Geddes et al. on Measuring Loudspeakers
http://www.diyaudio.com/forums/showthread.php?s=&threadid=120763
http://www.diyaudio.com/forums/showthread.php?s=&threadid=120763
Markus,
Do you see difference in measurements or listening over a period of playing time after the drivers had been first installed?
Do you see difference in measurements or listening over a period of playing time after the drivers had been first installed?
<standing on a hill in Austin, TX., ear cupped toward the Hudson, listening>
Is that Freedom Rock? Turn it up!
<straining to hear>
Is that Freedom Rock? Turn it up!
<straining to hear>
Thanks Zilch, so I guess I measured correctly. I maybe do some more measurements soon because I want to find out if and how the absorber in the back can be seen in a measurement. From listening with and without the absorber there's a huge difference in the perceived soundstage.
Best, Markus
Best, Markus
pos said:
Yah, beginning ~25 years ago:
http://www.aes.org/tmpFiles/elib/20081004/4573.pdf
Includes Linkwitz formula for vertical nulls @ 412.
Also see Fig. 10 @ 415.
How does Nathan 10 calc?
Then there's this. Click "Curves" here:
http://www.eighteensound.com/index.aspx?mainMenu=view_product&pid=178
Well scaling it in paint shop it looks like it's about 10" between the driver centers. Using the equation for a 2K crossover it's about 40-45 degrees. So 20 up and 20 down with no tilt. Using 1K double that assuming the spacing is right and I am not asleep at the wheel.
I don't get the nulls shown in the measurements???
Rob🙂
I don't get the nulls shown in the measurements???
Rob🙂
If those are nulls at crossover, it looks more like 1.2 kHz.
Also, we don't know which colors are which angles; they don't appear to be the same in the "Up" vs. "Down" plots.
It seems reasonable to assume that the lone single flatish line in each plot is on-axis, in which case the nulls appear immediately at 5°.
I'd expect Earl to be eager to clear up this issue, assuming there actually is one....
Assuming 10" C/C, I get alpha = 26.3°.
Also, we don't know which colors are which angles; they don't appear to be the same in the "Up" vs. "Down" plots.
It seems reasonable to assume that the lone single flatish line in each plot is on-axis, in which case the nulls appear immediately at 5°.
I'd expect Earl to be eager to clear up this issue, assuming there actually is one....
Assuming 10" C/C, I get alpha = 26.3°.
OOPS! Redo:ZilchLab said:
34.36°, almost half a nominal 70° beamwidth.
That ain't it, apparently....
Zilchlab, I'm too lazy to find the posts but Earl has said all along that the crossover is designed to make the flattest response at something like 20º off axis so the "lone flatish line"certainly isn't the on axis one.
"Zilchlab, I'm too lazy to find the posts but Earl has said all along that the crossover is designed to make the flattest response at something like 20º off axis"
Hello poptart
I believe those are horizontal polars not verticals. Flattest 20 degrees off axis with the toe in he recommends works. The vertical is what seems to be in question. The vertical listening window can't be toed in to the same extent. You could always tilt them on stands depending on height and distance if they have a narrow window. You would want that window be flat and lobe free over a reasonable coverage angle.
Lets see what Earl has to say. I am a bit confused by the posted measurements. I can't really tell what the first 40 degrees off axis looks like all that clearly. Would be better if the graphs posted were limited to that and each colored line was labeled for the measurement angle.
Rob🙂
Hello poptart
I believe those are horizontal polars not verticals. Flattest 20 degrees off axis with the toe in he recommends works. The vertical is what seems to be in question. The vertical listening window can't be toed in to the same extent. You could always tilt them on stands depending on height and distance if they have a narrow window. You would want that window be flat and lobe free over a reasonable coverage angle.
Lets see what Earl has to say. I am a bit confused by the posted measurements. I can't really tell what the first 40 degrees off axis looks like all that clearly. Would be better if the graphs posted were limited to that and each colored line was labeled for the measurement angle.
Rob🙂
I don't understand the question. What data set are you looking at? I can't comment on someone else's data, only my own. I will comment on the Nathan data shown on my web site, but I can't comment on the vertical data shown by Marcus, at least not in detail, until I have a chance to compare it with my own.
But regarding the two vertical nulls that DO occur in both Marcus data and mine, they will be at their deepest at "about" +- 20 degrees. There effect does begin to be noted earlier than that however. But these nulls are certainly NOT within the listening window when the speakers are in a reasonable sized room and on speaker stands. The nulls then actually help to minimize ceiling and floor reflections.
I have said before and I say again, I barely look at vertical responses because they do not have nearly the same weight in my mind as the horizontal.
But regarding the two vertical nulls that DO occur in both Marcus data and mine, they will be at their deepest at "about" +- 20 degrees. There effect does begin to be noted earlier than that however. But these nulls are certainly NOT within the listening window when the speakers are in a reasonable sized room and on speaker stands. The nulls then actually help to minimize ceiling and floor reflections.
I have said before and I say again, I barely look at vertical responses because they do not have nearly the same weight in my mind as the horizontal.
Here is my data with less angles (green = 0°, blue = 15°, yello = 30°, violet = 45°) and a bigger vertical spread. Mic-to-Speaker 1m, Window 4 ms = valid from 250 Hz.
Horizontal
Vertical upwards (There are room artefacts in the data - see Post #100 )
Vertical downwards (There are room artefacts in the data - see Post #100 )
Best, Markus
Horizontal
An externally hosted image should be here but it was not working when we last tested it.
Vertical upwards (There are room artefacts in the data - see Post #100 )
An externally hosted image should be here but it was not working when we last tested it.
Vertical downwards (There are room artefacts in the data - see Post #100 )
An externally hosted image should be here but it was not working when we last tested it.
Best, Markus
gedlee said:
Gedlee:
Kindly elaborate on your methods minimizing/controlling floor/ceiling reflections.
Thanks.
On the floor I use a two inch matress (Futon) covered with a rug so that it basically just looks like a rug. If you look closely you can tell that it is humped up. There is also the ubiuquitous coffee table to hold the beer bottles (or Bourbon in my case) - and the Cheez-its! So there is no floor reflection in my data.
The ceiling was major problem initially because my theater is in a basement with limited headroom - its only 7 feet in the HT, because of the drop ceiling, HVAC, joists, etc. So I built a sound reflector that basically takes all incoming wave and sends them back in the direction from wence they came. They then get scattered or absorbed by the speakers and back wall behind the speakers, which is very well damped. So nothing beyond about +- 15 degrees ever really gets out into the main room from the vertically radiating signals.
I thought that it might be interesting to show a "typical" horn woofer two way in the vertical direction so that we could all see what "typical" looks like.
The ceiling was major problem initially because my theater is in a basement with limited headroom - its only 7 feet in the HT, because of the drop ceiling, HVAC, joists, etc. So I built a sound reflector that basically takes all incoming wave and sends them back in the direction from wence they came. They then get scattered or absorbed by the speakers and back wall behind the speakers, which is very well damped. So nothing beyond about +- 15 degrees ever really gets out into the main room from the vertically radiating signals.
I thought that it might be interesting to show a "typical" horn woofer two way in the vertical direction so that we could all see what "typical" looks like.
- Status
- Not open for further replies.