TPL 150-H at 2kHz is about -6db at 40 degrees, where the 6ND410 is -2db. (..looking at actual measurements.)
On top of that, there is the dependence on diffraction elements for both drivers relative to the baffle. Narrow baffles tend to diffraction gain around the width of the baffle as it relates to wavelength. That diffraction can have the effect of making the result LESS directive.
Say a baffle of 8" with a wavelength around 1700 Hz, along with a crossover at 2 khz, and a steep cutoff for each high and low pass filter.
The result then would be very non-directive up to 2 kHz, and then a serious drop in pressure more than 15 degrees off-axis as freq.s increase.
In other words:
-it would be bad (if you value uniformity off-axis), VERY BAD.
As an alternative with the same drivers: consider a larger baffle of at least 14" width, a crossover around 2.5-2.7 kHz, and a crossover that is MUCH less "steep".
Trying to get my head around this. And struggling, so was looking to simulators to help me understand, but also didn't get to one that seems to answer this. Which simulator would you recommend?
I was using Bagby's Diffraction and Boundary Simulator, but it models baffle diffraction response on axis and I think doesn't show dispersion or modelled behavior off-axis.
Nevertheless, I played with this simulator and assumed:
- 5.5" piston diamater driver (for a 6" midrange)
- 24" height baffle, with driver centered vertically (it would have more drivers above and below)
- 8 feet listening distance
- 1" baffle edge radius
- Simulated baffle width at 8" (blue line) and 14" (grey line), driver always centered
Attachments
You can get directivity from JB's diff &bndry sim but it takes a little extra work.
Look in the upper right hand corner of the big black control box of the top left of the screen. There you can set the observation point in off axis angles both vertically and horizontally. By looking at multiple angles and manually recording data....
Look in the upper right hand corner of the big black control box of the top left of the screen. There you can set the observation point in off axis angles both vertically and horizontally. By looking at multiple angles and manually recording data....
My perspective is experience while also looking at actual off-axis curves for each driver. So it's not a simulation at all, just an educated guess.
I'd probably recommend VituixCAD with actual Infinite Baffle 1 meter 2.83Vrms measurements for each driver.
Software
It does support FRD Tools files, so you can generate a grouping of axial measurements from both drivers and have V-CAD import them.
FRD Consortium tools guide
Note: I look at 1 meter results.. and only after that backing off and doing 2 meters (and not more than that).
I'd probably recommend VituixCAD with actual Infinite Baffle 1 meter 2.83Vrms measurements for each driver.
Software
It does support FRD Tools files, so you can generate a grouping of axial measurements from both drivers and have V-CAD import them.
FRD Consortium tools guide
Note: I look at 1 meter results.. and only after that backing off and doing 2 meters (and not more than that).
Last edited:
Interesting discussion. The only problem is that many of these measurements cannot actually be heard in the typical living room where they are placed due to the confabulation of furniture, rugs, objects, and walls and floors.
They can only be heard in an anechoic chamber.
Even if the line array isn't big enough to couple with the floor and the ceiling, there are still benefits of lowered distortion because of each speaker covering maybe only 1/16 of the total sound output, and flatter frequency response since most speakers have a much flatter response at 1/16 of the output at the typical 1 watt at 1 meter level.
And then we have all the problems brought on by using a passive crossover(many of which can be ameliorated by electronic crossovers(analog or digital).
And to top it off, we don't hear the same, and hear a lot less over 50 years old.
Reflections? It is well known(the hare effect(spel?)) that any sound that comes into the ears is mitigated by the auditory cortex, such that unless it arrives longer than about 40 millisec after the direct radiated sound, your brain simply ignores it. So you may not be technically in the nearfield, but your brain may effectively interpret it that way simply because of how it perceives sound.
They can only be heard in an anechoic chamber.
Even if the line array isn't big enough to couple with the floor and the ceiling, there are still benefits of lowered distortion because of each speaker covering maybe only 1/16 of the total sound output, and flatter frequency response since most speakers have a much flatter response at 1/16 of the output at the typical 1 watt at 1 meter level.
And then we have all the problems brought on by using a passive crossover(many of which can be ameliorated by electronic crossovers(analog or digital).
And to top it off, we don't hear the same, and hear a lot less over 50 years old.
Reflections? It is well known(the hare effect(spel?)) that any sound that comes into the ears is mitigated by the auditory cortex, such that unless it arrives longer than about 40 millisec after the direct radiated sound, your brain simply ignores it. So you may not be technically in the nearfield, but your brain may effectively interpret it that way simply because of how it perceives sound.
The Haas limit does indeed make us almost ignore the sounds after the first peak. At least it would seem that way in a glance. Yet it will change perception when and from where the reflection hits the listening spot. I wouldn't even state a fixed number for that Haas limit as it varies with frequency.
I use a so called Haas Kicker and use it for a couple of reasons. This later arriving energy (at 20 ms in my case, user adjustable) can still change the perceived tonal balance of that first wave front. Not the reason for me to use it, it's what it can add to the front stage that makes it interesting. I've often said it was one of my most successful experiments.
However the most important imaging queues are determined in our brain in the first few milliseconds. So keeping the first reflections in check is still important. Especially within the first ~6 ms. Personally I cleaned up the first 20 ms the best I could. As seen from the listening position it looks like this:
(showing off what the cool new REW release brings in terms of Spectogram)
With an array of mids you'd need to keep an eye on parallel planes to the array. The array itself tends to average out a lot of other reflections in the room. Each driver will have its own specific reflection pattern as seen from position of the listener's ear. As long as these unique patterns per driver differ from the other drivers in that array they will average out nicely. The ones they share need some help like using damping panels.
All reflections will change our perception. A good polar response can make those reflections be more even across a wider frequency spectrum. However one can also use absorption at first reflections to deal with it's unwanted coloration. Later lateral arriving reflections(*) can actually really add something positive to the front stage. It can even fill in part of the cross talk dips that stereo as a concept suffers from.
(*) most positive when it's decorrelated from the original wave front. Diffused in nature without high frequency spectrum above, say 3 kHz.
I use a so called Haas Kicker and use it for a couple of reasons. This later arriving energy (at 20 ms in my case, user adjustable) can still change the perceived tonal balance of that first wave front. Not the reason for me to use it, it's what it can add to the front stage that makes it interesting. I've often said it was one of my most successful experiments.
However the most important imaging queues are determined in our brain in the first few milliseconds. So keeping the first reflections in check is still important. Especially within the first ~6 ms. Personally I cleaned up the first 20 ms the best I could. As seen from the listening position it looks like this:
(showing off what the cool new REW release brings in terms of Spectogram)
With an array of mids you'd need to keep an eye on parallel planes to the array. The array itself tends to average out a lot of other reflections in the room. Each driver will have its own specific reflection pattern as seen from position of the listener's ear. As long as these unique patterns per driver differ from the other drivers in that array they will average out nicely. The ones they share need some help like using damping panels.
All reflections will change our perception. A good polar response can make those reflections be more even across a wider frequency spectrum. However one can also use absorption at first reflections to deal with it's unwanted coloration. Later lateral arriving reflections(*) can actually really add something positive to the front stage. It can even fill in part of the cross talk dips that stereo as a concept suffers from.
(*) most positive when it's decorrelated from the original wave front. Diffused in nature without high frequency spectrum above, say 3 kHz.
Last edited:
Do you think that for the purposes of only discussing listening in the nearfield, that since sound only travels at about 1.13 feet per millisecond, that in a small room with furniture and rugs etc, that the perceived nearfield, as opposed to the natural nearfield could be bigger? For a line array that is large enough to couple with the ceiling and floor, which Griffin said was about 70% of the height of the wall, this makes no difference. But for an array which is smaller than being big enough to couple, could the perception effectively make a listener still within the nearfield because the duration of the reverb from walls-floor-ceiling, haven't reached the Haas Limit?
I know its a weird question, and may be on the order of snake oil to think that.
I've said it before in this thread. I do not have experience with smaller arrays, I opted to go with the floor to ceiling array concept.
If I were to build smaller arrays (no interest to do so at this time) I'd be looking into power shading (possibly frequency dependent).
For the purpose as discussed in this thread I'd probably opt to use the tweeter without waveguide and go with a frequency depending shaded array, and also including woofers.
Something like WWMMMTMMMWW, actively filtered.
If I were to build smaller arrays (no interest to do so at this time) I'd be looking into power shading (possibly frequency dependent).
For the purpose as discussed in this thread I'd probably opt to use the tweeter without waveguide and go with a frequency depending shaded array, and also including woofers.
Something like WWMMMTMMMWW, actively filtered.
The last one I built was had the midrange from floor to ceiling. This one will be at the 70%(or so) of the height from floor to ceiling that Jim Griffin specified as the minimum for coupling the ceiling and floor. My other one was a three way with a 12 inch woofer, 17 3.5 inch full ranges, and 32 Dayton domes with cut flanges giving the CTC at .90 inches, per channel.
This one will be a 4 way, 12 inch woofer, 8 inch mid woofer, 18 2.5 inch full ranges, and 7 ribbons. As before I will be using active crosses, but instead of using 2 ported subs, I will be using 2 sealed subs with equalization. The subs can handle 250 watts RMS, and I have 375 watts RMS in that amp, so plenty of power for equalizing below 39 hz.
I love line arrays. Its so nice to play Toccata and Fugue by Bach and close my eyes and think I in a church with a real organ.
Thanks for your kind reply.
Eric
Last edited:
![20171006_133903[1].jpg](/community/data/attachments/594/594052-2a4adcadcc8591f5ca48b59563d0d425.jpg)
I'M ASSUMING THAT that ribbon on the side is your mid. What is that?
And the woofer line looks like it couples too. Although my experience with the other array I built was that having a 2.5 -3 octave mid range coupling was all that was necessary, since its the critical range of the most sensitive hearing(see Fletcher Munsson curves).
I can't afford to buy four more 8 inch mid woofs to add to the existing 2, but I wonder what an additional coupling of that range would do!?
What could be bad about having more mid-bass.
More seriously, it could really help out to spread the mid-bass like that.
That's why I suggested the WWMMMTMMMWW for the OP's goals.
If you are talking to me.... It would exceed the WAF requirements. I can have a small narrow line that goes above the height of the stereo cabinet at 32 inches(meaning 28 more or so to reach the coupling of 70% of the 84 inch height at that part of the room). Putting more mid woofs into the package would give me too large a speaker for the part of the room its in.
I'm struggling to get her to accept a small narrow part that extends to 60 inches in that part of the living room, and then there is the cost. Even if she would accept it, it would add another 100 bucks to the cost. Maybe if I didn't already have 2 DVC xmax 15.2 mm woofers there already.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.