on the left and an open space on the right,
this acoustic asymmetry makes high-frequency reflections more complex and interference more pronounced.

During single-channel measurement and correction, the system measures the left and right channels separately, so “comb filtering” is not obvious;
however, during actual validation or listening, the two sound sources combine and can cause interference at certain frequencies, appearing as sharp dips.

We suggest you try:
Slightly adjusting the toe-in angle of the speakers;
Or adding curtains on the glass window side to see if the dips improve.

Thanks for the information. I’ll try adjusting the speaker angles and see what happens.

 

[mikerick39]

out of curiosity...
How is your microphone held?
nearby surfaces?
Have you tried it at 90 degrees?
;-)

I used a standard floor mic stand with mic located where my head typically is. Nearby surfaces are as in the pictures. I will try it at 90 and see how that goes. Thanks.

 

[Alexandr_2023]

I have a WiiM Amp Pro connected to two bookshelf speakers and two subs. I’m redoing room correction for the first time since doing it initially after buying the Amp Pro in September.

The bookshelves are just under six feet apart on top of a vintage stereo console and about 18 inches in front of large windows (see pics). The subs are hidden inside the stereo console directly beneath each bookshelf.

My listening position is about 10 feet away at the end of a wooden table typically covered in miscellaneous stuff. To my left is a large bank of floor to ceiling windows and to my right is open to my living room, and behind me is open to the kitchen area that is ten feet back. My floor is hardwood with a large rug under the table. Walls are plaster, and ceiling height is ten feet (see pics).

The problem:

The initial room correction measurements for both Left and Right above 1kHz look pretty good (see pic). However after WiiM’s equalization is applied (see pics) the measured result shows large dips between 4-7kHz and just over 16kHz (see pics).

Those large dips don’t correlate at all with the Eq applied by RC so where did they come from???

This problem is occurring when doing either L/R individual correction or the stereo correction. In fact it gets much worse when doing the Stereo correction.

My equipment:

- Calibrated UMIK-1 (horizontal, at seating position and height, centered between bookshelves and pointing straight ahead). Cal file uploaded into WiiM Home App.

- iPhone 16 Pro Max

- WiiM Amp Pro with latest 5.2.717556 firmware update

- Audio First Fidelia Bookshelf speakers (reference level anechoic frequency response and dispersion)

- Two Dayton Audio DIY 8” Ultimax 2 ported subs powered by one Aiyima A70 Stereo Amplifier

Anyone experienced in using WiiM’s room correction have any ideas or suggestions why this is happening and what may fix this? Thanks in advance for whatever advice you may have.

Note…I am only interested in help with the WiiM RC itself. I fully understand room treatments will help overall (not possible due to WAF), and I also understand that RC is often not recommended from around 300-500Hz and below due to bass nodes and nulls being uncorrectable via Eq.

It has been said a million times that only "Dirac Live" has a good correction. Everything else is just fooling around...

 

[Burnside]

It has been said a million times that only "Dirac Live" has a good correction.

By Dirac no doubt

Everything else is just fooling around...

I’m sure others using Lyngdorf’s Room Correction or Linn’s Space Optimisation - or even WiiM’s Roomfit - might disagree…

 

[Deleted member 18192]

By Dirac no doubt



I’m sure others using Lyngdorf’s Room Correction or Linn’s Space Optimisation - or even WiiM’s Roomfit - might disagree…

Can only write for me and here Space Optimisation in my main system is by far good enough. My room is not so complicated.
And the streamers in my office setup are simply not important enough to go to all the bit- and-pieces.

 

[harkpabst]

... and I also understand that RC is often not recommended from around 300-500Hz and below due to bass nodes and nulls being uncorrectable via Eq.

It' true that you cannot correct for nulls caused by room modes and cancellation, but usually the problem is still rather the other way around: RC makes most sense below 300-500 Hz. You seem to be correcting full range, which might or might not work well. If I were you I'd try to only correct up to 500 Hz, but the effect you see is mostly unrelated.

I don't see the point in doing a 90⁰ measurement since you have the cal file imported. Pointing to the ceiling will just result in lower than real SPL in the highs, which full range RC will try to correct for, so you end up with a treble boost.

This problem is occurring when doing either L/R individual correction or the stereo correction. In fact it gets much worse when doing the Stereo correction.

This confirms to me that @jiaxin.li (WiiM) is correct. What happens is not purely a measuring artifact, the acoustic phenomenon is real. However, it's a) not well represented by a single microphone compared to your two ears at a fixed distance and b) it's something you cannot correct for.

As mentioned above, I would limit RC to the frequency range clearly below the wavelength that equals the distance between your ears. Additionally, you may want to try the new beta feature that adds the "moving microphone method" to the standard swept sine measurement, as soon as it becomes available to you. If I remember correctly, it's currently available for the WiiM Home app on Android, only.

Playingwith toe-in is also a good idea. Ironically, a speaker with a more narrow directivity might have actually yielded better results with this kind of setup, but thanks to the very controlled DI you're at least free to experiment.

It might be worth investigating the small fee for the HouseCurve app to get a better feeling for what's happening in your room, independent of EQ.

 

[mikerick39]

It' true that you cannot correct for nulls caused by room modes and cancellation, but usually the problem is still rather the other way around: RC makes most sense below 300-500 Hz. Kyou seem to be correcting full range, which might or might not work well. If I were you I'd try to only correct up to 500 Hz, but the effect you see is mostly unrelated.

I don't see the point in doing a 90⁰ measurement since you have the cal file imported. Pointing to the ceiling will just result in lower than real SPL in the highs, which full range RC will try to correct for, so you end up with a treble boost.


This confirms to me that @jiaxin.li (WiiM) is correct. What happens is not purely a measuring artifact, the acoustic phenomenon is real. However, it's a) not well represented by a single microphone compared to your two ears at a fixed distance and b) it's something you cannot correct for.

As mentioned above, I would limit RC to the frequency range clearly below the wavelength that equals the distance between your ears. Additionally, you may want to try the new beta feature that adds the "moving microphone method" to the standard swept sine measurement, as soon as it becomes available to you. If I remember correctly, it's currently available for the WiiM Home app on Android, only.

Playingwith toe-in is also a good idea. Ironically, a speaker with a more narrow directivity might have actually yielded better results with this kind of setup, but thanks to the very controlled DI you're at least free to experiment.

It might be worth investigating the small fee for the HouseCurve app to get a better feeling for what's happening in your room, independent of EQ.

Thanks for the advice and great info. I will try the HouseCurve app and post a follow up.