Really nice setup!
Agreed completely. I think that center channels get a bad rap because too many people kludge them (toppled-MTM's, mismatches with the mains in terms of orientation and firing height).
I don't think that's an issue. I cannot recall hearing anything in which a matrix-derived center-channel system sounded worse than no hard center channel.
I think that it is quite apparent that the 1.4" throat is beaming a great deal at 8 kHz and above, especially when compared to my 1" devices. So the question is which is the bigger problem LF or HF directivity control - because its quite obvious that you cant have both.
I was amazed that your polars didn't show an effect from the BMS coax. They must have improved it considerably from the earlier ones that I tried. Theoretically I don;t see how you cannot have an effect at the crossover between the two different diaphragms no matter what kind of crossover you use.
I was amazed that your polars didn't show an effect from the BMS coax. They must have improved it considerably from the earlier ones that I tried. Theoretically I don;t see how you cannot have an effect at the crossover between the two different diaphragms no matter what kind of crossover you use.
I would imagine that the loss in pressure above 8 kHz is precisely *because* of the co-axial driver. Where the upper freq. portion of the driver doesn't even start to "kick-in" until 7 kHz.
Switching to the 18Sound ND1406A would likely display less loss in output at higher freq.s on and off-axis.
There is a lot of high energy midrange content in the additional lower octave of directivity control that the bigger waveguide provides. Voices and musical instruments sound very clear, distinct, and real.
My seats are within 22.5 degrees of the speaker axis. I don't notice the high frequency off axis roll-off at all.
I used a tiny bit of delay (0.04 ms) on the high frequency diaphragm to get the crossover performance shown.
I prefer your previous approach of sticking to the data and staying away from the subjective. The larger throat format clearly shows an increase in beaming when compared to the 1" (energy content is altogether another thing) and according to my experience the 2" should be worse yet. This all makes perfect sense, of course, and is why I stick with 1" throats. This then yields practical limits on the lower edge of the passband to just below 1 kHz or so because 1" drivers aren't going to go much lower than that. (The Summa gets down to about 800 Hz.)
And if you do want to talk subjective, then I don't find the higher crossover - 800 Hz versus 400 Hz - to be an audible problem either. Basically, I have looked at all the various throat sizes and concluded that the 1" was the best compromise. The mouth size is a different thing that is independent of the throat size. A larger mouth is always going to work better, but it appears from your data that there is diminishing return beyond my larger device.
And if you do want to talk subjective, then I don't find the higher crossover - 800 Hz versus 400 Hz - to be an audible problem either. Basically, I have looked at all the various throat sizes and concluded that the 1" was the best compromise. The mouth size is a different thing that is independent of the throat size. A larger mouth is always going to work better, but it appears from your data that there is diminishing return beyond my larger device.
Sticking to the objective data..
TRY LOOKING AT THE PLOTS WITH THE RADIAN 1.4" DRIVER!
The extreme loss in pressure you SUBJECTIVELY assume relates to the large diameter is not present to the degree it is with the co-axial BMS. (..and note the the Radian is rather limited in extension to about 15 kHz ..with a nasty resonance above 20 kHz not shown.) (..yes, there is some loss when compared to a 1" - but not a lot.)
Additionally, the spl-trace (EXCEPTING the high freq. loss from the co-axial), is overall *FAR* more uniform than that presented by the Suma.
Please give credit where it's due: strictly from what we have seen in this thread - it's a good bit better than the Suma (..at least for over-all uniformity off-axis). On the other hand (as Patrick noted), it's also good bit larger as well and *with* the coaxial driver does have the upper freq. loss in pressure (..though he could alter the gain on the upper freq. portion of the co-axial if he wanted to).
Last edited:
I believe that it is essentially impossible to achieve accurate reproduction of a waveform if it is divided by a crossover and sent to two drivers with significant physical offset of the radiation axes. With the larger waveguide that I am using, all content from 400 Hz on up radiates from a single point.
The point source coverage extended down to 400 Hz combined with the midrange directivity control leads me to prefer the larger waveguide. I hear the 400 to 800 Hz octave much more than the 8,000 to 16,000 Hz octave.
As a test one day when I had company over, I turned off the high frequency drivers (crossover at 7kHz). Nobody said a thing. When questioned, nobody had even noticed.
Sorry about the subjective comments, I should not have used them.
The point source coverage extended down to 400 Hz combined with the midrange directivity control leads me to prefer the larger waveguide. I hear the 400 to 800 Hz octave much more than the 8,000 to 16,000 Hz octave.
As a test one day when I had company over, I turned off the high frequency drivers (crossover at 7kHz). Nobody said a thing. When questioned, nobody had even noticed.
Sorry about the subjective comments, I should not have used them.
Inspires me to look more at horn systems for my own listening.
I agree with Scott - the Radian driver looks very good - only down 5dB at 45deg at 10k is very good. And the response across the horn's bandwidth is exceptionally flat. Exceptional performance! (too bad there so darn big)
And I'd like to agree with hulkss, I would guess the better directivity control between 400-800Hz (smack in the midrange) to be way more import than a few dB more drop above 10k. I've found errors in the high treble to be very forgiving. And hulkss, please don't apologize for the subjective comments - ultimately, audio is about the experience. The technical stuff just gets us there more readily.
Thanks for sharing your work!
Then how do you think the Danley Synergy horns work?
I guess the key word here is "significate distance".It is actually quite easy to get perfect phase alignment between two sources that are separated by a distance less than 1/4 wavelength. With a passive crossover, the low pass pole retards phase. This is why the lower frequency producing drivers are mounted more forward of the high frequency driver. It must be this way to get phase alignment. Another thing is the center of a driver's voice coil is not the apparent origin of its sound. The acoustic center of a driver can be several inches in front of the voice coil. All of this can be seen in the phase response measurements in the Danley SH-50. The patent application and white paper are very interesting to read. I've learned more about passive crossovers over the last year than I had known previously. Most of this is due to studying Tom Danley's designs.
Rgs, JLH
And for those of you how like big horns/waveguides from the low mids up - you might want to take a look at this:
http://www.hostboard.com/forums/altec-users-board/189416-802-8g-x-o-200-hz.html 36x26" mouth
With a 1" driver!
http://www.hostboard.com/forums/altec-users-board/189416-802-8g-x-o-200-hz.html 36x26" mouth
With a 1" driver!
I would *imagine* that as well.. naturally never having heard something extend that low with that type of off-axis response. In fact has Earl? If not how can this be stated?
:Anyway - a really nice rig hulkss!
(..and like cuibono, no need to apologize for a subjective comment - most appreciate it here! Earl doesn't, but would you know that? ..and he isn't shy about using favorable comments for advertising.
)
Last edited:
I said "significant physical offset of the radiation axes".
The Synergy and similar "Danley" horns are designed to have zero offset of the radiation axes between low, mid, and high sections of the horn. The drivers work together to act as a point source.
The Summa drivers are about 16" apart on the baffle. This is over one wavelength at the crossover frequency. Not at all a point source.
Here's a picture of one of my TD-2's with the cabinet back off.
An externally hosted image should be here but it was not working when we last tested it.
I completely concur with this and pay very little attention to above 8 kHz.
I don't have any particular problem with what you say, except that the directivity at the crossover in your design is not likely to be well matched above and below, and I strive for this. I have found this difficult to do as the crossover point goes lower and hence for this reason, as well as the required physical size requirements for going lower, I avoid it. From a marketable system standpoint going lower than I do is untenable.
From a subjective standpoint any advantage to using a lower crossover point is also completely unknown. It is clear from the fact that what I do now is always considered by those who hear it to be "extrodinary" that there isn't significant evidence that the higher crossover has any problems.
You seem to have gotten to where you are by the desire to use a LF horn and I understand how this desire would lead to a lower crossover point and a larger waveguide requirement. But as I have often said, I am not a fan of LF horns, so I am not likely to find this approach desirable.
Again, I don't buy the "wavefront/point source" argument. There is a perturbation in the vertical directivity due to this spacing, and the data shows that it is not that significant (that is "new" data by the way and the Summa has a "new" crossover which works a lot better than the ESP15, which is four years old now). I have said over and over that I do not consider the vertical polar response to be a significant aspect of a design. I seek the best horizontal response and I achieve that. I achieve my design criteria, and it is that criteria where we differ. You have not posted polars for your system in full range - with crossover and hence not much can really be said until you do that.
Again, your system should not be compared to a Summa, because its not the same physical size, etc. It's design criteria is different. It may be marginally better (maybe not either), but I have found that improvments get smaller and smaller with physical size, as secondary things get involved - cost being a major factor. The point is that I am trying to make "sellable" systems and bigger is not in the cards.
I'll measure the system full range and post the results some day.
That is exactly correct, I like low frequency horns. I use subwoofers as needed to keep the horn sizes reasonable.
I agree about the design goals. Summa's are quite big as they are.
An externally hosted image should be here but it was not working when we last tested it.
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.