#292 Bolserst wrote:
Would you expect measurements taken at 1.5m with Setup (A) & (B) to give essentially identical results?
After reading your recent article:
http://gedlee.azurewebsites.net/downloads/On Measureing loudspeakers.pdf
This observation was of particular interest:
"I occasionally look at the vertical, but not that often as I don’t find anything surprising or interesting. (Except in cases where I might suspect bad things to happen based on the design and then I usually find them - i.e. across the diagonal of a rectangular waveguide."
In my line of work providing sound reinforcement for live bands, both the horizontal and vertical response of loudspeakers are almost equally important, but had never really considered the diagonal response of rectangular horns (which I have used almost exclusively) separately. Being curious as to what "bad things might happen" I measured the response of my latest speaker design, the SynTripP, across the diagonal, as well as horizontally and vertically. The SynTripP is a 2-way, 2-part virtual single point source conical horn using two B&C 10CL51 10" for low frequencies (LF) in an offset bandpass bass reflex arrangement with a Celestion CDX14-3050 three inch diaphragm 1.4" exit high frequency (HF) driver at the horn apex. There are four injection ports for the 10" drivers located near the horn apex and four bass reflex ports at the end of the first horn section.
http://www.diyaudio.com/forums/mult...ual-single-point-source-horn.html#post4114406
The SynTripP's dispersion with the secondary horn expansion attached is 86 degrees horizontal x 18 vertical, with pattern control down to 310 Hz (-6 dB 42.5 degrees off axis) horizontal and 460 Hz vertical. Without the second horn section pattern control is lost below 430 Hz horizontal and 1100 Hz vertical.
Measurements showed the diagonal polars are a bit more consistent in the off axis level reduction in high frequency coverage compared to the horizontal, but show a bit of "waist-banding" (coverage that gets narrower, then widens back out) in a few frequency regions, though not to an extent that they would be apparent in any normal listening environment. Some of the differences may have been due to the test conditions, the diagonal measurement required propping the cabinet up, leaving a wedge shaped cavity between the cabinet bottom and the turntable.
The horizontal and vertical patterns of the SynTripP are quite consistent, the design of the grill frame using half again the angle of the 86 x 36 degree conical waveguide seem to be effective both in reducing diffraction and waist-banding. There was no fabric in the frame for the tests conducted so far, the fabric will fill in some small gaps around the frame, although I doubt it will make a measurable change in response.
The SynTripP tests before and after the eight port holes were cut confirmed several things:
1) The point of rotation makes a big difference in the HF response towards the outer regions of the horn coverage angle, my first tests with the rotation point slightly behind the horn apex made the HF response look like it was narrowing at 40 degrees off axis, the proper rotation point puts the -6 dB point at slightly beyond the initial horn wall angle.
2) Outside of 25 degrees off axis, the 10" driver's throat injection ports cause a 3 dB narrowing of pattern centered around 3500 Hz, the waist-banding lowers slightly in frequency the further off axis the measurement is taken. At 1900 Hz, the injection ports cause a 2 dB "midriff bulge", a widening of pattern. Even with the ports the SynTripP polars appear to be about as smooth as your NS15, though the 1" exit OS throat used on the NS15 causes narrowing above around 11 kHz, while the SynTripP does not narrow even at 20 kHz, remarkable for a 1.4" exit driver not using a reduced throat aperture. Although response above 10 kHz is relatively unimportant, I prefer that range to have even response just like any other octave, even though it will probably only be a memory to me in another 20 years or so. Vertical response is also important, the virtual single point source the offset LF drivers sharing the horn with the HF driver provides eliminates the vertical lobes in the crossover region a separate horn and bass driver location can not avoid.
3) 3 dB of waist-banding occurs around 750 Hz, which is reduced and lowered in frequency using the secondary horn expansion.
4) The BR ports have virtually no effect on HF polar response, covering them resulted in near identical frequency response.
5)Testing at a height of only 2 meters with mic at 4 meters from the cabinet front introduces ground bounce errors, the measured response at 4 meters off the ground is much smoother than what is shown in the polars below. I plan to re-test the polars at four meters high and four meters distant at some point.
I have found a few cabinets measured response using your polar mapping program scattered through this thread, but can't seem to find many of the ones you have mentioned. Could you please provide a link to where they are located (you may want to put the link in the OP)?
Art
Would you expect measurements taken at 1.5m with Setup (A) & (B) to give essentially identical results?
Recent tests confirm a problem illustrated in the "B" rotational point setup illustrated below.
After reading your recent article:
http://gedlee.azurewebsites.net/downloads/On Measureing loudspeakers.pdf
This observation was of particular interest:
"I occasionally look at the vertical, but not that often as I don’t find anything surprising or interesting. (Except in cases where I might suspect bad things to happen based on the design and then I usually find them - i.e. across the diagonal of a rectangular waveguide."
In my line of work providing sound reinforcement for live bands, both the horizontal and vertical response of loudspeakers are almost equally important, but had never really considered the diagonal response of rectangular horns (which I have used almost exclusively) separately. Being curious as to what "bad things might happen" I measured the response of my latest speaker design, the SynTripP, across the diagonal, as well as horizontally and vertically. The SynTripP is a 2-way, 2-part virtual single point source conical horn using two B&C 10CL51 10" for low frequencies (LF) in an offset bandpass bass reflex arrangement with a Celestion CDX14-3050 three inch diaphragm 1.4" exit high frequency (HF) driver at the horn apex. There are four injection ports for the 10" drivers located near the horn apex and four bass reflex ports at the end of the first horn section.
http://www.diyaudio.com/forums/mult...ual-single-point-source-horn.html#post4114406
The SynTripP's dispersion with the secondary horn expansion attached is 86 degrees horizontal x 18 vertical, with pattern control down to 310 Hz (-6 dB 42.5 degrees off axis) horizontal and 460 Hz vertical. Without the second horn section pattern control is lost below 430 Hz horizontal and 1100 Hz vertical.
Measurements showed the diagonal polars are a bit more consistent in the off axis level reduction in high frequency coverage compared to the horizontal, but show a bit of "waist-banding" (coverage that gets narrower, then widens back out) in a few frequency regions, though not to an extent that they would be apparent in any normal listening environment. Some of the differences may have been due to the test conditions, the diagonal measurement required propping the cabinet up, leaving a wedge shaped cavity between the cabinet bottom and the turntable.
The horizontal and vertical patterns of the SynTripP are quite consistent, the design of the grill frame using half again the angle of the 86 x 36 degree conical waveguide seem to be effective both in reducing diffraction and waist-banding. There was no fabric in the frame for the tests conducted so far, the fabric will fill in some small gaps around the frame, although I doubt it will make a measurable change in response.
The SynTripP tests before and after the eight port holes were cut confirmed several things:
1) The point of rotation makes a big difference in the HF response towards the outer regions of the horn coverage angle, my first tests with the rotation point slightly behind the horn apex made the HF response look like it was narrowing at 40 degrees off axis, the proper rotation point puts the -6 dB point at slightly beyond the initial horn wall angle.
2) Outside of 25 degrees off axis, the 10" driver's throat injection ports cause a 3 dB narrowing of pattern centered around 3500 Hz, the waist-banding lowers slightly in frequency the further off axis the measurement is taken. At 1900 Hz, the injection ports cause a 2 dB "midriff bulge", a widening of pattern. Even with the ports the SynTripP polars appear to be about as smooth as your NS15, though the 1" exit OS throat used on the NS15 causes narrowing above around 11 kHz, while the SynTripP does not narrow even at 20 kHz, remarkable for a 1.4" exit driver not using a reduced throat aperture. Although response above 10 kHz is relatively unimportant, I prefer that range to have even response just like any other octave, even though it will probably only be a memory to me in another 20 years or so. Vertical response is also important, the virtual single point source the offset LF drivers sharing the horn with the HF driver provides eliminates the vertical lobes in the crossover region a separate horn and bass driver location can not avoid.
3) 3 dB of waist-banding occurs around 750 Hz, which is reduced and lowered in frequency using the secondary horn expansion.
4) The BR ports have virtually no effect on HF polar response, covering them resulted in near identical frequency response.
5)Testing at a height of only 2 meters with mic at 4 meters from the cabinet front introduces ground bounce errors, the measured response at 4 meters off the ground is much smoother than what is shown in the polars below. I plan to re-test the polars at four meters high and four meters distant at some point.
I have found a few cabinets measured response using your polar mapping program scattered through this thread, but can't seem to find many of the ones you have mentioned. Could you please provide a link to where they are located (you may want to put the link in the OP)?
Art
Attachments
Last edited:
The are in the database connected to the PolarMap program on my website, that is in the OP I believe.
see GedLee LLC
If you read the write up then you went right by it.
Probably did go by it, a Mac can't open the application. Posting PDF pictures of the various polar maps you have created (imported) would make viewing possible for those of us not using a Windows platform, and make it possible for the growing number of iPhones, iPads, etMc.
As I don't know how to convert my Smaart files to fit your polar map program's Holm Impulse import format, I'd also be very interested in seeing those polar responses in the "old fashioned" frequency response format as used in the previous post #321.
My creative left brain hemisphere likes the polar map presented looking like a Rorschach test, but the right rectilinear reptilian hemisphere clings to the simple amplitude on the vertical scale and frequency on the horizontal scale...
Art
Last edited:
Art, cool project! If you send me the impulses as wav files I could do a contour plot in ARTA. I would want to gate the impulse because what you show, IMO, really isn't that great with the ground reflection present and I think this would be clear in a contour plot. Maybe I'm reading your screen grab wrong. Data beyond 42.5° would really be necessary as well.
It would be nice if you could show us the horizontal polar fr plots by themselves, with a standard 5dB scale in your thread.
It would be nice if you could show us the horizontal polar fr plots by themselves, with a standard 5dB scale in your thread.
Nate,
Only did magnitude (and phase) response tests, no impulse files. Will post the horizontal polar response in the SynTripP thread with 1 dB increment scale out to 90 degrees.
Art
Originally Posted by weltersys
Posting PDF pictures of the various polar maps you have created (imported) would make viewing possible for those of us not using a Windows platform.
Art
Buying a Windows laptop when I already own a decent Mac laptop just to look at the polar response of a few speakers presented in a different visual format does not rate high on my "to buy" list, neither does purchasing Bootcamp (which still requires purchase of Windows to operate many programs), VMware Fusion, or Parallels.
That said, when my present Mac computers break down, I may be joining the PC world, the differences between PC/Mac user friendliness ain't what it used to be when I got hooked on Macs 27 (or so) years ago, Macs may no longer be worth the price premium and incompatibility issues anymore. That said, I really like being able to just speak into the computer and have it type what I say with the stock program that comes with the g5 computer.
Art
Posting PDF pictures of the various polar maps you have created (imported) would make viewing possible for those of us not using a Windows platform.
Art
Posting your polar maps in a PDF format would cost you nothing (other than a few minutes time) and widen your viewing audience and their potential interest in measurement technology.
Buying a Windows laptop when I already own a decent Mac laptop just to look at the polar response of a few speakers presented in a different visual format does not rate high on my "to buy" list, neither does purchasing Bootcamp (which still requires purchase of Windows to operate many programs), VMware Fusion, or Parallels.
That said, when my present Mac computers break down, I may be joining the PC world, the differences between PC/Mac user friendliness ain't what it used to be when I got hooked on Macs 27 (or so) years ago, Macs may no longer be worth the price premium and incompatibility issues anymore. That said, I really like being able to just speak into the computer and have it type what I say with the stock program that comes with the g5 computer.
Art
Last edited:
Windows does that as well. I have never used it, so I don't know how well it works, but it is there.
You're still using a PowerPC? Wow. I don't think any not-a-Macs would have lasted the dozen-odd years that machine lasted you.
As for Dr. Geddes' program, here's an interesting development that may bring it to people like you and me sometime in the future: Microsoft to Open Source .NET Stack, Expand Platform to OS X, Linux - Mac Rumors
My "new" 2011 Mac is a 2.5 GHz Intel Core i5, I forgot the name designation change from my (dead) g5.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.