I'm sure you know auto-tech in Poland? Their waveguides are extremely well built with metal flanges / mounting plates etc. I'm not aware of a way to print or route or in any way make a waveguide of that size in that quality for that kind of money.
Sorry, I meant more the first part of your sentence: "I still believe there are benefits to non axi-symmetric waveguides depending on the use and implementation". So could you show an example of such implementation where the non-axisymmetric shape would be really beneficial? I'm asking becaue I'm not aware of any.
How much is the SEOS-24? I don't see prices.
How much is the SEOS-24? I don't see prices.
Last edited:
Ah yes. Caveat, some of this is based on my specific needs, not necessarily universally applicable. Listening distance is only about 3 meters, so CTC distance is at some point an issue plus vertical real estate is limited as there also needs to be a large screen on top of the center speaker. Both of these contraints are specific to my situation and go against an 'ideal' speaker, but that's the environment they have to live in.
Then in principle I believe a narrower vertical dispersion than 90 degrees would be beneficial. Problem is how do you get there withouth making things worse, i.e. the dreaded pattern flip. This is of course always a matter of trade off as it's simply not possible to transition seamlessly from a symmetrical device (the woofer) to a non-symmetrical, no question.
I think there are two ways to at least mitigate the problem to a degree where it becomes a choose your poison question. One is Wayne Parham's approach of directing the nulls just outside of the waveguides vertical pattern, thus achieving an abrupt transition, but at least without much widening. I've never heard any of his speakers, but I like the approach in theory. The other (which I'm favoring) is to use dual horizontal woofers underneath the waveguide. That leads to even lower pattern control in the horizontal plane and a far from perfect, but in my opinion acceptable transition at about double the frequency in the vertical. This also leads to a total radiation area including the woofers which is about the same size as an axi-symmetric waveguide on its own.
The best solution would be a MEH based on your software (somebody just mentioned that as well a couple of pages back), but for the time being I favor the dual woofers underneath a non-symmetrical waveguide as my compromise of choice. Happy to be convinced otherwise though!
Then in principle I believe a narrower vertical dispersion than 90 degrees would be beneficial. Problem is how do you get there withouth making things worse, i.e. the dreaded pattern flip. This is of course always a matter of trade off as it's simply not possible to transition seamlessly from a symmetrical device (the woofer) to a non-symmetrical, no question.
I think there are two ways to at least mitigate the problem to a degree where it becomes a choose your poison question. One is Wayne Parham's approach of directing the nulls just outside of the waveguides vertical pattern, thus achieving an abrupt transition, but at least without much widening. I've never heard any of his speakers, but I like the approach in theory. The other (which I'm favoring) is to use dual horizontal woofers underneath the waveguide. That leads to even lower pattern control in the horizontal plane and a far from perfect, but in my opinion acceptable transition at about double the frequency in the vertical. This also leads to a total radiation area including the woofers which is about the same size as an axi-symmetric waveguide on its own.
The best solution would be a MEH based on your software (somebody just mentioned that as well a couple of pages back), but for the time being I favor the dual woofers underneath a non-symmetrical waveguide as my compromise of choice. Happy to be convinced otherwise though!
BTW, the rollback can be addeed to a non-axisymetric shape as well and then use it as free standing. So far I haven't even tried to play with that much in Ath as the round shapes work so well but it's a possibility.
Last edited:
Granted there's no 'best' in audio, I do think MEH's have two incredible benefits. a) freely selectable angle of CD throughout their entire passband or at least down to Schroeder and b) good timing / step response due to their (almost) concentric nature. No matter how perfect your step response is in a conventional multiway speaker, that's always just for one point in space. And large waveguides are even detrimental in that regard as many benefits as they may have in other respects.
OK, when it comes to opinions, impressions and ideas, I can tell you what I think would work best: a large free standing round OS-SE waveguide with 1.4" driver such as the HF1440, crossed over to a cardioid-like open baffle woofer around 500 - 600 Hz, supplemented by some subwoofer(s). I just can't imagine anything better at the moment.
Last edited:
I'm not sure how my inputs were particularly subjective or impressionistic, but I actually agree. 🙂
The cardioid-like midwoofer is just hard to get right, but I love the idea. Another option would be to do it like Genelec with the W371a, might be doable with a 500 - 600 Hz crossover and a lot easier to experiment with than with a physically generated cardioid.
The only drawback compared to a MEH is that the timing / step response is again only correct for one point in space. How much that matters with such a relatively low crossover point is another question.
The cardioid-like midwoofer is just hard to get right, but I love the idea. Another option would be to do it like Genelec with the W371a, might be doable with a 500 - 600 Hz crossover and a lot easier to experiment with than with a physically generated cardioid.
The only drawback compared to a MEH is that the timing / step response is again only correct for one point in space. How much that matters with such a relatively low crossover point is another question.
Well you showed no data.
As the phase/step response was shown many times to be virtually irrelevant to sound quality (at least it has been never shown otherwise, despite all the attempts), I don't put this aspect high on my priority list. Why do you think it is so important?
What data was I supposed to show? My inputs were purely conceptual. And I don't know how I could have made it more clear that I was not making absolute statements, but that from a set of compromises I would choose certain ones for a specific situation and I tried to make arguments as to why.
When it comes to the step response I wrote explicitely that its relevance is questionable, but it is interesting that pretty much all accounts I read from people about MEH state what stands out is the coherence and clarity which I find hard to attribute to anything else than their quasi coincidental nature.
When it comes to the step response I wrote explicitely that its relevance is questionable, but it is interesting that pretty much all accounts I read from people about MEH state what stands out is the coherence and clarity which I find hard to attribute to anything else than their quasi coincidental nature.
You explicitly wrote that the "good timing / step response due to their (almost) concentric nature" is an "incredible benefit". It may be my fault to deduce that you consider this at least relevant, if not important. When I ask why exactly, you say that's because people say it. Don't get me wrong but if this is not just a story-telling, than what is? There's simply no data to support that.
I haven't optimized for size just yet.
Here I have reduced diameter from 639mm to 542mm, with only very little change in the response:
so ok you've got two sim's showing little to no difference can we use the forum to get a consensus as to it's difference or even quantifiable perception to everyone...which should show whether or not it's significant?
Ha, touché. I'm actually not that gung-ho on step response, guess I shouldn't even have used the term. It seems I was trying to rationalize to myself the subjective benefits of concentric sound sources by some questionable metric and of course you immediately saw the flaw in the argument.
But there is something about a well designed coaxial loudspeaker (and I have no way to support that scientifically) that just feels 'right'. And I suppose I in someway hope (yes, hope) that a MEH can have the benefits without the drawbacks of conventional coaxials.
So I definitely got a little carried away with the "incredible benefits" stuff...
But ironically what so attracted me to the Tritonia and that design study you posted was that I immediately thought it might just be possible to use the subwoofer in the enclosure to make the front radiating woofer cardioid, which is what you suggested as being the best choice at the moment, just done digitally (hence the reference to the Genelec W371a).
But there is something about a well designed coaxial loudspeaker (and I have no way to support that scientifically) that just feels 'right'. And I suppose I in someway hope (yes, hope) that a MEH can have the benefits without the drawbacks of conventional coaxials.
So I definitely got a little carried away with the "incredible benefits" stuff...
But ironically what so attracted me to the Tritonia and that design study you posted was that I immediately thought it might just be possible to use the subwoofer in the enclosure to make the front radiating woofer cardioid, which is what you suggested as being the best choice at the moment, just done digitally (hence the reference to the Genelec W371a).
Hello Mabat!
Do You have any chance to measure a driver from BMS in the existing sand waveguide?
I have used (1" models) 4524, 4538,4540, 4544, 4548 (!), 4550 and 4552.
Would love to test the new one, 5530... So far the best ever is the 4552 and 4544 not too far behind.
Also, how about measurements using the PWT response as the reference . . .
Do You have any chance to measure a driver from BMS in the existing sand waveguide?
I have used (1" models) 4524, 4538,4540, 4544, 4548 (!), 4550 and 4552.
Would love to test the new one, 5530... So far the best ever is the 4552 and 4544 not too far behind.
Also, how about measurements using the PWT response as the reference . . .
No, I don't have any 1" BMS at hand.
What are the differences between them? Do you have all of them measured (under the same conditions)?
What are the differences between them? Do you have all of them measured (under the same conditions)?
Last edited:
