Project May on the Lansing forum might be of interest.
Attachments
Last edited:
I have never used two woofers so I don;t have a lot of experience. Multiple sources over the same frequency range will always create problems that we don't want so its best not to do that unless it is absolutely necessary, and I have not found this to be the case.
Vertical spacing would be preferred however as it is the horizontal directivity that is so critical, not the vertical.
KA is impy defined as twoPI*frequency * radius / c - anything else is not correct.
Both comb filtering and directivity stem from phase differences in sound path, so they are similar phenomena, but yet the two terms imply different things. Comb filtering usually applies to room reflections and not free field issues. Directivity is a free-field issue that does not depend on room reflections, although one can certainly argue thet they result in the same thing in the same situations.
The underlined statement is clearly true.
The use of a cap or a dome in theory should result in the exact same results. However, it can be difficult to know what the wavefront looks like on either surface and so one has to make some assumptions. These assumptions will generate differences if either one is not correct. For example, the flat case would have a great deal of phase and amplitude variation across it since the wavefront is not flat in that surface. The spherical dome would be more correct for a uniform velocity, but still assumptions about this wavefront may be in question, leading to erroneous results.
Most of the problem with diffractions slots are due to the large reflection that occurs from this interface. These reflections move back down the device to the diaphragm causing large peaks and dips in the response. For the most part the outer portion of the device has little effect on these problems. A longer "adapter" between the diaphragm and the diffraction slot will lower the frequencies of the resonance peaks creating more of them in the passband. Its best not to have any internal reflections in a waveguide or horn. Nothing can be done with diffraction that cannot be done without it, although a diffractionless waveguide will tend to be bigger that a diffraction horn for the same coverage.
This is, of course, a big issue with me. Horizontal aberrations lie in the same plane as our ears and as such are far more serious than vertical ones, almost to the point that we can ignore the vertical as long as we take great care with the horizontal. There are some issues with floor and ceiling bounce in small rooms, but I take care of those in the room, not in the speaker design.
No, I do not like them. All of my early speaks had them and they all had the same harsh sound. It was not until I made a waveguide and heard how clean a diffractionless device is that I fully understood the problems with diffraction slots.
Modern simulations can mitigate the problems with diffraction, but never eliminate them as long as there is an abrupt change in the wall contour. Diffraction is highly audible and all efforts to eliminate it will pay big dividends.
Loading, to me, is all but irrelevant. The length of the horn doesn't really affect loading nearly as much as the flare rate.
I think that this is completely the opposite. High directivity minimizes the very early reflections (VER) that are so detrimental to imaging. WIde directivity - many VER, narrow directivity - fewer VER.
I design for the highest directivity that I can get in a reasonable size. The higher and flatter the DI(f) the better. The NS-15 and the NA-12 have almost identical frequency response and DI(f), but the NS-15 DI(f) is a few dB higher and this makes it sound just a little better.
I somehow missed this whole conversation. It's the size and the smooth roundover. I use Euler spiral (clothoid) for the termination, typically from about a half of the depth of the horn outward - it turned out as a really convenient shape for this. Now I have so plenty of such results that I almost could make a proper study for the "axial issues" (as I'm quite busy these days I hoped someone else might do it since I released the tool but that seems not going to happen).
Originally I started to develop on non-axisymmetric designs but when I fiddled with different roundovers and horn sizes I quickly realised that I can still make quite perfect axi-symmetric waveguides. So now I'm kind of returning back to those as I still see some points in the axial symmetry in general and I believe it is a good thing, especially when got rid of these axial issues. It can be done!
Of course "it can be done", I've been doing this and known these relationships for decades.
In your models how do you specify the source velocities? These aren't really known exactly are they?
OceanWay's smaller monitor also has 2 pole filter. Impressing sounding speaker.
Pro2A – OceanWay Audio
Consumer version is this.
Ocean Way Audio Speakers
$7000, but you can build a pair with SB+Seas for much less than 10% of that.
It can be a great DIY project!
Last edited:
Now it's constant. Either the velocity or the acceleration (mostly). I don't do anything more.
Earl, for me it was a surprise (or a discovery in fact) - now I know you know it. It's hard to really appreciate without own experiments. That being said, I still haven't seen any of your waveguides performing as good as this (quite a small) one - to tease you a bit
Attachments
Now it's constant. Either the velocity or the acceleration (mostly). I don't do anything more.
Earl, for me it was a surprise (or a discovery in fact) - now I know you know it. It's hard to really appreciate without own experiments. That being said, I still haven't seen any of your waveguides performing as good as this (quite a small) one - to tease you a bit
To me that looks quite like my largest waveguide. I would certainly not say that it is any better, or worse.
Uniform velocity is a simplifying assumption that is not entirely correct. I have measured standing waves across a waveguide mouth resulting from the edge diffraction. This will not have a uniform velocity contour.
Last edited:
Marcel, thanks for joining in!
It felt a bit awkward posting your sim results and comments.
To me it was the first step, just to see the basic trends. Any ideas how to make the sims more realistic would be highly appreciated. I'm really not an expert in this - I can't even pretend. I just use what is readily available (ABEC) the best way I can...
Oh, maybe I misunderstood. As a source velocity I meant the wavefront at the throat. From that on the BEM simulation takes over. I make no other assumptions about the velocities anywhere else.
What you saw earlier (a cylinder or a cap) "in front" of the WG is just a way how ABEC organizes the calculations. It called "interface" and its actual shape shouldn't really matter much.
Last edited:
Would this still matter in case of an oversized waveguide, say: 1 meter in diameter for a crossover at about 800Hz?
Last edited:
For anyone's interest, this is how it looks like for a straight cone in a infinite baffle (as all the sims I made so far are for IB) -
Ro808: I make no assumptions about the velocity at the mouth - that is really calculated and I would guess it will be quite right. I only assume a flat wavefront at the throat.
Ro808: I make no assumptions about the velocity at the mouth - that is really calculated and I would guess it will be quite right. I only assume a flat wavefront at the throat.
Attachments
Last edited:
I'm glad you say that because this horn is only 374 x 351 x 130 mm.
Attached is the STL file. It's basically an OSWG with a bit of induced non-axisymmetry.
Attachments
