"How about a ladder delay network for the ribbon and put it flush with the mouth of the horn?'
Simple delay may not work alone with the ribbon above the midrange horn. In a horizontal orientation it may work because of the wide coverage angle of the ribbon in that orientation. It depends on what the DI of the midrange horn and the ribbon are, how far away you are and your seated height. You may have to aim the ribbon down into the horns sweet spot for it to integrate well. As it is you may have to be very close too sitting on the midrange horns axis because of that horns directivity. Can't be in two places at once.
Rob
Simple delay may not work alone with the ribbon above the midrange horn. In a horizontal orientation it may work because of the wide coverage angle of the ribbon in that orientation. It depends on what the DI of the midrange horn and the ribbon are, how far away you are and your seated height. You may have to aim the ribbon down into the horns sweet spot for it to integrate well. As it is you may have to be very close too sitting on the midrange horns axis because of that horns directivity. Can't be in two places at once.
Rob
I haven't thought through the method of mounting the ribbon, as you can see from the drawings. I'm thinking of a little wooden sled that would ride on the two horizontal pipes and also elevate the tweeter by several inches, and maybe have a tilt feature so it can aim downward at the listener.
My listening height is somewhere around 40~44 inches, if I recall right. The centerline on my Ariels is 36 inches, and I've found I have slouch into an uncomfortable position to get exactly on the centerline - in practice, I aim them upward just a bit.
When it comes to time-aligning, I have to agree with the cat in Boston. The various electronic schemes just don't work very well. I've heard a number of different ones (they're very popular in European high-end circles for some reason), and hifi wasn't the word that came to mind. John Atwood and I agreed that the sound was always better (usually a lot better) when these gadgets were switched out of the signal path.
Doing a time-delay right involves professional-grade 192/24 ADC/DAC conversion and professional-grade signal processing with extremely high quality, low-jitter master-clock synchronizing systems. But - if you have access to recording-studio-quality signal processing equipment, or even better, design your own, go for it! But don't kid yourself that a modified Behringer is sonically transparent - it isn't, and it is a long way from recording-studio quality.
I'd point the reader to my little Klipsch Chorus project where I replaced the horrible stock tweeter in the cabinet with an external horn supertweeter that sat on the top of the cabinet, almost at the rear edge.
Aside from the balance issues with the speaker - which I could have spent a little more time remedying until it was just right - the high frequencies were greatly improved. No more phasey and screechy Klipsch highs. Instead, it was airy and spacious, and actually had some 3D imaging with depth, the first time I'd ever heard that from a Klipsch loudspeaker. Crossovers and time alignment really matter, particularly with horns.
The weird thing about the RAAL tweeters is that they are indeed very directional in the vertical plane - enough so that they are effectively 3 to 7 dB more efficient at the listening position than the "so many dB at 1 meter" numbers would indicate. So some kind of gizmo to point the tweeter at the listening position is going to be necessary.
It may seem like I am taking a rather casual attitude to frequencies above 7 kHz. Well, I've played with sharp-cut lowpass filters that could be set to 5, 7, and 10 kHz, and there's not a lot of information above 7 kHz. Switch the lowpass in and out, and the sense of "air" and spaciousness changes, but not much else. The only instruments that change timbre are triangles and xylophones, if I recall right.
Peaks, though, are plenty audible, and draw your attention to this part of the spectrum in an undesirable way - peaks in this region give a sort of "chrome-plated", spotlighted quality to the sound that many magazine reviews mistake for "detail". It's not - it's just peaking, similar to turning up the "unsharp masking" in Photoshop too high.
Not that I want to give 7~20 kHz short shift - if I didn't care all that much, I'd go the usual prosound route of electrically boosting the HF in the same region where the diaphragm is breaking up and the phase plug is getting phasey. That way lie the peaks and edgy sound I want to avoid.
P.S. Thanks for the link, Agent5. Much appreciated, they look like useful gizmos.
My listening height is somewhere around 40~44 inches, if I recall right. The centerline on my Ariels is 36 inches, and I've found I have slouch into an uncomfortable position to get exactly on the centerline - in practice, I aim them upward just a bit.
When it comes to time-aligning, I have to agree with the cat in Boston. The various electronic schemes just don't work very well. I've heard a number of different ones (they're very popular in European high-end circles for some reason), and hifi wasn't the word that came to mind. John Atwood and I agreed that the sound was always better (usually a lot better) when these gadgets were switched out of the signal path.
Doing a time-delay right involves professional-grade 192/24 ADC/DAC conversion and professional-grade signal processing with extremely high quality, low-jitter master-clock synchronizing systems. But - if you have access to recording-studio-quality signal processing equipment, or even better, design your own, go for it! But don't kid yourself that a modified Behringer is sonically transparent - it isn't, and it is a long way from recording-studio quality.
I'd point the reader to my little Klipsch Chorus project where I replaced the horrible stock tweeter in the cabinet with an external horn supertweeter that sat on the top of the cabinet, almost at the rear edge.
Aside from the balance issues with the speaker - which I could have spent a little more time remedying until it was just right - the high frequencies were greatly improved. No more phasey and screechy Klipsch highs. Instead, it was airy and spacious, and actually had some 3D imaging with depth, the first time I'd ever heard that from a Klipsch loudspeaker. Crossovers and time alignment really matter, particularly with horns.
The weird thing about the RAAL tweeters is that they are indeed very directional in the vertical plane - enough so that they are effectively 3 to 7 dB more efficient at the listening position than the "so many dB at 1 meter" numbers would indicate. So some kind of gizmo to point the tweeter at the listening position is going to be necessary.
It may seem like I am taking a rather casual attitude to frequencies above 7 kHz. Well, I've played with sharp-cut lowpass filters that could be set to 5, 7, and 10 kHz, and there's not a lot of information above 7 kHz. Switch the lowpass in and out, and the sense of "air" and spaciousness changes, but not much else. The only instruments that change timbre are triangles and xylophones, if I recall right.
Peaks, though, are plenty audible, and draw your attention to this part of the spectrum in an undesirable way - peaks in this region give a sort of "chrome-plated", spotlighted quality to the sound that many magazine reviews mistake for "detail". It's not - it's just peaking, similar to turning up the "unsharp masking" in Photoshop too high.
Not that I want to give 7~20 kHz short shift - if I didn't care all that much, I'd go the usual prosound route of electrically boosting the HF in the same region where the diaphragm is breaking up and the phase plug is getting phasey. That way lie the peaks and edgy sound I want to avoid.
P.S. Thanks for the link, Agent5. Much appreciated, they look like useful gizmos.
If not mentioned earlier, I'll be measuring and auditioning the Lambda TD12M and TD15M midbass drivers, along with the Altec/GPA 414-16 Alnicos, and the 18Sound 12NDA520's. Out of those four drivers, a favorite should emerge. Sonics and ease of crossover design are at the top of the list.
Now to get the HF frame and bass panels built - I'll be looking up the baffle-opening diameter, overall diameter, bolt-circle diameter, and flange depth of these four drivers.
Now to get the HF frame and bass panels built - I'll be looking up the baffle-opening diameter, overall diameter, bolt-circle diameter, and flange depth of these four drivers.
Lynn Olson said:
Actually, that is the exact unit that that crazy cat used. I should have linked to this thread.
http://www.goodsoundclub.com/Forums/ShowPost.aspx?postID=4832
An externally hosted image should be here but it was not working when we last tested it.
Oh boy, we're in the soup now, we'll attract the Cat's attention. I was really hoping that wouldn't happen.
There's some very interesting points to the Macondo, particularly the attention paid to time-aligning. I don't know why horns are so picky about this, but they are - and much more so than direct-radiators. Many of the most serious and legitimate complaints about horns go away when the horn has a low-diffraction profile and the system has good time alignment. Very few commercial horn systems meet these simple requirements - styling usually seems to win out.
What's weird is that I agree with the Cat, and Dr. Geddes, about 2/3's of the time (not on the same issues of course). It's the other 1/3 that makes the sparks fly.
There's some very interesting points to the Macondo, particularly the attention paid to time-aligning. I don't know why horns are so picky about this, but they are - and much more so than direct-radiators. Many of the most serious and legitimate complaints about horns go away when the horn has a low-diffraction profile and the system has good time alignment. Very few commercial horn systems meet these simple requirements - styling usually seems to win out.
What's weird is that I agree with the Cat, and Dr. Geddes, about 2/3's of the time (not on the same issues of course). It's the other 1/3 that makes the sparks fly.
Re: It's not a Paragon - or a Parthenon
There are a number of issues here. I will address three -
First is the bass low mid baffle(s) are not extended to the top. This will present problems with weak reinforcement in the low range that eq won't properly fix. The speaker will sound thin compared to a proper baffle.
Second the alignment between the 12 and compression driver must be much closer then you show in your picture if you want the system to really 'gel' in the midrange. Because of the depth of the horn you choose you must compromise in either poor alignment between the compression driver and 12 or facing unwanted reflections from the 12 off the bottom of the horn.
Third the ribbon is going to reflect off the top of the horn. This may not be a major problem for you. Some people seem to have less sensitive hearing up there and some may actually like the scattering of the top octave off the top of the horn.
Lynn Olson said:
There are a number of issues here. I will address three -
First is the bass low mid baffle(s) are not extended to the top. This will present problems with weak reinforcement in the low range that eq won't properly fix. The speaker will sound thin compared to a proper baffle.
Second the alignment between the 12 and compression driver must be much closer then you show in your picture if you want the system to really 'gel' in the midrange. Because of the depth of the horn you choose you must compromise in either poor alignment between the compression driver and 12 or facing unwanted reflections from the 12 off the bottom of the horn.
Third the ribbon is going to reflect off the top of the horn. This may not be a major problem for you. Some people seem to have less sensitive hearing up there and some may actually like the scattering of the top octave off the top of the horn.
I wasn't thinking in terms an active ladder delay, but passive. Like the example in this project:
http://www.zaphaudio.com/ZD5.html
I don't know if it is practical with as large as an offset is needed with deeper horns.
http://www.zaphaudio.com/ZD5.html
I don't know if it is practical with as large as an offset is needed with deeper horns.
"any comments on what you see here?"
I have one comment on the draft design. I would not make the side panels parallel with each other, it will create resonances. It would be better to splay the side panels out a little bit such that they form a 105 degree angle to the front panel.
Also, the picture does not say anything about the thicknesses. I would make the panels at least 1 inch thick, and 1 1/4 or 1 1/2 inch thick would be better.
I have one comment on the draft design. I would not make the side panels parallel with each other, it will create resonances. It would be better to splay the side panels out a little bit such that they form a 105 degree angle to the front panel.
Also, the picture does not say anything about the thicknesses. I would make the panels at least 1 inch thick, and 1 1/4 or 1 1/2 inch thick would be better.
diffraction diffusion confusion
Hi Lynn:
As stated, non-parallel rear wings would be a plus, but another option is to do 'holey' panels. This way there's less standing waves, but you get the benefit of breaking up the cancellation pathlength, which should even out the dipole peak/dip setup.
I used this with the (short) rear support braces on my own current OBs, and they sound superb. Whether the effect was significant... well, damned if I know.
Hi Lynn:
As stated, non-parallel rear wings would be a plus, but another option is to do 'holey' panels. This way there's less standing waves, but you get the benefit of breaking up the cancellation pathlength, which should even out the dipole peak/dip setup.
I used this with the (short) rear support braces on my own current OBs, and they sound superb. Whether the effect was significant... well, damned if I know.
Lynn Olson said:
Lynn, to avoid horizontal lobing keep mid-woofer LP at around 500Hz
CONS
- a small format compressor driver wount like that - and
- its quite contradicitonal to the RAAL pattern
An externally hosted image should be here but it was not working when we last tested it.
Greetings
Michael
Lynn Olson said:
Its funny you should make such a statement.. I said something similar before, and I *think* its because of their combination of knowledge and particularly their polarity of "opinion" (..or rather that they perceive such to be fact - often to the detriment to all else).
Still, lots of good information from both sources if you can tolerate it.
Lynn, why do you stick with the 10"-12" mid-woofer idea anyway?
I have settled to two 6,5" having its break up at around 5kHz making a 2-2,5kHz LP XO realistic.
SPL of such an arrangement in OB can be around 120dB down to 300-350Hz at very low Doppler IM of ~ 1%
When split load is used like I proposed way back in that thread
"...split the SPL by putting the second driver in series and bypass it with a capacitor ..."
"...Even better of course is to do that "active" – feed any speaker with its dedicated amp and insert a simple two resistor / one capacitor network at the input of the amp's..."
there aren't any unsymmetric lobing issues whatsoever.
The second driver comes in at the lowest frequencies, actually to reduce cone movment to one half at the HP XO point only.
I doubt that a low weight mid cone blends less well than a horn with your Raal.
Greetings
Michael
I have settled to two 6,5" having its break up at around 5kHz making a 2-2,5kHz LP XO realistic.
SPL of such an arrangement in OB can be around 120dB down to 300-350Hz at very low Doppler IM of ~ 1%
When split load is used like I proposed way back in that thread
"...split the SPL by putting the second driver in series and bypass it with a capacitor ..."
"...Even better of course is to do that "active" – feed any speaker with its dedicated amp and insert a simple two resistor / one capacitor network at the input of the amp's..."
there aren't any unsymmetric lobing issues whatsoever.
The second driver comes in at the lowest frequencies, actually to reduce cone movment to one half at the HP XO point only.
I doubt that a low weight mid cone blends less well than a horn with your Raal.
Greetings
Michael
I was kind of board today so I whipped up a couple of big OB cabinet concept 3D models using four cone drivers for the 500-2500 or 500-4000 Hz range.
This is mostly for fun and for ideas. And to see what kind of shapes can be gotten away with for some large baffle OB speaker. The only criterion I had is that the baffle is no more than 48 inches wide and can be cut from a single sheet of standard size MDF or plywood.
This is mostly for fun and for ideas. And to see what kind of shapes can be gotten away with for some large baffle OB speaker. The only criterion I had is that the baffle is no more than 48 inches wide and can be cut from a single sheet of standard size MDF or plywood.
Attachments
Good Articles
Imperfect Sound Forever
and
Turn Me Up!, an industry initiative to restore dynamic range to CD's. Many good links here, including the article above.
As for the continued interest in big midranges - well, the answer is simple: Dynamics! I've been listening to pairs of 5.5" drivers for fifteen years now, and it's time for a change. A small midrange just doesn't sound like a large one, regardless of efficiency. I guess the word I'm reaching for is "presence" - a piano sounds more like a real piano, and less like a hifi trying to imitate a piano. More palpable, more "in-the-room", more real.
I think part of the reason horn speakers sound the way they do is the effective aperture of the horn/waveguide, which at lower frequencies is the same size as the horn-mouth itself. The 1.4" compression driver doesn't sound anything at all like a 1.4" dome midrange, but much more like a 15" driver, in terms of physical impact and sensation of presence. This is the secret of why a 15" Tannoy Dual Concentric sounds big. Well, in terms of effective radiating area, it is!
The distortion and headroom arguments have a lot of merit, but the special aspects of horns and big drivers are also apparent at very low listening levels, right down to a whisper - there's a sense of latent power, and of a certain type of sheer physicality to the sound. I strongly suspect this is due to the effective radiating area of the drivers themselves, or in the case of horns, the mouth area.
I do see a place for 5~6" midranges, and that's covering the range up to 4 kHz, where it can meet a small tweeter. The Ariels have a 3.8 kHz crossover with the LF/HF drivers phase-matched within five degrees (the reverse-phase null is 25 to 30 dB deep), and they sound like the scaled-up minimonitors they are. They cruise up to 105 dB without too much trouble, but I can hear compression setting in around 90~95 dB - with a subwoofer highpass filter engaged.
I've heard enough large-diameter midranges (Bastanis Apollo), front-horn Lowther/AERs, and all-horn systems to want to have that kind of engagement, presence, and physicality in my own system. But I don't want the annoying and gross colorations that have plagued high-efficiency systems for many decades.
With the work done by Dr. Geddes, Jean Michel Le Cleac'h, and Newell & Dr. Holland, we can move past the diffraction horns of the JBL Bi-Radial and the Altec Manta-Ray, and aim for much better time response, with an emphasis on rapid settling time. This emphasis on settling time is something I've focussed on since 1979, with my last speaker for Audionics, and intend to carry forward with the new design.
Other designers, who I respect, have more emphasis on directionality and the polar pattern, but I've always put greater weight on first-arrival time characteristics. The Ariel violates standard guidelines in terms of driver spacing and choice of crossover frequencies, but I'm quite satisfied with the crossover integration and spatial qualities.
Imperfect Sound Forever
and
Turn Me Up!, an industry initiative to restore dynamic range to CD's. Many good links here, including the article above.
As for the continued interest in big midranges - well, the answer is simple: Dynamics! I've been listening to pairs of 5.5" drivers for fifteen years now, and it's time for a change. A small midrange just doesn't sound like a large one, regardless of efficiency. I guess the word I'm reaching for is "presence" - a piano sounds more like a real piano, and less like a hifi trying to imitate a piano. More palpable, more "in-the-room", more real.
I think part of the reason horn speakers sound the way they do is the effective aperture of the horn/waveguide, which at lower frequencies is the same size as the horn-mouth itself. The 1.4" compression driver doesn't sound anything at all like a 1.4" dome midrange, but much more like a 15" driver, in terms of physical impact and sensation of presence. This is the secret of why a 15" Tannoy Dual Concentric sounds big. Well, in terms of effective radiating area, it is!
The distortion and headroom arguments have a lot of merit, but the special aspects of horns and big drivers are also apparent at very low listening levels, right down to a whisper - there's a sense of latent power, and of a certain type of sheer physicality to the sound. I strongly suspect this is due to the effective radiating area of the drivers themselves, or in the case of horns, the mouth area.
I do see a place for 5~6" midranges, and that's covering the range up to 4 kHz, where it can meet a small tweeter. The Ariels have a 3.8 kHz crossover with the LF/HF drivers phase-matched within five degrees (the reverse-phase null is 25 to 30 dB deep), and they sound like the scaled-up minimonitors they are. They cruise up to 105 dB without too much trouble, but I can hear compression setting in around 90~95 dB - with a subwoofer highpass filter engaged.
I've heard enough large-diameter midranges (Bastanis Apollo), front-horn Lowther/AERs, and all-horn systems to want to have that kind of engagement, presence, and physicality in my own system. But I don't want the annoying and gross colorations that have plagued high-efficiency systems for many decades.
With the work done by Dr. Geddes, Jean Michel Le Cleac'h, and Newell & Dr. Holland, we can move past the diffraction horns of the JBL Bi-Radial and the Altec Manta-Ray, and aim for much better time response, with an emphasis on rapid settling time. This emphasis on settling time is something I've focussed on since 1979, with my last speaker for Audionics, and intend to carry forward with the new design.
Other designers, who I respect, have more emphasis on directionality and the polar pattern, but I've always put greater weight on first-arrival time characteristics. The Ariel violates standard guidelines in terms of driver spacing and choice of crossover frequencies, but I'm quite satisfied with the crossover integration and spatial qualities.