Help read this SPL measurement graph?

[MalVeauX]

Hello,

I had a minute yesterday to attempt to measure a monitor out in my 5 acre field. Kind of a pseudo anechoic chamber, if you will. I realize its not the same. But it's better than measuring in a room, if I wanted to get a more accurate measurement of the speaker's ability without room influence. But even then, I think I picked up some stuff even in the field! Durrrrr. So I'm trying to understand what it may be if anyone sees what's obvious and can recommend what to do.

I dragged some extension cables out into my field and took my laptop, dac and an amplifier along with a speaker stand and a speaker (bookshelf monitor). The stand is wooden and sand filled with isolation foam on its shelf that the speaker is resting on. It was standing up in grass on sand. Maybe 80 feet away is a shed (to the rear left of the speaker facing out into my 5 acre field). There's an oak tree on the other side of my field. Nothing else around to reflect other than a wall of oak trees to the far right over 200+ yards away. But, maybe that mattered? Could a tree or tree line over 100 yards away disrupt the measurement?

Here's a single speaker measurement, at 1 meter distance, mic was positioned pointing at the tweeter to simulate ear level response.

I notice what looks like comb filtering going on?
What could be causing that out there in the field?
It's like there's a reflective boundary somewhere?
Ideas?

Just trying to learn more about this. I don't expect a flat line or anything, but mainly wanted to be able to learn more about the graph, what is causing that, and what I can do to correct for it if possible so I know what's going on.

Very best,

 

[TLS Guy]

Hello,

I had a minute yesterday to attempt to measure a monitor out in my 5 acre field. Kind of a pseudo anechoic chamber, if you will. I realize its not the same. But it's better than measuring in a room, if I wanted to get a more accurate measurement of the speaker's ability without room influence. But even then, I think I picked up some stuff even in the field! Durrrrr. So I'm trying to understand what it may be if anyone sees what's obvious and can recommend what to do.

I dragged some extension cables out into my field and took my laptop, dac and an amplifier along with a speaker stand and a speaker (bookshelf monitor). The stand is wooden and sand filled with isolation foam on its shelf that the speaker is resting on. It was standing up in grass on sand. Maybe 80 feet away is a shed (to the rear left of the speaker facing out into my 5 acre field). There's an oak tree on the other side of my field. Nothing else around to reflect other than a wall of oak trees to the far right over 200+ yards away. But, maybe that mattered? Could a tree or tree line over 100 yards away disrupt the measurement?

Here's a single speaker measurement, at 1 meter distance, mic was positioned pointing at the tweeter to simulate ear level response.

I notice what looks like comb filtering going on?
What could be causing that out there in the field?
It's like there's a reflective boundary somewhere?
Ideas?

Just trying to learn more about this. I don't expect a flat line or anything, but mainly wanted to be able to learn more about the graph, what is causing that, and what I can do to correct for it if possible so I know what's going on.

View attachment 47617

View attachment 47618

Very best,

The details are scant, but if you did use a proper technique, I would guess this is a small sealed speaker with a crossover somewhere between 2 and three K. It has severe problems at crossover with likely and out of phase at crossover. I suspect it is a sealed speaker with poor alignment, with excessive ripple before the falling response.

So either there is a highly suspect measurement technique, or it is a lousy speaker.

 

[shadyJ]

You are getting lots of ground bounce cancellation. You need to window the response to filter out later arriving signals then the acoustic reflection off of the ground. If your speaker stand is like 3 feet tall, you are probably only getting a 3 or 4ms window before acoustic reflections corrupt your measurement.

Bring up that measurement in REW, and go to the 'Filtered IR' tab. In the upper left hand corner of that window, there is a drop down box that has two options: 'dBFS' and '% FS.' Make sure it is set to '% FS' and take a screenshot with -10ms to +20ms window. Post that window here so I can see the reflection point.

 

[MalVeauX]

The details are scant, but if you did use a proper technique, I would guess this is a small sealed speaker with a crossover somewhere between 2 and three K. It has severe problems at crossover with likely and out of phase at crossover. I suspect it is a sealed speaker with poor alignment, with excessive ripple before the falling response.

So either there is a highly suspect measurement technique, or it is a lousy speaker.

Probably my technique. The crossover is around 1900hz. It's not sealed, rear ported.

You are getting lots of ground bounce cancellation. You need to window the response to filter out later arriving signals then the acoustic reflection off of the ground. If your speaker stand is like 3 feet tall, you are probably only getting a 3 or 4ms window before acoustic reflections corrupt your measurement.

Bring up that measurement in REW, and go to the 'Filtered IR' tab. In the upper left hand corner of that window, there is a drop down box that has two options: 'dBFS' and '% FS.' Make sure it is set to '% FS' and take a screenshot with -10ms to +20ms window. Post that window here so I can see the reflection point.

Thanks, I figured it was something to do with boundaries since the combing effect was obvious to my amateur eyes (I think?). The stand is around 32 inches or so. Foam added another 2 inches. Then the speaker itself. The field is grass (tall grass though).

The IR screen shot (if I got the values right):

Very best,

 

[shadyJ]

That's the right window but the scale is far too large to be useful. You need to narrow it down to a -10ms to +20ms scale. That scale that you posted shows a -100ms to +200ms tine window.

 

[MalVeauX]

That's the right window but the scale is far too large to be useful. You need to narrow it down to a -10ms to +20ms scale. That scale that you posted shows a -100ms to +200ms tine window.

Doh, just saw the units. Duh. Hopefully this is correct now:

Very best,

 

[shadyJ]

Probably my technique. The crossover is around 1900hz. It's not sealed, rear ported.



Thanks, I figured it was something to do with boundaries since the combing effect was obvious to my amateur eyes (I think?). The stand is around 32 inches or so. Foam added another 2 inches. Then the speaker itself. The field is grass (tall grass though).

The IR screen shot (if I got the values right):

View attachment 47619

Very best,

go to the 'limits' button in the upper right hand corner. Set 'left s' to -0.010 and 'right s' to 0.030.

 

[MalVeauX]

Sorry, just realized it was in the hundred's place.

Very best,

 

[MalVeauX]

go to the 'limits' button in the upper right hand corner. Set 'left s' to -0.010 and 'right s' to 0.030.

Very best,

 

[shadyJ]

Doh, just saw the units. Duh. Hopefully this is correct now:

View attachment 47620

Very best,

Ok that's good. Your ground reflection is occurring at about 3.5ms. What you need to do to filter that out is go to the 'IR Windows' button on the top of the REW screen. Click that, and in the window that pops up, there is a count box labeled 'Right Window (ms):' The default there is 500. Set it to 3.5.

Than post the frequency response. It would help if the response window had a 50dB Y-axis scale, so just set it for 50dB to 100dB.

 

[shadyJ]

This is the screen, although from an older version of REW:

See where it says 'Right Window (ms):' In that image, it is set to 4.4. You need to set yours to 3.5.

This is a page that describes what has to be done.

 

[MalVeauX]

This is the screen, although from an older version of REW:

See where it says 'Right Window (ms):' In that image, it is set to 4.4. You need to set yours to 3.5.

This is a page that describes what has to be done.

Very best,

 

[MalVeauX]

If there's a better way to measure, I'm happy to learn it. I had read about using open fields for this, but obviously there's problems there too! I can't setup an anechoic chamber in my house as I just don't have extra space for it nor the capital to flesh it out, plus, I have no real application for this other than learning and enjoying tinkering with this stuff to learn more about it. But if there's a better way to measure a speaker, simply, without the room influence as much, I'd try that too.

I've read about burying a speaker facing up, but I'm not going that far...

Very best,

 

[shadyJ]

View attachment 47626

Very best,

OK, better. Such a short window doesn't give you much low-frequency resolution, so you can just shorten the window to 500Hz. In order to measure bass response, you would need to use another measurement technique. But at least now er can see what is going on above 1kHz.

You say that the crossover is at 1.9kHz, but it really looks to be around 3kHz. You have a nearly 8dB dip at 3kHz that is symptomatic of crossover cancellation. If it is not a crossover issue, then that tweeter has some serious problems. The tweeter is pretty rough above that point as well.

If you really want to characterize that speaker, you need more than just the on-axis response though. Next time you go measuring outdoors, capture the response at 15, 30, and 45 degrees as well. you might also want to get the 10, 20, and 30-degree response on the vertical axis as well- both above and below the tweeter.

Also, the higher you can elevate the speaker, the less you have to window out from the impulse response, and the more accuracy you can get in lower frequencies. I measure speakers at a 7.5' height, and that only gives me useful information down to 200Hz. Your ground reflection isn't allowing you any serious accuracy below maybe 800Hz or so.

 

[shadyJ]

If there's a better way to measure, I'm happy to learn it. I had read about using open fields for this, but obviously there's problems there too! I can't setup an anechoic chamber in my house as I just don't have extra space for it nor the capital to flesh it out, plus, I have no real application for this other than learning and enjoying tinkering with this stuff to learn more about it. But if there's a better way to measure a speaker, simply, without the room influence as much, I'd try that too.

I've read about burying a speaker facing up, but I'm not going that far...

Very best,

You are off to a good start. If you can get the speaker higher off the ground, and also devise an easy way to take off-axis measurements to precise angles, then you will be doing pretty good. You will also want to use a speaker that is a known quantity so you know how close you are getting to accurate measurements. As it is, there is no way to be sure that your measurements are truly accurate. Try to get your hands on a speaker that has been measured a lot. The KEF LS50 has a lot of measurements. The Revel M16 is pretty well-documented. The JBL LSR 305P mk2. The Dayton MK4xx speakers. There are others...

If you want to capture bass accurately, you should probably use groundplane measurements. That is a whole other can of worms...

 

[MalVeauX]

Thanks for all the help & info!

I will try again some time and maybe use different hardware. And I'll see what I can do to get it higher up off the ground. A ladder and a board maybe? I need to refresh setup too. I had just setup at 1 meter, on axis, straight at the tweeter roughly. I wanted to make sure my 50 watt amp I dragged out could handle referencing at 75db at 1 watt or less and have lots of room to spare, and then do the measurements.

I assembled this speaker, so I'm just trying to learn from the entire process. It may be poor because I'm poor at putting it together. It sounds good in my office and I'm happy with it. But measuring is a whole different ball game completely and while I'm not throwing stones, again, I'm just trying to learn from the whole process to know more about all of it.

I got the crossover and tweeter from Dennis Murphy. I got the cabinet and woofer from Parts Express (Dayton).

My build log:

I want new bookshelves, but I don't need them. UGH!

Thanks all, super helpful! So I'll cup it in the bottom chamber, and looks like none of it goes in the upper chamber. I can always take it apart and move it or add more and once it's sounding right, I'll glue it. I plan to secure the crossover to the bottom of the cabinet so it doesn't flop...

forums.audioholics.com

Very best,

 

[shadyJ]

Thanks for all the help & info!

I will try again some time and maybe use different hardware. And I'll see what I can do to get it higher up off the ground. A ladder and a board maybe? I need to refresh setup too. I had just setup at 1 meter, on axis, straight at the tweeter roughly. I wanted to make sure my 50 watt amp I dragged out could handle referencing at 75db at 1 watt or less and have lots of room to spare, and then do the measurements.

I assembled this speaker, so I'm just trying to learn from the entire process. It may be poor because I'm poor at putting it together. It sounds good in my office and I'm happy with it. But measuring is a whole different ball game completely and while I'm not throwing stones, again, I'm just trying to learn from the whole process to know more about all of it.

I got the crossover and tweeter from Dennis Murphy. I got the cabinet and woofer from Parts Express (Dayton).

My build log:

I want new bookshelves, but I don't need them. UGH!

Thanks all, super helpful! So I'll cup it in the bottom chamber, and looks like none of it goes in the upper chamber. I can always take it apart and move it or add more and once it's sounding right, I'll glue it. I plan to secure the crossover to the bottom of the cabinet so it doesn't flop...

forums.audioholics.com

Very best,

You can use a ladder. Make sure that whatever you set the speaker on does not stick out in front of the speaker though. That would be a source of reflection. A 1-meter distance is fine, and aiming the mic at the tweeter is generally a good idea.

 

[Pogre]

I'm learning a lot in this thread, so I'm tagging it! Really appreciate your input Shady. I've had some long standing questions about some of these settings.

 

[TLS Guy]

If there's a better way to measure, I'm happy to learn it. I had read about using open fields for this, but obviously there's problems there too! I can't setup an anechoic chamber in my house as I just don't have extra space for it nor the capital to flesh it out, plus, I have no real application for this other than learning and enjoying tinkering with this stuff to learn more about it. But if there's a better way to measure a speaker, simply, without the room influence as much, I'd try that too.

I've read about burying a speaker facing up, but I'm not going that far...

Very best,

There is something not right about this, as the bass roll off is not like a sealed and certainly not a ported speakers.