Share your in-room measurements?

[TurtlePaul]

I don't know, but if your speakers cross over to the tweeters between 3K and 7K could just one tweeter be wired with the wrong polarity? Probably just the right one? As I said, I'm stumped, but might it be worth reversing the wires to the right tweeter only, just to see?

As far as audibility is concerned, I'd bet that, at that frequency, the dip would be audible on some music.

It is most likely microphone alignment. The reason summed left and right response is not used for high frequency is because the wavelengths get too short. 4 khz has a wavelength at of about 3.4 inches. If his microphone is a mere 1.7 inches closer to one speaker than the other then it will measure a textbook cancelation notch at 4 khz (as we can see).

Fortunately our head has two ears which are spaced apart so the sound waves reach our ears at different times and we get different nulls. There is a big round thing between our ears that shields each ear from cross cancelation from the opposite direction. All this means that instead of hearing nulls our brain is able to interpret this as soundstage and directional audio.

 

[garyrc]

It is most likely microphone alignment. The reason summed left and right response is not used for high frequency is because the wavelengths get too short. 4 khz has a wavelength at of about 3.4 inches. If his microphone is a mere 1.7 inches closer to one speaker than the other then it will measure a textbook cancelation notch at 4 khz (as we can see).

Fortunately our head has two ears which are spaced apart so the sound waves reach our ears at different times and we get different nulls. There is a big round thing between our ears that shields each ear from cross cancelation from the opposite direction. All this means that instead of hearing nulls our brain is able to interpret this as soundstage and directional audio.

Does this mean that brachycephalic listeners get a different impression of stereo geometry than dolichocephalic ones? Is there just the right speaker cable or mid- field chrystal focal object, or perhaps a softwood diffuser that will correct for that, making everyone hear like the presumably mesophallic sound engineers?

 

[HiMu]

Here are quick left+sub & right+sub only 2 measurements averaged from my diy treated 29m3/1035ft3 non-cuboid room with a sloped ceiling in a leaky old wooden house.

Genelec 8340 pair + 7360 sub, with listening distance of ~2.2m from speakers, ~3.2m from sub. Sub integration and placement height needs some attention to get rid of that 100hz dip, 450-500hz "middle finger" is probably back wall reflection, 1.3khz is from the absorber surface behind the listening position.

Room has 14 absorbers ranging from 10cm/4" to 40cm/16" deep that cover most of the back wall, side walls, a 3 panel ceiling cloud and 1 behind both speakers. Combined thickness of all the 120x60cm /4'x2' panels is absolutely ridiculous 3.6m/12'. Absorber materials are Isolina flax insulation, Hunton Nativo wood fiber and Rockwool, with gas flow resistivity of 4k, 5k and 8.8k, respectively.

I started to build this room 4-5 years ago mostly during winter months. Calculated RT60 for a room this size/volume is 0.17s and it's barely inside the ITU-R BS1116 tolerances, but being so small, those figures are not as useful. Originally this was a synth noodling room /home studio but it has evolved more towards media /listening room, and the treatments will change a bit due to that. A fun little learning experience!

Genelec calibration doesn't do much to the speakers anymore, but the sub has 10 filters in use. The lowest room resonances are between the outer walls of the house in the 15-22hz range.
I'm very happy how it sounds.

 

[Flaesh]

mesophallic

exactly?))

 

[czt]

Raised the sub to half room height (on a stand and tray) in "Neumann" corner position (facing right wall). L, R RTA 1/48, periodic pink noise.

 

[garyrc]

exactly?))

Woops!

 

[changer]

After listening to a commercial low cardioid speaker I was disappointed with my chipboard brick and reworked my room EQ. I used the 9 available filters from the Hypex room EQ bank from 45 Hz to 660 Hz.

These graphs still show a 2 dB higher response <100 Hz than I am using now, coming from an even more extended bass shelf/room curve. It's not cardioid, but it does sound much better already. Top curve MMM (white noise periodic), below single spot sweep. Both 1/6 smoothing. I don't know why with MMM the <40 Hz region is not correctly measured, which is why I included the single spot. Thanks to a room mode, bass is extended to 30 Hz.

 

[Axo1989]

Does this mean that brachycephalic listeners get a different impression of stereo geometry than dolichocephalic ones? Is there just the right speaker cable or mid- field chrystal focal object, or perhaps a softwood diffuser that will correct for that, making everyone hear like the presumably mesophallic sound engineers?

TIL three words about heads !! Hopefully you mean mesocephalic ...

 

[changer]

Apparently, room modes are excited in varying degrees at different SPL levels. I did the MMM measurement at 10 dB higher SPL (black line) to find out whether the unconvincing measurement results below 40 Hz would change, and yes, this solved the issue, from the perspective of the measurement graph result. I arrived at a ca. 2 dB bass shelf below 100 Hz.

What I also noticed was that some modes, first and most the 45 Hz and the 60 Hz reacted differently to the increase in SPL. Finally, the single point sweep shows significantly lower bass response. Out of convenience, I decided to stick to the medium loud MMM (black line), but wonder what is the more faithful bass response depiction, the MMM or the sweep.

 

[Deleted member 48726]

Apparently, room modes are excited in varying degrees at different SPL levels. I did the MMM measurement at 10 dB higher SPL (black line) to find out whether the unconvincing measurement results below 40 Hz would change, and yes, this solved the issue, from the perspective of the measurement graph result. I arrived at a ca. 2 dB bass shelf below 100 Hz.

View attachment 348446

What I also noticed was that some modes, first and most the 45 Hz and the 60 Hz reacted differently to the increase in SPL. Finally, the single point sweep shows significantly lower bass response. Out of convenience, I decided to stick to the medium loud MMM (black line), but wonder what is the more faithful bass response depiction, the MMM or the sweep.

That would somehow oppose physics. Resonance amplify 1:x what energy is being the source.

 

[ernestcarl]

Apparently, room modes are excited in varying degrees at different SPL levels. I did the MMM measurement at 10 dB higher SPL (black line) to find out whether the unconvincing measurement results below 40 Hz would change, and yes, this solved the issue, from the perspective of the measurement graph result. I arrived at a ca. 2 dB bass shelf below 100 Hz.

View attachment 348446

What I also noticed was that some modes, first and most the 45 Hz and the 60 Hz reacted differently to the increase in SPL. Finally, the single point sweep shows significantly lower bass response. Out of convenience, I decided to stick to the medium loud MMM (black line), but wonder what is the more faithful bass response depiction, the MMM or the sweep.

How repeatable are your measurements?

And, are you certain it's not the varying noise levels picked up by the microphone? The microphone can be sensitive to movement and probably to a much smaller degree one's own body...

I do MMM at relatively higher SPLs (wearing hearing protection) and the resulting difference between MMM and sweeps aren't too "significant" in the bass (at least with my listening room and speakers).

Same subwoofer and speaker, but different main listening positions in the room

*applied a LF tail end to the MMM, and SPL leveled to match with the single-point measurement

 

[ernestcarl]

single point sweep shows significantly lower bass response.

BTW, this can happen if the FR significantly varies (peaks and dips) around the area of your mic position.

 

[Count Arthur]

My latest measurements. Nearfield, desk set-up, Neumann KH-150 monitors with Dynaudio 18S subwoofer.

It sounds good to me, but I don't pretend to understand all the graphs and welcome any feedback.

 

[OCA]

My latest measurements. Nearfield, desk set-up, Neumann KH-150 monitors with Dynaudio 18S subwoofer.

It sounds good to me, but I don't pretend to understand all the graphs and welcome any feedback.

View attachment 348489


View attachment 348490


View attachment 348491


View attachment 348492


View attachment 348493


View attachment 348494


View attachment 348495

It seems to me like you still need 3 more peaking filters at around 130Hz, 22Hz and 50Hz in order of importance.

 

[Count Arthur]

It seems to me like you still need 3 more peaking filters at around 130Hz, 22Hz and 50Hz in order of importance.

The first graph, is a combination of 4 measurements, averaged, then smoothed, which is used to calculate the PEQ filters.

The subsequent graphs, are all based on a single, un-smoothed measurement. Those additional graphs are not available for the averaged measurement.

Here are the PEQ filters that REW generated and you can see it has done something at the frequencies you suggested:

 

[OCA]

The first graph, is a combination of 4 measurements, averaged, then smoothed, which is used to calculate the PEQ filters.

The subsequent graphs, are all based on a single, un-smoothed measurement. Those additional graphs are not available for the averaged measurement.

Here are the PEQ filters that REW generated and you can see it has done something at the frequencies you suggested:

View attachment 348512

What's your DSP engine, are you sure the filters are reflected in it as is (same Q types, etc). Btw, you can comfortably use sharper filters with higher Q values (allow narrow filters ticked) below 200Hz. I'd use 1/48 octave band smoothing and use REW's auto EQ between 20-200Hz with that (with 0 dB boost). You will get a much smoother response in that area.

 

[Count Arthur]

What's your DSP engine, are you sure the filters are reflected in it as is (same Q types, etc). Btw, you can comfortably use sharper filters with higher Q values (allow narrow filters ticked) below 200Hz. I'd use 1/48 octave band smoothing and use REW's auto EQ between 20-200Hz with that (with 0 dB boost). You will get a much smoother response in that area.

I use JRiver, for local files, which allows me add unlimited PEQ filters, as far as I know, and I can enter them to match REW's suggestions:

I can also use the checkboxes to individually enable or disable specific filters to hear what they are doing and whether the effect is positive or not. You can see that those last couple, are -0.7dB and -0.9dB, and probably aren't really necessary.

If I'm not using JRiver, I can use the PEQ function in my RME DAC, but I'm limited to just five bands, so I pick the ones that I think are the most significant:

 

[OCA]

I use JRiver, for local files, which allows me add unlimited PEQ filters, as far as I know, and I can enter them to match REW's suggestions:

View attachment 348520

I can also use the checkboxes to individually enable or disable specific filters to hear what they are doing and whether the effect is positive or not. You can see that those last couple, are -0.7dB and -0.9dB, and probably aren't really necessary.

If I'm not using JRiver, I can use the PEQ function in my RME DAC, but I'm limited to just five bands, so I pick the ones that I think are the most significant:

View attachment 348523

REW "Generic" EQ uses proportional Q, I am not sure but I think JRiver also uses the same Q type so I don't think that's the reason but rePhase for example uses constant Q type by default so you might wanna check. Btw, you can use configurable PEQ under Generic in REW and set it up with the exact limits of your RME DAC (i.e 5 bands, Q limit, step size, etc.) and after that REW will optimize the EQ for these specific settings.

In general, you can do a lot better than simple PEQ with JRiver DSP and REW but you're on the right track

 

[GaryY]

Here are my room measurement using PEQ of DMP-A8.
Today I also tried FIR filter (because of curiosity), but outcome was not really better, especially big deep ~60Hz was much deeper.
There is some difference by grill, but not so much better or worse at least for me.
With some room treatment, I may be able to improve 100~200Hz, but I'm not planning to do.

Black : f226be grill off
Red : f226be grill on
Sub : Arendal 1961 1s

 

[IamJF]

Here are my room measurement using PEQ of DMP-A8.
Today I also tried FIR filter (because of curiosity), but outcome was not really better, especially big deep ~60Hz was much deeper.
There is some difference by grill, but not so much better or worse at least for me.
With some room treatment, I may be able to improve 100~200Hz, but I'm not planning to do.

Black : f226be grill off
Red : f226be grill on
Sub : Arendal 1961 1s

Did you measure both speakers at once? Can you check left and right separate?
The high frequency fall off looks to much to be real?