@fluid I was commenting on JC:s posts #43 - 52, if there was an impulse or step response referenced in there I must have missed it and am really sorry.
What you are stating is self evident and no there’s no need to waste energy on arguing about it. I can only conclude we’re talking at cross purposes, must be a trait of mine since my wife and kids point this out to me all the time.
What you are stating is self evident and no there’s no need to waste energy on arguing about it. I can only conclude we’re talking at cross purposes, must be a trait of mine since my wife and kids point this out to me all the time.
First off, I am not a native English speaker... and I have never used the words "embellishing " or "nefarious". In fact the sentence you refer to was as subjunctive with "if" (i hope this is the correct translation). And my reference was solely your own data first shown by NicoB down to 300Hz and contour lines and you newly presented graph only down to 500Hz. This changes substantially the presentation of the same data.
And the paragraph regarding frequency domain and time domain was asked as a question! I still believe that the underlying math implies a linear system which may be the case but Klippel already showed in an extensive presentation the many non-linear components are present in a speaker.
All my discussions from were based on facts, partially assumptions. You posted here in the Yuichi thread and you advertise your product that it improves many aspects of the Yuichi. So it should be allowed to discuss about this or do you think that criticism about your products is not allowed if you directly state your product "improves" the discussed horn in this thread? Here is the text from your site:
How should a potential customer interpret this text? My interpretation would be that I get the same loading (or even better), the same horizontal and vertical directivity control (or even better) and the dispersion caused by the fins in the Yuichi is avoided ("no dispersion fins"). What has been improved?
What you accuse me of being unprofessional in the technical and personal aspect I reject in the strongest possible terms! I have designed so many horns and have solved the underlying math that I generally know what I am talking about or what I ask. And I have given so much to the community - for free...
Related to the discussion on throat adapters I've been working on bent dual driver adapters that go from round to rectangular with exponential area expansion. I look forward to the results of the side wall 'bumps' as my adapter/horn does not behave as well as I would like and perhaps this is related to wavefront distortion as I have prioritized exponential area expansion. Unfortunately its a bit tricky to extract this information from AKABAK...
On the subject of distortion measurements dynamic microphones work fine for at least low frequency distortion measurments.
On the subject of distortion measurements dynamic microphones work fine for at least low frequency distortion measurments.
Ok, I have no dog in this fight, and I sincerely hope that the discussion can be brought back to a civil and constructive exchange.
But I must say that your tone did (and still does) come across as rather aggressive.
I would tone it down.
Just my point of view.
Respectfully,
Marco
No worries, post #50 shows the files are ARTA .pir which are impulses.
My intent was not to argue but to try and explain a position that seemed obvious to me but not to you, so cross purposes.
I have had a lot of private communication with @docali which still continues, regardless of how any forum communication comes across he is a very knowledgeable, respectful and generous person. That has had it's own communication struggles but I am sure no malintent was ever meant. He is passionate about horns and sometimes that can lead to frustration. Anyone who has been involved in an audio forum for more than five minutes has probably felt it too.
Me too hopefully we can leave discussion of your commercial products aside. I don't ascribe anything untoward in the comments you have made but it is hard to fully separate commercial interest from general comment even if it is only in the mind of the reader.
Can you explain a little more what this means to you?
Apart from rearranging the fins for which data is presented, doing much more becomes quite a departure even though visually it might appear similar. Modelling of the fins is quite difficult as each channel really requires it's own subdomain, but subdomains aren't all created equally and results can vary so practical confirmation is important.
@docali has really pushed the envelope with what can be done with fin horn designs and I have seen the results of many different options.
One of the options that appeared during the development I really like the idea of and docali has released the full model for open use on his website.
https://sphericalhorns.net/2023/04/09/the-mk3b2-radial-fin-horn/
You have the CAD skills and CNC to make one and see what you think. Measure it and see how well the modelling holds up. I understand if you are reluctant but hopefully that doesn't stop you. I enjoy your measurements and explorations, lets see if we can come together despite the inauspicious start.
No offense trying to be projected here, but I haven't seen anything from Joseph Crowe ever being given for free, whether it was advice, dimensional drawings or CAD files. Coming in here and schmoozing into a thread, being a strictly for profit guy, isn't right. People like Mbrennwa have shared so much on here, which is extremely selfless and generous of him. Not saying it always needs to be like that, but for profit people should clearly identify their intentions first.
Considering how little the "Big Guys" disclose about their horns, particularly the underlying math, ANY open-source contributions are very welcome, from any contributor.
One of things that has seriously held back audio over the last century has been trade secrecy, with the actual research data trapped in corporate silos. Articles in the AES Journal are purposely vague and slide into marketing exercises, leaving the buying public knowing almost nothing but unintelligible reviewer-speak. This is particularly bad with horns and waveguides, leaving the curious audiophile in the dark.
This forum is one of the very few to be outside of corporate control, and the more data that is published, the better. We can legitimately argue about which set of measurements correlate best with the subjective experience ... that's been going on for at least a half-century, and will not resolve any time soon.
For example, I put greatest weight on rapid, resonance-free decay, and uniform diaphragm loading over frequency. Dr. Geddes, and others, put greater weight on a tailored dispersion pattern that is uniform with frequency. But most of all, we need data. I thank all of the contributors here for that.
One of things that has seriously held back audio over the last century has been trade secrecy, with the actual research data trapped in corporate silos. Articles in the AES Journal are purposely vague and slide into marketing exercises, leaving the buying public knowing almost nothing but unintelligible reviewer-speak. This is particularly bad with horns and waveguides, leaving the curious audiophile in the dark.
This forum is one of the very few to be outside of corporate control, and the more data that is published, the better. We can legitimately argue about which set of measurements correlate best with the subjective experience ... that's been going on for at least a half-century, and will not resolve any time soon.
For example, I put greatest weight on rapid, resonance-free decay, and uniform diaphragm loading over frequency. Dr. Geddes, and others, put greater weight on a tailored dispersion pattern that is uniform with frequency. But most of all, we need data. I thank all of the contributors here for that.
Last edited:
@Lynn Olson - I couldn't agree more in every aspect. Trade secrecy is a huge problem. I understand the worry of theft, but some consumers are wanting to know what they're getting for their hard earned money. Thats what patents are supposed to be for, but this country has so many loop holes in patent law that its impossible to rely on it alone to avoid theft, especially from the far east.
You hit the nail on the head regarding spectral decay vs coverage. Most pro sound applications put little emphasis on THD and CSD in favor of efficiency and coverage uniformity. Thats the difference between a hifi WG vs one designed for sound reinforcement.
You hit the nail on the head regarding spectral decay vs coverage. Most pro sound applications put little emphasis on THD and CSD in favor of efficiency and coverage uniformity. Thats the difference between a hifi WG vs one designed for sound reinforcement.
I'll agree to disagree for the sake of avoiding more clutter to this thread. I can msg you with details if you want an explanation.
A potentially big missed opportunity is however that he doesn't (to the best of my knowledge) offer any sample design for a diy persom to build and try out, like some other merchants do. This is a huge step in inspiring a diy consumer's confidence in other higher price products being offered.
With the fins I would expect the same kind of practical problems that emerged with the wave shaping device, and that's the real wavefront emanating from the driver. While it's certainly possible to optimize the fins for a coherent source in a simulation, an actual output of a real driver will be probably far from that. As a result different fin sections will be driven by more or less different amplitudes and phases which will get only worse down the horn...
Well things have evolved here ^^.
Energy is not free: A horn radiates energy according to his form, so a on-axis flat horn like the one of Joseph or a JMLC for example cannot be CD (Constant Directivity).
When we add a CD behavior we will create a bell form response, that is inevitable, physics laws.
What is pushed to the off axis is taken from on axis basically.
So a CD horn needs EQ, people that are allergic to DSP will not like this kind of horn. it’s just a matter of choice if you want CD behavior or a flat on axis response, but both not.
So Joseph Horn is not CD, it’s just a matter of choice.
In more, fins horns will always have a particular dispersion, they concentrate the +/-30 in constant SPL when on a regular horn it’s +/-15°, it’s due to fins behavior and cannot be reproduced otherwise and it’s what gives the natural in-room listening experience.
For linear response, if we talk about whether the global form of the raw response is linear or not, it’s not the point, a flat raw response just indicate that the horn is not CD, no more.
It’s the mid-range narrowing and beaming (accidents) that are important and it’s mainly relative to the mouth, rounded help but rounded with precise simulation (FEA helps a lot on it).
It’s very hard to remove the midrange narrowing completely without doing a very large horn. Midrange beaming yes. But it’s not just about rounded flare, it’s also a matter of acceleration/deceleration (Mabat will be agree with this
)
Then there is the finger response due to distance between fins exit, that is another “problem” that can be completely solved too.
For the dispersion angle, 90° it’s good for our in-room listening distance (2.5 to 5/6m), further we will have to reduce the coverage (80/70/60°), closer to open it to 100/110° as the monitoring speaker does, with a tweeter in a WG like Somasonus it’s good.
In a way yes, it's what I have done, very difficult to produce indeed. In some month I will publish a polar response of mine Bi-radial.
I will put the Docali one in FEA for see what's happen right now.
Energy is not free: A horn radiates energy according to his form, so a on-axis flat horn like the one of Joseph or a JMLC for example cannot be CD (Constant Directivity).
When we add a CD behavior we will create a bell form response, that is inevitable, physics laws.
What is pushed to the off axis is taken from on axis basically.
So a CD horn needs EQ, people that are allergic to DSP will not like this kind of horn. it’s just a matter of choice if you want CD behavior or a flat on axis response, but both not.
So Joseph Horn is not CD, it’s just a matter of choice.
In more, fins horns will always have a particular dispersion, they concentrate the +/-30 in constant SPL when on a regular horn it’s +/-15°, it’s due to fins behavior and cannot be reproduced otherwise and it’s what gives the natural in-room listening experience.
For linear response, if we talk about whether the global form of the raw response is linear or not, it’s not the point, a flat raw response just indicate that the horn is not CD, no more.
It’s the mid-range narrowing and beaming (accidents) that are important and it’s mainly relative to the mouth, rounded help but rounded with precise simulation (FEA helps a lot on it).
It’s very hard to remove the midrange narrowing completely without doing a very large horn. Midrange beaming yes. But it’s not just about rounded flare, it’s also a matter of acceleration/deceleration (Mabat will be agree with this
)Then there is the finger response due to distance between fins exit, that is another “problem” that can be completely solved too.
For the dispersion angle, 90° it’s good for our in-room listening distance (2.5 to 5/6m), further we will have to reduce the coverage (80/70/60°), closer to open it to 100/110° as the monitoring speaker does, with a tweeter in a WG like Somasonus it’s good.
In a way yes, it's what I have done, very difficult to produce indeed. In some month I will publish a polar response of mine Bi-radial.
I will put the Docali one in FEA for see what's happen right now
.
I scaled and perspective corrected the scanned page @fluid posted, overlayed the XY points from the table on the previous page. It's a lot closer than the hand drawn yuichi plans I tried previously but still seems a bit off. Also the 6 vertical measurements shown differ from the ones in the table..... I wonder if scaling and tracing the scanned graphic would be more accurate
The vertical profile and fins in the V2 CAD file are created from an excel calculator generating a CSV and a spline going through them, I am sure that these are accurate to how they should be for the intended design. The horizontal does not have a similar calculator and is therefore open to some interpretation. The values from one of the drawings goes not give an easy curve to match. Yuichi states in the information document that the drawings are indicative and not absolute, it is a demonstration of how a horn of this type can be built and not a definitive plan. There is no huge discrepancy between the two versions here. Either would work and produce something quite similar. The radiation impedance shows that they both work closely to the intended Hypex horn. The horizontal polar in post 1 shows you how the V2 surface model will perform (within the bounds of the predictive ability of BEM), I personally think this is closer to what was intended than can be had from religiously tracing the images.
