• 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!

My loudspeaker measurements (work in progress)

TimVG

Major Contributor
Joined
Sep 16, 2019
Messages
1,181
Likes
2,573
I'll be doing some loudspeaker measurements as time allows

The speakers are raised so that the acoustical axis about 2m from the ground (room as a height of 4m) - first lateral reflections are 4m away.

The measurements are made quite close at 50cm - but I take care in finding the acoustical axis of the speaker in question. I tried at 1m and further, but resolution got much worse in the lower frequency while making little to no difference as you go higher.

This is by no means a controlled environment, nor is my equipment high-end. This is done using a Dayton UMM6 and a simple laptop with a line out. Take these with a grain of salt is what I'm saying. While the measurements can be compared to one another, as the relative differences are correct, I am not claiming they are entirely accurate in itself.

Please not I am diverging from the ANSI 2034 standard - I have to do these at my work during breaks (it's a nice large room), so I try to these as efficient as possible.

Black = on-axis
Green (listening window) = 0° - 10° - 20° - 30° horizontal average plus up and down 10° from the 0° axis
Red (early horizontal reflections) = 40° - 50° - 60° - 70°- 80° horizontal average
Bottom = early horizontal reflection directivity index

I'll be adding more and more as time allows
 
Last edited:

Krunok

Major Contributor
Joined
Mar 25, 2018
Messages
4,600
Likes
3,065
Location
Zg, Cro
Is bottom curve (early reflections) calculated within REW as green curve - red curve?
 
OP
TimVG

TimVG

Major Contributor
Joined
Sep 16, 2019
Messages
1,181
Likes
2,573
Is bottom curve (early reflections) calculated within REW as green curve - red curve?

Divided by and scaled in level. Substracting showed some very weird anomalies
 

Juhazi

Major Contributor
Joined
Sep 15, 2018
Messages
1,717
Likes
2,897
Location
Finland
What IR gating is used? I would like to see more smoothing, eg. 1/6.
Is the bottom curve supposed to represent DI (directivity index) ?
In general these look ok, but I prefer to see all 0-90deg and vert. curves overlaid, please show them too with "indexed" graphs! (veerticals with -20dB offset)
If you take the measurementt at say 100dB at 0.5m, distortion curve would be interesting too!

I used my outdoor measurements of a coaxial speaker as reference. Sorry, no vertical off-axis. Because the speaker was on a 80cm stand, we see lots of reflections below 1000Hz. Smoothing by ERB here smooths out low response. Directivity-wise we can still see what is happengn. Low end response is useless in my measurements. Distrtion graphs have faulty titles, gating and smoothing is set by REW automatically, not by ERB setting!

exa tile.jpg
 
Last edited:
OP
TimVG

TimVG

Major Contributor
Joined
Sep 16, 2019
Messages
1,181
Likes
2,573
1/24 octave smoothing. No gating.
Since I only have about 20 min to set up and measure for each session I can't do all you request but I'll see about sharing the mdat files
 

617

Major Contributor
Forum Donor
Joined
Mar 18, 2019
Messages
2,404
Likes
5,296
Location
Somerville, MA
Tim, learn about gating. These measurements look quite good though. That sort of sine wave ripple in 1-2khz area will go away with proper gating.
 

Krunok

Major Contributor
Joined
Mar 25, 2018
Messages
4,600
Likes
3,065
Location
Zg, Cro
Tim, learn about gating. These measurements look quite good though. That sort of sine wave ripple in 1-2khz area will go away with proper gating.

:facepalm:

Cut that crappy attitude. Such measurements are supposed to be done in anechoic environment without gating.
 

617

Major Contributor
Forum Donor
Joined
Mar 18, 2019
Messages
2,404
Likes
5,296
Location
Somerville, MA
:facepalm:

Cut that crappy attitude. Such measurements are supposed to be done in anechoic environment without gating.

It's pretty easy to get useful measurements without an anechoic chamber. Apologies if I sounded condescending.
 

Juhazi

Major Contributor
Joined
Sep 15, 2018
Messages
1,717
Likes
2,897
Location
Finland
1/24 octave smoothing. No gating.
Since I only have about 20 min to set up and measure for each session I can't do all you request but I'll see about sharing the mdat files

So you saved the .mdat of REW? Then just open the file and change gating and, smoothing as you like!
One feature I am missing is normalization of measurements in ovelay, like Omnimic can do.

Soundstage measurements are done in an anechoic chamber and use long gating (REW default is 1000ms if I remember right) Anyway one should always study the impulse response to see where reflections start to happen, and this sets the guidelines for analysis. An experienced person can "read through" those anomalies caused by reflections and short gating.

exa filt IR.jpg
 
Last edited:

Krunok

Major Contributor
Joined
Mar 25, 2018
Messages
4,600
Likes
3,065
Location
Zg, Cro
It's pretty easy to get useful measurements without an anechoic chamber. Apologies if I sounded condescending.

So you saved the .mdat of REW? Then just open the file and change gating and, smoothing as you like!

Such measurements are actually supposed to be done in anechoic environment. I also don't really understand what kind of gating are you guys proposing without even seeing the on-axis IR.
 

617

Major Contributor
Forum Donor
Joined
Mar 18, 2019
Messages
2,404
Likes
5,296
Location
Somerville, MA
Such measurements are actually supposed to be done in anechoic environment. I also don't really understand what kind of gating are you guys proposing without even seeing the on-axis IR.

A lot of high performance speakers are designed without anechoic chambers. A combination of gated and nearfield measurements can result in measurements which are as good or better than those produced in an anechoic chamber.
 

Krunok

Major Contributor
Joined
Mar 25, 2018
Messages
4,600
Likes
3,065
Location
Zg, Cro
A lot of high performance speakers are designed without anechoic chambers. A combination of gated and nearfield measurements can result in measurements which are as good or better than those produced in an anechoic chamber.

Look, it's actually quite simple, just read carefully and pay attention:

1) ANSi 2034 (aka spinorama) measruements are supposed to be done in anechoic environment. Any reflections will cause serious deviation from correct results

2) you can't make "better" in-room measurements than those in anechoic environment

Now let's wait for the OP to give more info instead of your trash talk.
 

Krunok

Major Contributor
Joined
Mar 25, 2018
Messages
4,600
Likes
3,065
Location
Zg, Cro
What IR gating is used? I would like to see more smoothing, eg. 1/6.
Is the bottom curve supposed to represent DI (directivity index) ?
In general these look ok, but I prefer to see all 0-90deg and vert. curves overlaid, please show them too with "indexed" graphs! (veerticals with -20dB offset)
If you take the measurementt at say 100dB at 0.5m, distortion curve would be interesting too!

View attachment 34926

All your THD data is under the noise which makes it highly unreliable. I believe we have learned recently from John Mulcahy that distortion should be done by RTA stepped sine when measured in-room.

Besides, ANSI 2034 doesn't mention distortion at all. Neither does Toole.
 
OP
TimVG

TimVG

Major Contributor
Joined
Sep 16, 2019
Messages
1,181
Likes
2,573
Gating doesn't show any more useful info than what's been presented here. I'm on mobile now but will post more tomorrow.
 

617

Major Contributor
Forum Donor
Joined
Mar 18, 2019
Messages
2,404
Likes
5,296
Location
Somerville, MA
Krunok, the gating removes reflections from the measurement. Believe me it works. Earl Geddes used a turntable with two sheets of melamine separated by grease to design his speakers which have extremely refined off axis performance. It is relatively cumbersome to set gating over a bunch of measurements but it can be done with free software. The low frequency response is measured separately in the near field, where the magnitude of reflections far smaller than that of the DUT. You then splice them together. This enables measurement down to 1hz if desired, provided your mic is up to it.

I agree with your comments regarding distortion measurement, which is really hard to do with consumer grade equipment.
 

dc655321

Major Contributor
Joined
Mar 4, 2018
Messages
1,597
Likes
2,235
Look, it's actually quite simple, just read carefully and pay attention:

1) ANSi 2034 (aka spinorama) measruements are supposed to be done in anechoic environment. Any reflections will cause serious deviation from correct results

2) you can't make "better" in-room measurements than those in anechoic environment

Now let's wait for the OP to give more info instead of your trash talk.

@617 is not wrong. Please lose the holier-than-thou attitude.
Anechoic chambers cannot provide reliable data at the lowest frequencies. Nearfield or ground-plane can.
Blend those techniques with gated measurement for frequencies > 0.5-0.8kHz, and results comparable to anechoic measurements can be had.
 
OP
TimVG

TimVG

Major Contributor
Joined
Sep 16, 2019
Messages
1,181
Likes
2,573
Since the measurements are performed in the nearfield as explained, and since the room is large and the speaker optimally positioned, gating does not provide any more useful info. A very near field measurement of mid/low frequency driver would be useful to show its behavior with regards to resolution lost below 1kHz
 

Krunok

Major Contributor
Joined
Mar 25, 2018
Messages
4,600
Likes
3,065
Location
Zg, Cro
Krunok, the gating removes reflections from the measurement. Believe me it works. Earl Geddes used a turntable with two sheets of melamine separated by grease to design his speakers which have extremely refined off axis performance. It is relatively cumbersome to set gating over a bunch of measurements but it can be done with free software. The low frequency response is measured separately in the near field, where the magnitude of reflections far smaller than that of the DUT. You then splice them together. This enables measurement down to 1hz if desired, provided your mic is up to it.

I agree with your comments regarding distortion measurement, which is really hard to do with consumer grade equipment.

I don't need your "lessons" M8, I have done it myself. What I don't understand is how you proposed to do the gating without seeing the IR.
 
Top Bottom