Yamaha offers the speaker in a wooden enclosure, but it's quite a bit more expensive.
The Yamaha enclosure isn't as inert as the Summas I used to own. But not too many boxes are, the Summas are dead like granite.
If one wanted to deaden the Yamahas further, it would be fairly easy. Just remove the amplifier from the back and add some deadening inside.
I'm not an expert on FIR, but as I understand it, it allows you to use steep slopes while still preserving good phase behavior.
Here's a frequency and phase measurement I did on it:
In this measurement, you can see the phase is pretty darn good. Over a span of six octaves the Yamaha has less than 45 degrees of phase rotation. Not too shabby! The 'stock' frequency response of the Yamaha exhibits a "U-shape", similar to what JBL used to do for decades. This will be easy to fix with a MiniDSP.
The Yamaha DSR-112 has been a great standard setting box. It's been the most often recommended <$1000 speaker on live-sound forums for at least a half-dozen years.
It's FIR utilization probably kicks in around 500Hz. If it were used any lower in frequency, latency would be unacceptable for live use. For a HPF when used with subs, it uses what measures like a standard IIR LR4 at 120Hz.
Patrick, I've never seen the DSR available in wood..could you point me to it? Thx.
There's a recently introduced DZR line that will probably phase out the DSR, but not many reviews yet, and currently $2-300 higher. It touts more and better linear-phase FIR tuning, as does nearly every new pro-sound box on the market. The DSR-112, which is already a very good buy, may become a hell of a great buy when Yamaha promotes the phaseout.
Ok, sorry to post to my post...
But I should have said, I've never seen the dsr-112 in more than one cabinet version....which I took to be plastic from this thread.
Which I mistakenly took.....Because you clearly said 'DXR' which I misread...my apologies.
Yep dxr = plastic. And dsr = wood.
The ones that I had were built like a surfboard:
very thick foam wrapped in carbon fiber
The baffle was ridiculous, something like four inches thick
IIRC, rigidity goes up exponentially, so a 4" thick baffle is sixteen times more rigid than a 1" thick baffle.
Earl would have to chime in here, but IIRC, the original (foam) construction was too labor intensive and he switched to a mold that was poured with some type of plastic.
Last edited:
Aaaah! I hadn't considered that! My wife has been complaining that there's a bit of a lag on movies. That must be the FIR filtering.
I see it that way too.
Same. I believe that the asymptote of the profile at any ways around the axis can be brought back from an ellipsoid at a finite point from the source. The OS profile can be drawn from that. The tricky bit seems to be coordinating the starting points for these profiles and incorporating the shape change. The mouth size for each can follow from the wanted low end. The roundover can probably be estimated. The horn should be symmetrical.
I estimated mine using an ellipse rather than an ellipsoid and I'm not sure of the accuracy of this. However, it produced a couple of positives. The sound was very similar to the round equivalent. I think I prefer the round version but there could be two reasons, one of them is that round was much easier to build accurately by hand.
Then the elliptical version cut the ceiling by a few dB. However this isn't enough to sway me on its own. There are other ways to deal with this. Making the waveguide elliptical doesn't make it shorter (vertically).
From what I can gather, it seems that latency would be a bigger issue with live, or video editing, than with movie viewing. I’m guessing the FIR is not using an inordinate number of taps, and all the phase aligning is around the crossover point, not the HPF, in order to minimize latency.
Or maybe your wife is more tuned into latency than average. Do you have other digital entities in your playback? Does it happen with other speakers, or only the Yammies?
Hi Allen
The bit I haven't worked out is the throat. Diffraction effects cause a kind of inversion, where the directivity needs to be narrow the throat can be wide, and vice versa. So for, say, a horn 80 H x 50 V we need the throat narrower horizontally. That's easy to understand, but for a rectangular(ish) pattern we need some kind of diamond shaped throat, to spread the diagonals wider. I don't know how to calculate this exactly. Some of the Klipsch horns have diagonal fillets near the throat, called "mumps" IIRC, and I suspect this is part of the reason they work well. Just to keep on the thread topic, I want to do this for a 1.5" throat, very similar to the OP's requirement. The Klipsch would be too expensive to ship to Australia, even the JBL is steep. Is there any interest in a home built horn of similar dimensions?
Best wishes David
Hello all I found a series of interesting posts about the throat diameter effect on HF dispersion:
Geddes on Waveguides
To paraphrase Geddes claims (and having checked out his data I would agree) that the HF dispersion is not a function of the throat diameter but of the waveguide and phase plug of the driver. He avoids using large drivers because they generally need a lot of EQ to bring the response up. So taking that into account and my desire to be able to do sustained high output for PA use the 4592nd is 2" is looking good.
Geddes on Waveguides
To paraphrase Geddes claims (and having checked out his data I would agree) that the HF dispersion is not a function of the throat diameter but of the waveguide and phase plug of the driver. He avoids using large drivers because they generally need a lot of EQ to bring the response up. So taking that into account and my desire to be able to do sustained high output for PA use the 4592nd is 2" is looking good.
Another quick question, if I were to get a 1.4" driver and mount it on a 1.5" throat horn would I have any problems?
Looking at the different throats of the BMS coaxes I think it may be a native 1.4" throat driver that for a 2" throat size a length of horn is added:
4592ND = 113mm body length (2" hole)
4595ND = 79 mm body length (1.5" hole)
4594ND = 79 mm body length (1.4" hole)
I have emailed them asking about this, so hopefully they can clarify.
Also the 2" Throat Horn Bolt-On 18"x10"For Assorted Bolt On 2"Exit Drivers 90°x 40° from ZXPC arrived and looks to be of good quality and has a smooth flare. I calculate that it should lose vertical pattern control at 2520 Hz and Horizontal at 626 Hz. I'm of two minds to use it for this project or to try and get something bigger as in my application I can't do the standard trick of using the directivty of a 12" or 15" woofer to bring the vertical lower in frequency.
The same calculations on the JBL 365359-001 give loss of Vertical at 1333 Hz and Horizontal at 370 Hz. So that's looking good apart from the possible issues with 1.5" throat.
I have also been looking at the SEOS-30 but I suspect its very expensive.
Looking at the different throats of the BMS coaxes I think it may be a native 1.4" throat driver that for a 2" throat size a length of horn is added:
4592ND = 113mm body length (2" hole)
4595ND = 79 mm body length (1.5" hole)
4594ND = 79 mm body length (1.4" hole)
I have emailed them asking about this, so hopefully they can clarify.
Also the 2" Throat Horn Bolt-On 18"x10"For Assorted Bolt On 2"Exit Drivers 90°x 40° from ZXPC arrived and looks to be of good quality and has a smooth flare. I calculate that it should lose vertical pattern control at 2520 Hz and Horizontal at 626 Hz. I'm of two minds to use it for this project or to try and get something bigger as in my application I can't do the standard trick of using the directivty of a 12" or 15" woofer to bring the vertical lower in frequency.
The same calculations on the JBL 365359-001 give loss of Vertical at 1333 Hz and Horizontal at 370 Hz. So that's looking good apart from the possible issues with 1.5" throat.
I have also been looking at the SEOS-30 but I suspect its very expensive.
Yes, been there. There is more than comparing the lengthening of the path via curved walls, to lengthening the paths with a throat plate type phase plug. For example when an OS profile widens the pattern past the source it appears to be guiding diffraction out in a way that it coincides with the wanted wavefront in an apparently fairly successful compromise.
I stuck with the latter for tweeter use, but seriously considered the former for below the crossover.
It is the opposite. The 4592 came first, and the 4594/4595 came later, and are truncated versions. I am surprised to see the 4594 and 4595 are the same length, as I always assumed that the 4594 was shorter, opened up at the same rate, and got to 1.4” sooner than the 4595 got to 1.5”.
The 4594/4595 have another upgrade over the 4592. The HF magnet is larger, and is outside the VC instead of inside the VC. The 4593 is a 1.4” driver that has the same magnets as the 4592.
Why not use the 4595 with the JBL instead of the 4594?
So I got a reply from BMS and they stated:
(I think 4504 is a typo and that's the 4595 as it would match other information on the web if so)
which is pretty much the same as what you said. Regarding my own project I was attracted to the 1.4" version as its a more standard size than 1.5" as its rather expensive to just buy drivers with every throat size. However I now have a new plan where I use the flare I have and a cheaper 2" driver to make a smaller about 1/4 cost version of my speaker to decide if the concept is good or not.
Hello James
Your link depends on the number of posts per screen, people use different values.
Can you link to a specific post, at the start of the discussion?
Best wishes
David
I'm now thoroughly confused about the subject of dispersion and throat diameter. My current thinking on this is that if special care is not taken the throat does narrow the pattern beyond one wavelength. However there exist approaches such as the attached equivalent throat that can push this frequency up to the point where its not an issue even with a 2" driver. Also if there is diffraction this can widen the pattern as well. The post I was trying to link to is: #7503
*I think probably the easiest thing will just to measure a 2" driver on the horns I have and see what dispersion pattern I get.