• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

Room acoutics issue 300-3000Hz

ChristianN

Member
Joined
Dec 3, 2020
Messages
38
Likes
2
Hi All,
I'm hoping for some advise. I have setup a 2 channel stereo for listening in my odd shaped basement, using Dynaudio Evoke50 speakers, which should have nice even frequency response, but in my room I miss around 6Db in the whole spectrum of 300->3000Hz. I had expected to deal with room modes in the lower bass area on certain frequnecies, but not this high up and not over such a wide area.

So any ideas why this could be or how to remedy? ( more absorbers, diffusers, position, should i DSP is all down to a liniar curve, would an absorber of diffuser be beneficial at the end of the hallway

Further info:
I did measure the speaker up really close, 6 inch and then it has quite liniar response, hence it must be the room ?
The room is concrete but both front and back has quite a mix of diffuser and absorbtion. There are 30Cm bass traps in 3 of 4 corners.
The left wall has 1 absorber in the 1.reflextion point.
The right wall is not there, its a hallway, maybee that has something to do with it.
There is heavy carpets ( 3 cm all over ) and 2cm acustic cealing.
The room is overall 30m2 but oddly shaped.
I placed the speaker for optimum soundstage, but have also tried other positions 1m around those spots, but the 6DB dip is the same.
I measured other listening positions 1m in all directions but the 6DB is the same as well allthough varying.
The amp is an Hegel H30, Preamd Primare, Streamer Primare ND5, graph is both channels.
I plan to integrate Subs at a later stage, and will eventually hence use DSP anyway, so that could be used if beneficial.
evoke50 mangler 6db i mid bass.jpg
kælder3.jpg
 

DWPress

Addicted to Fun and Learning
Forum Donor
Joined
May 30, 2018
Messages
962
Likes
1,369
Location
MI
I would give REW EQ a try - cut the peaks up to 300dB and small boost to the problem area if your amps can still drive your speakers to your desired SPL at LP with a -6dB cut or so. If you're going to integrate a sub soon then you'll need something like REW anyway so may as well start playing. The top end seems kind of bright so maybe some more targeted absorbtion for the rear wall but EQ might do it as well with a small gentle low pass filter to match your midrange boost.
 

Frgirard

Major Contributor
Joined
Apr 2, 2021
Messages
1,737
Likes
1,040
Hi All,
I'm hoping for some advise. I have setup a 2 channel stereo for listening in my odd shaped basement, using Dynaudio Evoke50 speakers, which should have nice even frequency response, but in my room I miss around 6Db in the whole spectrum of 300->3000Hz. I had expected to deal with room modes in the lower bass area on certain frequnecies, but not this high up and not over such a wide area.

So any ideas why this could be or how to remedy? ( more absorbers, diffusers, position, should i DSP is all down to a liniar curve, would an absorber of diffuser be beneficial at the end of the hallway

Further info:
I did measure the speaker up really close, 6 inch and then it has quite liniar response, hence it must be the room ?
The room is concrete but both front and back has quite a mix of diffuser and absorbtion. There are 30Cm bass traps in 3 of 4 corners.
The left wall has 1 absorber in the 1.reflextion point.
The right wall is not there, its a hallway, maybee that has something to do with it.
There is heavy carpets ( 3 cm all over ) and 2cm acustic cealing.
The room is overall 30m2 but oddly shaped.
I placed the speaker for optimum soundstage, but have also tried other positions 1m around those spots, but the 6DB dip is the same.
I measured other listening positions 1m in all directions but the 6DB is the same as well allthough varying.
The amp is an Hegel H30, Preamd Primare, Streamer Primare ND5, graph is both channels.
I plan to integrate Subs at a later stage, and will eventually hence use DSP anyway, so that could be used if beneficial.
View attachment 185692View attachment 185693
Have You measured the two speakers in the same time?
If yes
then it's useless : one speaker by measure.
else
where is the other speaker
End if
 
OP
C

ChristianN

Member
Joined
Dec 3, 2020
Messages
38
Likes
2
Hi

@DWPress I tried to REW EQ, and using my laptop and EqualizerAPO used the filters ( 15 filters) to modify the sound, which resulted in the graph below, which seems right. It sounds very much different, so need to figure out if its also better. Esp. the high notes are markedly numbed.
evoke50 mangler 6db i mid bass filtered.jpg
 
OP
C

ChristianN

Member
Joined
Dec 3, 2020
Messages
38
Likes
2
Hi @Frgirard
Have You measured the two speakers in the same time?
If yes
then it's useless : one speaker by measure.
else
where is the other speaker
End if
The measurement are for both ( Orange ), but i did measure them individually as well. they are very much alike ( Green and blue ), see below for all graphs.

Response.jpg
br
 

puppet

Senior Member
Joined
Dec 23, 2020
Messages
446
Likes
284
Interesting that this null falls right in the mid range department .. 430-3.5khz.
Is the mid range driver equidistant to the floor and ceiling by chance?
 
OP
C

ChristianN

Member
Joined
Dec 3, 2020
Messages
38
Likes
2
Interesting that this null falls right in the mid range department .. 430-3.5khz.
Is the mid range driver equidistant to the floor and ceiling by chance?
The mid range driver is 1m off the floor and 1,2m from the ceiling. So not exactly in the middle. but just 10cm off. If I measured freq. response af 2m height it has the same mid range dip.

But the tweeter is actually midway at 110cm
 
Last edited:

Thomas_A

Major Contributor
Forum Donor
Joined
Jun 20, 2019
Messages
3,422
Likes
2,407
Location
Sweden
What is the frequency response 1 meter from the speaker, gated measure? What mic do you use?
 

DWPress

Addicted to Fun and Learning
Forum Donor
Joined
May 30, 2018
Messages
962
Likes
1,369
Location
MI
It sounds very much different, so need to figure out if its also better. Esp. the high notes are markedly numbed.
Yes, it's going to sound different if you've gotten used to the dip and bright top end. Play some music you are intimately familiar with to get a feel for your EQ. Mess around with the EQ settings a bit more if it's easy to export to EAPO.

As mentioned, it is best to equalize the L/R speakers individually but it looks like both are following the same trend measurement wise. You're using a measurement mic with calibration files? Doing this as a MMM measurement or sweeps?
 
OP
C

ChristianN

Member
Joined
Dec 3, 2020
Messages
38
Likes
2
Yes, it's going to sound different if you've gotten used to the dip and bright top end. Play some music you are intimately familiar with to get a feel for your EQ. Mess around with the EQ settings a bit more if it's easy to export to EAPO.

As mentioned, it is best to equalize the L/R speakers individually but it looks like both are following the same trend measurement wise. You're using a measurement mic with calibration files? Doing this as a MMM measurement or sweeps?
Hi
Yes, the sound is better in the bottom, but high notes is off, kinda like if the speakers are out of phase, which they are not.
I use an Umik-1 with the std. batch calibration file, so its a not to say if my particular mic is actually in order.

I have been using sweeps ? also it seems that the NOISE button in measurement is disabled, even i was to use it.

I could try an individual EQ to see if chances it a bit, but the big dip in mid bass remains. So was still looking to see if I could do something with the room, so I could go more easy on the DSP.

br
 
Last edited:
OP
C

ChristianN

Member
Joined
Dec 3, 2020
Messages
38
Likes
2
What is the frequency response 1 meter from the speaker, gated measure? What mic do you use?
Hi
I use a std. Umik-1 mic with batch calibration file.
A measurement of left and right speaker, 1m i front of the speaker at the height of the mid bass speaker unit. Gated af 1-6ms produces the following. I'm not familiar with the gateing, but it would seem that it's not a liniar respons either due to the speaker of some other component ?
gated measured 1m.jpg
 

DWPress

Addicted to Fun and Learning
Forum Donor
Joined
May 30, 2018
Messages
962
Likes
1,369
Location
MI
Odd that your speakers dip then rise in the HF. If you like all the HF excitement (have grown used to it) the just try EQ up to 200-300, add a broad peak filter with low Q for the middle to match up. The fewer the filter bands usually the better. Gating helps eliminate sidewall, ceiling/floor bounce but those things exist at LP. Perhaps try the MMM method from LP to see how things correlate to the sweeps.

It would also be helpful if you included the legend at the bottom of the REW screen. Click the camera icon at the top left of the measurement window and it'll give you options on what info to include instead of just a screenshot from your monitor. We don't know what smoothing you're using among other things.
 

Thomas_A

Major Contributor
Forum Donor
Joined
Jun 20, 2019
Messages
3,422
Likes
2,407
Location
Sweden
Hi
I use a std. Umik-1 mic with batch calibration file.
A measurement of left and right speaker, 1m i front of the speaker at the height of the mid bass speaker unit. Gated af 1-6ms produces the following. I'm not familiar with the gateing, but it would seem that it's not a liniar respons either due to the speaker of some other component ?View attachment 185765
From these measurements it seems that it is you speaker and not the room that is causing the rather poor results at listening position (bass excluded). I would say that it is one of the more serious errors with a dip 1-2 kHz and peaking at 4 kHz. If you use 1/6 smoothing you should be able to create EQ filters to make a linear response on axis*, lifting 800 Hz-2 kHz and a dip centered at 4.5 kHz. However, off-axis is very important too so you should take at least 3 measurements at 0 degree, +/-15 degree and average them on axis. Also your speakers are quite large so at 1 meter distance, measurements may show some errors.

*Or IMO even better IMO, create a lite more energy 1.8 khz and less energy centered 3.5 kHz and 8-9 kHz.
 
Last edited:
OP
C

ChristianN

Member
Joined
Dec 3, 2020
Messages
38
Likes
2
From these measurements it seems that it is you speaker and not the room that is causing the rather poor results at listening position (bass excluded). I would say that it is one of the more serious errors with a dip 1-2 kHz and peaking at 4 kHz. If you use 1/6 smoothing you should be able to create EQ filters to make a linear response on axis*, lifting 800 Hz-2 kHz and a dip centered at 4.5 kHz. However, off-axis is very important too so you should take at least 3 measurements at 0 degree, +/-15 degree and average them on axis. Also your speakers are quite large so at 1 meter distance, measurements may show some errors.

*Or IMO even better IMO, create a lite more energy 1.8 khz and less energy centered 3.5 kHz and 8-9 kHz.
Hi thanks
I will try that, with a quick fiddling i can get REW to do almost what you suggest, but only if I set it to accept individual gains of 18Db and overall 18 db as well, which i kindy worrying numbers! Here's is what it suggets below. But it just based on yesterdays gaited response curve, so I would need to go back and do it including room response with averages off -axis measure and all, when I hope to get some time this evening.

Would you rather just based the correction on the "speaker alone measurement" or including the room response ?
eq.jpg
 

goat76

Major Contributor
Joined
Jul 21, 2021
Messages
1,269
Likes
1,385
Maybe you can borrow another pair of speakers and measure them as well, just to make sure it's not your measurement equipment that is behaving strangely?
 

Thomas_A

Major Contributor
Forum Donor
Joined
Jun 20, 2019
Messages
3,422
Likes
2,407
Location
Sweden
Hi thanks
I will try that, with a quick fiddling i can get REW to do almost what you suggest, but only if I set it to accept individual gains of 18Db and overall 18 db as well, which i kindy worrying numbers! Here's is what it suggets below. But it just based on yesterdays gaited response curve, so I would need to go back and do it including room response with averages off -axis measure and all, when I hope to get some time this evening.

Would you rather just based the correction on the "speaker alone measurement" or including the room response ?
View attachment 185827
I would treat speaker and room separately, but both would be included in the EQ in the end. I would do it step-wise, speaker first and do some listenings of music/female voices/choirs and then go for room modes.

I do not use REW for EQ, so I pass to someone else to help you out there. (I use Audyssey EQ < 100 Hz only).
 
OP
C

ChristianN

Member
Joined
Dec 3, 2020
Messages
38
Likes
2
Odd that your speakers dip then rise in the HF. If you like all the HF excitement (have grown used to it) the just try EQ up to 200-300, add a broad peak filter with low Q for the middle to match up. The fewer the filter bands usually the better. Gating helps eliminate sidewall, ceiling/floor bounce but those things exist at LP. Perhaps try the MMM method from LP to see how things correlate to the sweeps.

It would also be helpful if you included the legend at the bottom of the REW screen. Click the camera icon at the top left of the measurement window and it'll give you options on what info to include instead of just a screenshot from your monitor. We don't know what smoothing you're using among other things.
Hi thank you

That corresponds well with what @Thomas_A suggested, so will try that next. and also look into the MMM method.
Here is a better screenshot. I included another measure as i found a "Dali Rubicon LCR" speaker to measure, just to see the difference, same measurement, 1m avay, same gaiting 1-6ms. smoothed 1/6. but also it seems to exibit the dip, so it makes me wonder if amp, preamp or even the laptop does something wierd.
with another SPK2.jpg
 

Thomas_A

Major Contributor
Forum Donor
Joined
Jun 20, 2019
Messages
3,422
Likes
2,407
Location
Sweden
Could be something up with the microphone, especially >8 kHz. Try measure the speakers away from boundaries, reduce gating to 5 ms.
 

Thomas_A

Major Contributor
Forum Donor
Joined
Jun 20, 2019
Messages
3,422
Likes
2,407
Location
Sweden
Hi thank you

That corresponds well with what @Thomas_A suggested, so will try that next. and also look into the MMM method.
Here is a better screenshot. I included another measure as i found a "Dali Rubicon LCR" speaker to measure, just to see the difference, same measurement, 1m avay, same gaiting 1-6ms. smoothed 1/6. but also it seems to exibit the dip, so it makes me wonder if amp, preamp or even the laptop does something wierd.View attachment 185829
Here are my simple settings from another software; half Hamming, 7 ms duration, speaker at its nominal "position" near wall, about 1 meter from the speaker, 200 Hz-20 kHz, no smoothing. Left and right.

Graph.png
 
Top Bottom