Thanks dump truck. Looks a little closer. I'll try with my normal amp some other time. The t-amp is convenient, but not sure I can rely on it. I've had weird low frequency problems before as well with it. The mic is from Cross Spectrum, so that should be fine. I suppose I may have been a touch off axis above or below. Bottom line, you can see the response is smoother than the picture being painted here using Brad's measurement.
Thanks again. I don't have photoshop, so I appreciate it.
Thanks again. I don't have photoshop, so I appreciate it.
That's not how it works Tom. You need to pick a design philosophy and/or guru. From there the course of action is much clearer (e.g. start a crusade). Do what you believe is right, don't think about what is right. Way too stressful
This is way off-topic, but it's way more important too.
A few posts back, Tom mentioned the Moore tornado.
Well, I found out a way we can all help. A friend of mine runs a catering business near Moore and today he loaded up the truck with food and went onsite to give free food to anyone that asked. He is mostly feeding rescue workers and displaced residents, but he'll feed anyone that asks. Basically, he is just going out there every day feeding people until the food runs out.
So, if you want to help, maybe send lunch money as a donation. See the link below:
A few posts back, Tom mentioned the Moore tornado.
Well, I found out a way we can all help. A friend of mine runs a catering business near Moore and today he loaded up the truck with food and went onsite to give free food to anyone that asked. He is mostly feeding rescue workers and displaced residents, but he'll feed anyone that asks. Basically, he is just going out there every day feeding people until the food runs out.
So, if you want to help, maybe send lunch money as a donation. See the link below:
Re CD horns, in Sound Systems
I've yet to hear a CD horn that sounds natural, & is pleasing & comfortable to listen to, even for short periods ! Titanium diaphragms & CD horns = BAD, IMO
I think the DS is the problem too, & it affects @ least the upper Mid as well.
Closely matched dispersion between the Mid & Horn @ the Xover point is a worthwhile goal, in fact i'd say Essential.
A 5 db variation tested @ one location/room, will Not be a 5 db variation in another ! Of course it's nice to see flat-ish fR curves, but in Real rooms/buildings & outdoors, things change. Sometimes for better or worse. Flat-ish fR curves are however indicative of good engineering, but it's the end result that matters = Great sound quality/stereo & enduring listenability = Non annoying.
I've yet to hear a CD horn that sounds natural, & is pleasing & comfortable to listen to, even for short periods ! Titanium diaphragms & CD horns = BAD, IMO
I think the DS is the problem too, & it affects @ least the upper Mid as well.
Closely matched dispersion between the Mid & Horn @ the Xover point is a worthwhile goal, in fact i'd say Essential.
A 5 db variation tested @ one location/room, will Not be a 5 db variation in another ! Of course it's nice to see flat-ish fR curves, but in Real rooms/buildings & outdoors, things change. Sometimes for better or worse. Flat-ish fR curves are however indicative of good engineering, but it's the end result that matters = Great sound quality/stereo & enduring listenability = Non annoying.
I agree with you about titanium diaphragms. Some aren't too bad but most are a little bit too harsh for my tastes. Some are downright grating.
Also agree with you on the effects of the diffraction slot in the throat of a constant directivity horn. Earlier, we were talking about some popular CD horns with diffraction slots that created this kind of ripple.
I think that smooth amplitude response is very important, probably the most important single aspect with regards to sound quality.
We're talking about sources with nearly constant directivity, so the power response tracks the on-axis response. It can be argued that this means the on-axis response can be equalized flat, and in so doing, response is made smooth througout the pattern. And this is true - That's one of the features of constant directivity - Power response matches on-axis response.
The most obvious example is the required first-order EQ to conjugate mass-rolloff. There has to be a mechanism in place to provide this, be it electrical or acoustic, or a combination of the two.
But I'm not comfortable with a series of tank circuits in the crossover to equalize away periodic ripple. And if the designer only chooses one or two of the largest peaks, it still usually leaves two or three that are almost as bad. All of the examples we've looked at in this thread with periodic ripple have several peaks larger than 3dB between the peak and the adjacent vallleys. So using tank circuits to shave off multiple peaks isn't attactive to me.
And equalization of ripple conveniently side-steps the root cause, which is what caused the response ripple in the first place. Internal reflections are the cause of ripple in horns. So one might consider that even if you equalize the response of a CD horn flat, the disruption of wavefront propogation is still happening.
Also agree with you on the effects of the diffraction slot in the throat of a constant directivity horn. Earlier, we were talking about some popular CD horns with diffraction slots that created this kind of ripple.
I think that smooth amplitude response is very important, probably the most important single aspect with regards to sound quality.
We're talking about sources with nearly constant directivity, so the power response tracks the on-axis response. It can be argued that this means the on-axis response can be equalized flat, and in so doing, response is made smooth througout the pattern. And this is true - That's one of the features of constant directivity - Power response matches on-axis response.
The most obvious example is the required first-order EQ to conjugate mass-rolloff. There has to be a mechanism in place to provide this, be it electrical or acoustic, or a combination of the two.
But I'm not comfortable with a series of tank circuits in the crossover to equalize away periodic ripple. And if the designer only chooses one or two of the largest peaks, it still usually leaves two or three that are almost as bad. All of the examples we've looked at in this thread with periodic ripple have several peaks larger than 3dB between the peak and the adjacent vallleys. So using tank circuits to shave off multiple peaks isn't attactive to me.
And equalization of ripple conveniently side-steps the root cause, which is what caused the response ripple in the first place. Internal reflections are the cause of ripple in horns. So one might consider that even if you equalize the response of a CD horn flat, the disruption of wavefront propogation is still happening.
Last edited:
Great discussion!
Although i am not a huge horn fan i can honestly say that i havent found one that sounds 'right' to me. Perhaps its the diffraction slot since most ive heard used one. Biradials and annular ring types sounded better to me and to my limited experience of horns.
Whilst CD or at least controlled directivity is a worthwhile goal, not at the expense of amplitude ripple, a bit like high order dsp filtration (to my ears at least).
Funny thing is the short horn loaded ribbons i use, which are supposedly CD sound great AND 95dB/W. (awaits flaming for mentioning ribbons) however i suspect the lack of throat restriction, be it slots plugs or rings, is my personal justification for that (sorry no science behind that, just personal assertion of opinion)
Although i am not a huge horn fan i can honestly say that i havent found one that sounds 'right' to me. Perhaps its the diffraction slot since most ive heard used one. Biradials and annular ring types sounded better to me and to my limited experience of horns.
Whilst CD or at least controlled directivity is a worthwhile goal, not at the expense of amplitude ripple, a bit like high order dsp filtration (to my ears at least).
Funny thing is the short horn loaded ribbons i use, which are supposedly CD sound great AND 95dB/W. (awaits flaming for mentioning ribbons) however i suspect the lack of throat restriction, be it slots plugs or rings, is my personal justification for that (sorry no science behind that, just personal assertion of opinion)
Last edited:
Refreshing! A horn head owning up to temporal aberration inherent to horns. But also internal reflections of enclosures in general are sources of temporal aberration. Combat with analog methods is highly limited.
Effective control with DSP is feasible:
Here is how one of my two way speaker's measurements look:
Independent woofer and tweeter raw response measurements mounted:
An externally hosted image should be here but it was not working when we last tested it.
Based on responses, including complex temporal behavior, filters for EQ and crossover are made, and woofer and tweeter now measure as:
An externally hosted image should be here but it was not working when we last tested it.
This is how speaker measures with both drivers running:
An externally hosted image should be here but it was not working when we last tested it.
And this is how the speaker response looks when one of the drivers has leads reversed:
An externally hosted image should be here but it was not working when we last tested it.
Response recording of speaker playing 1kHz square wave:
An externally hosted image should be here but it was not working when we last tested it.
Response recording of speaker playing 66Hz square wave:
An externally hosted image should be here but it was not working when we last tested it.
Percussion of drum being struck. It is about 94ms of recording from CD. Top track is recording on CD, lower track is recording of speaker playing CD:
An externally hosted image should be here but it was not working when we last tested it.
This is 10ms zoom from above:
An externally hosted image should be here but it was not working when we last tested it.
Physical design is very important for good results. Drivers coincident to quarter wavelength through crossover region is key.
Tom Danley's designs are great candidates for advanced DSP.
Classic horn loaded box is not; horn has one set of temporal issues, box loaded driver another. Very properties desired from horn make it impossible to get it's acoustic center within quarter wavelength of box loaded driver's acoustic center through crossover region.
So true! This is something I find often overlooked in the DIY community, but it is really important in any speaker that is physically large but and has a midwoofer run to high frequencies. Standing waves inside the enclosure will develop, and they can adversely affect midrange response.
In fact, I am certain this is what makes some box speakers sound "boxy". Because when you take steps to put the sound sources (drivers/ports) in the right positions, anomalies from standing wave modes can be greatly mitigated. This combined with damping material spaning the cross-section in addition to lining the walls will do wonders.
It's really easy to see in measurements too. I've found 3-5dB midrange peaks in speakers with arbitrary driver placement. It's more common than not, unless one pays attention to it.
So the idea of designing solely with a box modeling program that uses (Helmholtz) cavity resonance and T/S parameters is not enough to give satisfactory results, in my opinion.
It isn't that the programs are wrong - they're almost always very accurate below 100Hz - but from about 200Hz to 500Hz, we need to look at the standing wave modes. Martin King's spreadsheets are great for this.
But no matter how much modeling is done, the final loudspeaker must be measured with drivers (and ports, if vented) in place. The insulation must be in place too. Only then can one know if the design is good.
Last edited:
I'm glad you mentioned that, Bill. I meant to respond to that comment but got busy and forgot.
This goes to a couple points I've made, some in forum discussions but mostly in private emails that then get presented in public by others, with varying amounts of "translation" by being paraphrased. It's not even really my argument, but rather one that comes from studying work from guys like Keele and Leach. That and from doing hundreds of computer simulations and dozens of horn and waveguide designs, and then measuring them and seeing these trends.
The first point is there is a myriad measurements from various sources that reasonably agree. I do not consider any of the charts presented in this thread thus far to have been unfairly "cherry picked" to mislead anyone. They are consistent and show true data, as should be obvious by the fact that they agree with one another.
The second point is a conical horn or waveguide has depth that is a function of wall angle. Without roundover, you can pick a wall angle and mouth area, and the depth is then known. So depth and wall angle are related and either one can be chosen to describe the device.
The secondary expansion (or roundover) does two things. It changes the pattern at low frequencies, mitigating waistbanding. And it increases mouth area, making it no longer fixed by the depth and initial wall angle. For a basshorn, increasing mouth area almost always makes response smoother because mouth area is undersized. But this is not the case for a tweeter, and may actually increase ripple.
I do agree with you generally, that both approaches have merit. It boils down to a set of competing priorities, so you pick your optimizations for what is most important to you. But there should be no disagreement on the data. Even simulations like from Hornresp show these trends.
There are off-axis charts of most models, but you're right, the cornerhorn models need current charts. The last charts I did of them were with Speaker Workshop, and I only did every 15 degrees. So I need an update. In fact, all models need new charts. This summer, hopefully I'll be able to do that.
The SEOS was specifically designed to reduce unwanted reflections from ceiling and floor bounce.
Wayne, I'm confused why your site clearly says your corner horns are "constant directivity" because this:
"reduces unwanted reflections from the ceiling and floor. This arrangement provides the most uniform directivity possible." And that the 'resulting sound quality from that type of set up is nothing short of breath taking'.
So you say constant directivity provides the most uniform directivity. Hmmmm. Seems like there's some paradoxical logic going on after reading this new thread. Or is it the chicken and the egg, who came first debate?
You seem to have changed your ideas a couple times to better suit the debate, which is the worst thing you can do in 'science', if I recall. If you plan on sticking with this new one, you might want to change your wording on your site.
In the end, I'm pleased to know that you think a constant directivity waveguide like the SEOS provides the most uniform directivity possible.
Wayne, I'm confused why your site clearly says your corner horns are "constant directivity" because this:
"reduces unwanted reflections from the ceiling and floor. This arrangement provides the most uniform directivity possible." And that the 'resulting sound quality from that type of set up is nothing short of breath taking'.
So you say constant directivity provides the most uniform directivity. Hmmmm. Seems like there's some paradoxical logic going on after reading this new thread.
Or is it the chicken and the egg, who came first debate?You seem to have changed your ideas a couple times to better suit the debate, which is the worst thing you can do in 'science', if I recall. If you plan on sticking with this new one, you might want to change your wording on your site.
In the end, I'm pleased to know that you think a constant directivity waveguide like the SEOS provides the most uniform directivity possible.
Hi Wayne
In reality, VERY VERY VERY FEW loudspeakers are array-able at least acoustically speaking (physically that’s easy).
If you array two or more of nearly any of them and play music or even better pink noise and walk the pattern, it is very clear they DO NOT combine, they interfere and have a very audible seam (try it).
To combine, one must have a horn shaped like the old Unity or Synergy horns AND have an acoustically small gap between the inner horn walls.
The need for constant directivity is partly because the listening plane is more than one or a couple seats and if one wants it to sound the same in every seat, you MUST have constant directivity.
In the case of the Synergy horns, often they are often used singly and not in arrays.
Also though, when one provides sound for a large space, a high directivity CD source properly aimed and positioned can supply sound that not only sounds the same everywhere (same frequency response) but also has a very small SPL variation (in some cases + - 1 or 2 dB) from the closest to farthest seat (much much better than any concert line array (who’s frequency response is also different in every location and why we have replaced so many of them in sports stadiums etc).
Best,
Tom
In reality, VERY VERY VERY FEW loudspeakers are array-able at least acoustically speaking (physically that’s easy).
If you array two or more of nearly any of them and play music or even better pink noise and walk the pattern, it is very clear they DO NOT combine, they interfere and have a very audible seam (try it).
To combine, one must have a horn shaped like the old Unity or Synergy horns AND have an acoustically small gap between the inner horn walls.
The need for constant directivity is partly because the listening plane is more than one or a couple seats and if one wants it to sound the same in every seat, you MUST have constant directivity.
In the case of the Synergy horns, often they are often used singly and not in arrays.
Also though, when one provides sound for a large space, a high directivity CD source properly aimed and positioned can supply sound that not only sounds the same everywhere (same frequency response) but also has a very small SPL variation (in some cases + - 1 or 2 dB) from the closest to farthest seat (much much better than any concert line array (who’s frequency response is also different in every location and why we have replaced so many of them in sports stadiums etc).
Best,
Tom
Tom, yes, that's one of the reasons I point out to some of the newbies for constant directivity horns being so important in prosound.
Scaper, I'm not sure I take your meaning. The constant directivity cornerhorn is designed for constant directivity, hence the name.
Do not be confused by the title of this thread and think I propose something other than controlled directivity. I've always chosen that approach.
What I am suggesting is the idea that a horn designed for prosound needs to be able to "play nice" in arrays, but a horn designed for home theater and high-fidelity does not. This allows slightly different optimizations. But I still would prefer uniform directivity in a home setting for spectral balance, seat-to-seat uniformity and imaging.
Scaper, I'm not sure I take your meaning. The constant directivity cornerhorn is designed for constant directivity, hence the name.
Do not be confused by the title of this thread and think I propose something other than controlled directivity. I've always chosen that approach.
What I am suggesting is the idea that a horn designed for prosound needs to be able to "play nice" in arrays, but a horn designed for home theater and high-fidelity does not. This allows slightly different optimizations. But I still would prefer uniform directivity in a home setting for spectral balance, seat-to-seat uniformity and imaging.
Ah, Scaper, I think I see what you're talking about now, so I think I can explain myself better.
The constant directivity cornerhorn provides truly constant directivity. It's one of the few configurations I know of that can do this. It kind of "cheats", in that it uses the walls to constrain the pattern at low frequencies, down to the point where room modes take over.
This is distinctly different than the pattern produced by a matched-directivity two-way speaker. A speaker that combines a direct-radiating midwoofer with a waveguide does not produce constant directivity, but rather provides collapsing directivity. Radiation is omnidirectional at bass frequencies, and beamwidth narrows up to the crossover point, where the waveguide takes over. This is not constant directivity, but if the transition is smooth, then at least the spectral balance is pretty uniform. It is a useful compromise, in my opinion, and sounds pretty good.
The constant directivity cornerhorn provides truly constant directivity. It's one of the few configurations I know of that can do this. It kind of "cheats", in that it uses the walls to constrain the pattern at low frequencies, down to the point where room modes take over.
This is distinctly different than the pattern produced by a matched-directivity two-way speaker. A speaker that combines a direct-radiating midwoofer with a waveguide does not produce constant directivity, but rather provides collapsing directivity. Radiation is omnidirectional at bass frequencies, and beamwidth narrows up to the crossover point, where the waveguide takes over. This is not constant directivity, but if the transition is smooth, then at least the spectral balance is pretty uniform. It is a useful compromise, in my opinion, and sounds pretty good.
Earlier in this thread Scraper was proposing co-developing a perfect 'uniform' waveguide. Don't the terms constant-directivity vs uniform-directivity vs matched-directivity really apply to fully integrated speaker systems, and not to just the waveguide component? Having said this, wouldn't you consider both the SEOS-12 and H290C to be constant-directivity in their passband?
Hello Wayne
I have question about how a corner horn is in fact constant directivity?? May be I am being a purist but if you put on in an anechoic chamber how would it look? I remember the corner horn that was at the MWAF, the first or second one, many years ago. It was a 15 in a reflex box with a triangle spacer to nest it into the corner with a 10" and a horn on top??
So essentially it was a 15 in a reflex box loaded into the corner with the corners walls as reflectors. Aside from flipping the pattern 180 degrees wouldn't it look like any other box in the anechoic chamber??
The point I am making is if it doesn't control directivity in an anechoic chamber how can anyone claim that it does in a room?? If that's the case I can take any 15' box, drop it in a corner and claim it's CD if the room walls are what is making it CD below where the driver size is dictating the directivity.
To me it just looks like any other 15" driver that's an omni down low and increases as you go up in frequency. So nothing out of the ordinary.
Am I missing something ??
Rob
I have question about how a corner horn is in fact constant directivity?? May be I am being a purist but if you put on in an anechoic chamber how would it look? I remember the corner horn that was at the MWAF, the first or second one, many years ago. It was a 15 in a reflex box with a triangle spacer to nest it into the corner with a 10" and a horn on top??
So essentially it was a 15 in a reflex box loaded into the corner with the corners walls as reflectors. Aside from flipping the pattern 180 degrees wouldn't it look like any other box in the anechoic chamber??
The point I am making is if it doesn't control directivity in an anechoic chamber how can anyone claim that it does in a room?? If that's the case I can take any 15' box, drop it in a corner and claim it's CD if the room walls are what is making it CD below where the driver size is dictating the directivity.
To me it just looks like any other 15" driver that's an omni down low and increases as you go up in frequency. So nothing out of the ordinary.
Am I missing something ??
Rob
Sorry a bit snarky and we don't need any more snark. Wayne stipulates that you need ~4 ft of wall (x,y,z) and ~6 ft of no obstructions (x,y,z) for his Corner Horns to perform optimally. The 'constraining' starting from just above the room nodes is part of the design. Anechoic would be breaking his 'speaker cabinet'?!?
What'cha breakin cabinets for, Mon?
I'm slowly building my wife up to some higher-WAF version of them ... slowly ... clear acrylic or 'plexiglass' may be involved ... slowly ... she's a decorator
... you guys have it so easy ...PS I believe you could use them outside or anechoic but you'd need what 6 ft (x,y,z) of 'wall', Wayne? I remember that guy ... hmmm, would 1" ply be enough? ... cool ....
Last edited:
Wayne, no offense intended, but that again sounds like circular logic.
You clearly say that constant directivity reduces unwanted reflections from the ceiling and floor. And that this arrangement provides the most uniform directivity possible.
This was your argument until the SEOS came to be. The SEOS was clearly designed to do exactly what you have always claimed would be the best.
So I'm pleased to know that you think that a great constant directivity waveguide like the SEOS provides the most uniform directivity possible.
I agree that the DIY Community did an incredible job with that design.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.