On second thought. If directivity didn't matter then room treatment wouldn't matter. Increasing directivity reduces the room... Room treatment (damping) reduces the room.... increasing directivity increases accuracy... increasing Room treatment increases accuracy. Spl at listening position is direct+indirect... increasing the direct portion only increases accuracy.
Almost forget what we were debating; Whats more potent to the increase in accuracy.... Diaphragm Material vs Directivity?
We would need anechoic measurements of the two competitors and then in room measurements. It was already said that Be and a well designed Aluminum diaphragm are more the same the different perception.
we also went into this; "The difference between say an exponential and a CD horn on the systems designed listening axis WRT first arrival, using the precedence effect, is going to be minimal."
If I were to stick to the first debate...I'd want to compare a be dome tweeter vs a waveguided Ti or Al Diaphragm. I think that waveguided drivers will have higher accuracy than a dynamic BE driver, in a room.
In the second part of our discussion, how can we set the premise? 60degrees vs...What? For my waveguide I might be looking at going from 60 degrees to 15 from 1000hz to 20khz.
Last edited:
Apologies. The driver used in the polar measurment of the K402 was Radian 950BePB.
The driver in the other horn I shared in comparison was JBL 2450SL (aquaplas coating).
I very much agree with you that we don't have much to back of audibility of HOMs and the importance of matching the exit angle. But I don't outrule it either. When you're comparing several different aspects at the same time, it's challenging to know which exactly matters and correlates to what we hear. And that's often the case.
Personally I believe there are more important reasons why the K402 isn't (at that's in my opinion of course) a stellar horn. In several areas though, it's very good.
I’m not sure why HOM’s are being mentioned at all.
Any potential audible HOM’s require very high Spl: rare for most in-home critical listening. Of course a poor coupling between the driver and horn can cause not just diffraction but also reflection and even high freq. resonance that can produce a shrill result.
Any potential audible HOM’s require very high Spl: rare for most in-home critical listening. Of course a poor coupling between the driver and horn can cause not just diffraction but also reflection and even high freq. resonance that can produce a shrill result.
This comparison doesn't make a lot of sense in real life in my opinion.
To get high resolution you need a controlled room and relatively close listening distance. A normal 1" dome works perfectly in such conditions. Add a wavegudie and you can fit HF distribution to your midrange and get a little more SPL. You can put midrange and tweeter close together -> smaller listening distance.
With a horn and compression driver you can get even further away as with the waveguide - but not like double the distance! You still need a controlled room!
You NEED to get further away cause of the large distances betwen drivers. You suddenly have tweeter membranes with 75mm and bigger instead of 25mm. You get resonances and often H2.
BUT you get "unlimited" dynamics and in have benefits in a shi??y room. Both can be very important and more important as the last bit of resolution >10kHz.
But you will never get best resolution - cause your room can't do it anyways.
That is not correct. Here is a page from an Earl Geede's presentation. You can look at HOM's as group delay. Higher SPL increases audibility.
Rob
I am not sure there are any research backing up any claim that you need "above critical listening levels" of SPL for HOM to be audible, and I am not that sure this is what Geddes meant either. The relative amount of HOM will be constant unless we are talking about very extreme situations where air is no longer linear.
HOM is not group delay in itself, but like any reflection/diffraction, it can be rather delayed sometimes. However, like with reflections, if the delay is significant, we tend to filter it as a separate source rather than as distortion, so this is not a linear relationship either.
We can probably look at the following as fact:
Here are some of my opinions:
HOM is not group delay in itself, but like any reflection/diffraction, it can be rather delayed sometimes. However, like with reflections, if the delay is significant, we tend to filter it as a separate source rather than as distortion, so this is not a linear relationship either.
We can probably look at the following as fact:
- HOM is any change in the wavefront that deforms it from being an ideal wavefront (originating from a single point, line or surface).
- We will have some level of HOM in any practical application as ideal sources do not exist.
- Varying levels of HOM can be determined by the loudness relative to the derived ideal wavefront, the delay compared to the first arriving part of the wavefront, and the angular deviation. But we can not scale audibility directly from this.
- HOM is a known audible factor outside the horn world. We know it from edge diffractions, reflections, lobing etc.
- HOM is not minimum phase.
Here are some of my opinions:
- In horns we need to pay special attention to this due to the scale of the horn compared to the wavelengts it is set to reproduce.
- Like with other types of distortion, the audibility of HOM will reach a sweetspot. Above that our ability to separate the main signal from distortion decreases. Below that, our ability to actually hear the distortion decreases. (Since Geddes has worked on harmonic distortion with his wife some time ago, I am pretty confident this is what he meant as well.)
- HOM is particularly audible on full range line array speakers with multiple small drivers. This aligns with reflections from dome tweeters within about 5-15cm from the dome. They are often extremely audible and can not med EQ-ed away.
- HOM seems to be a larger issue when it occurs early in the horn and has time to develop, while at the same time being contained in the horn and directed to the listener.
- Simulations show that HOM is commonly generated by rapid changes in cross section of a waveguide.
- The subjective impression can be several different things, depending on what the brain filter it as. A large delay and angular deviation can be experienced as a separate source. Shorter delay can be experienced as deviations in frequency response or lack of spatial focus.
This is interesting. You imply that you are measuring HOMs inside horns, like across the horn throat, etc. with a microphone. Can you describe your setup to do this? Do you have one that you still use?
Personally, I've only though about how this would be done. There is a PhD thesis that includes this subject (Richard C. Morgans, December 2004, University of Adelaide) that describes measuring HOMs (but I don't remember anything on the audibility of them).
I've found more recently that those using only simulations are not getting very good results, as evidenced by the "phasing plug adaptor" (sic) debacle that occurred about a year or so ago on these forums. I was told in no uncertain terms by some well known guys here that it "couldn't work", until of course the patent application was posted to USPTO and the measured (anechoically) results showed exactly what was described by the author/inventor. I don't trust simulation-only approaches, and that episode made me into a hardened cynic on that subject: one must measure--not simulate--what one is proposing.
Until we start to measure HOMs and correlate to personal listening, as a group of interested amateurs here, I don't believe this discussion is going to progress much further--even if some single member currently has a measurement rig that can handle this. It will take more than just one person that says it's audible or not, I firmly believe.
Also note: I'm not really the world's proponent of the K-402 horn. If I had the resources currently, I think I would go back and do the throat geometry variations and measurements that Roy D. only implied that he did. (As you might recall, even building test articles on patented gear in the U.S. is an infringement.) So he never published anything on what he found. But he implied there were big problems with the curved throat geometries (especially including oblate spheroid) in terms of their resulting polar coverage responses, something that the K-402 was apparently optimized to do as well as possible, and from my personal measurements it does it--up to a half wavelength across the 2" throat entrance (6.8 kHz). Above that frequency, the entrance throat diverging lens does its job (and well, I might add).
Building these large horns from scratch is difficult enough, and measuring HOMs is another difficulty, as well as proposing psychoacoustic studies that can separate what human hearing systems can hear and what sounds offensive or perhaps just detectable levels of HOMs by frequency and by degree.
Also note that the study would have to be sensitive to internally reflected HOMs that go back into the phase plug/diaphragm area and creates other issues, so the exact design of the compression driver (e.g., dome diaphragm, ring radiator, dual ring radiator, etc.) would have to be placed into the study, since you can't just wave your hand and say none of that makes any difference.
A lot of work yet to do. Perhaps I'll receive an inheritance soon from an unknown wealthy relative who expired and willed it all to me. I won't hold my breath, however.
Chris
Strongly agree. We need both HOM defining measurements, and then audibility studies of those.
Personally, I see HOM's as internal horn reflections...in all directions...like described a hundred years ago..
I have a hard time visualizing how they might vary by SPL. But maybe their audibility does...who knows.
I've tried to at least measure HOM's, horn reflections, that return from the horn back into throat, by using a coaxial CD as both the source and the mic.
Drive one section, and use the other as a mic. Works pretty well....I just don't know what to do with the data yet.
Vanes tend to be problematic for tailoring a wavefront, especially the more they need to achieve.
The OS profile has been found to widen higher frequencies (beyond the source size) with the least struggle.
Here is an example of vanes..
1. Forcing out a flat wavefront,
2. Beginning with a preferred wavefront shape.
The OS profile has been found to widen higher frequencies (beyond the source size) with the least struggle.
Here is an example of vanes..
1. Forcing out a flat wavefront,
2. Beginning with a preferred wavefront shape.
Last edited:
I was referring to the audibility of HOM’s - and according to Earl it is dependent on Spl.
This is easy to agree with, when it comes to accuracy and directivity, High directivity has virtually the same affect as taking a room, and controlling it. Both paths, reduce indirect sound. In my discussions in the past, psychoacoustics come into play, and that is still something I am learning and cannot repeat accurately what I was told in my thread without looking it up, but it has to do with what @Robh3606 brought up about First arrivals. I believe it was said that our brain can process out some of the indirect sound? The indirect sound becomes "air" as in spaciousness. What I believe is that regardless of it being perceived as spaciousness/ambience, masking is still happening if it can, meaning, if the decay spl and frequency, is near, matching, or higher than the direct energy, it can mask the direct energy. Because the low level details relationship to rest of the mix, is already facing masking issues, meaning, the low level direct sound is already facing masking issues in the face of the higher level direct sound, it doesn't take much more masking to dilute their existence. Whatever low level content that would have been there looking only at its direct energy, will be swamped out by the decay simply said. As we get higher and higher past the where we have high sensitivity, its already an uphill battle.
with this said, an anechoic chamber would give the clearest view of direct sound, regardless of the being impractical, it exposes the truth. How much meat you leave on the table is up to you.
Last edited:
I think this illustrates a bit of what we are struggling with. HOM is not reflections, it is deviation from a simple wave front. I see people asking for data, but it is not clear what type of data they are asking for. Some kind of agreement, or proff that HOM actually exist? IDK, but we know for a fact that HOM exist and we know what it is.
We do not have lots of data on what different types of HOM sounds like to a group of listeners who have agreed on a description, but this sounds to me a bit like when people mix blind testing with "proving what sounds best". We can do preference testing, but it is extremely hard when we can not change one parameter and be sure we do not change another in the process.
We could measure this by looking at the shape of the waveform arriving at a single point in space. That would tell us something about how the wave is affected at this point, but not the direction of the incoming wave.
Another way we could measure is by arranging an array of microphones outside the horn opening with an exact geometry, then measure the impulse response and compare all of them to see the actual shape of the waveform of the transient exiting the horn. That would show about the same as we get in a good simulation.
The question is, what are you looking for?
I like to use a totally different approach. I have made horns for 17+ years, and I have tested tons of different horns with many different drivers. I also have a fairly large group of people around me that are equally interested in the topic, and sometimes we even strongly disagree, but most of the time we find this to be an oportunity to dig deeper as our ultimate goal is the same.
But sometimes we stumble upon situations where tuning a system simply does not bring us any level of satisfaction. I have experienced this on many different speakers and in many different rooms. There can be numerous different reasons why this happens but where I most often find myself not being able to track it in my measurements is when I work with some types of horns, and when I work with wide bandwidth line arrays. It is like you are "almost there", but there is nothing left to fix. The sound just does not want to "release" from the speakers, and some frequencies seem to claim my attention without any obvious reason.
There are some types of speakers we can simulate and show that they generate different types of HOM. To me, the pattern is extremely clear. When I started working on tuning sound, I had no idea why something just did not want to sound good, while others were really easy to play with. I spent maybe 20 years of looking for something I had no idea what was. When I started learning about wave propagation, HOM and ways to simulate this, I finally started to see a pattern, literally. Where HOM was present, I also had these experiences.
Then we can say that "ok, why don't you show some proof? Proof of what? That I think stuff did not sound right? And why? So the judge can give his verdict? That is not why I work in this field. If we had a tool, then ok, I would love to have it and use it. But for now this is really simple. I have heard so many systems throughout the years that struggles without revealing why on basic measurements. I know for a fact that all of these are know cases for HOM. I also know that when I avoid these types of solutions, things tend to just work.
So if HOM is not the cause, the method still works. And I find it very likely that it is the cause, simply because it sounds exactly as bad as very close range reflections, and they are just as well hidden in the measurements, and we know for a fact that that IS HOM. So if it sounds like something we know for a fact is HOM, it shows HOM on simulations, and we have no reason to suspect anything else, then why do we even care?
Also thanks to @AllenB for showing what HOM can look like. All deviations in the pressure wave will cause pressure to start leaking in new directions. This is how HOM starts, and then it develops over time (microseconds). When a neck has a very abrupt change in expansion, the edge of the pressure wave is no longer contained and it will start leaking into the low pressure zones outside the pressure wave. This is what causes wide dispersion, and it is also what causes HOM. Any change in cross section will create some level of HOM, but a sharp edge with a huge difference in the profile angle is an absolute worst case scenario. We do not have to check if this is true, this has been measured and demonstrated as long as we have made loudspeakers. It is even being used deliberately in horns, because we know of its effect.
That is ok, but I am not sure it is correct, or correctly understood. Is he here to comment on it?
BTW, the thread is getting “de-railed”, we need to start talking about “best” compression drivers again (or their alternatives).
In that context the Beryllium vs Aluminum CD’s objective comparison here is interesting:
https://audioxpress.com/article/voi...udio-475bepb-8-1-beryllium-compression-driver
In that context the Beryllium vs Aluminum CD’s objective comparison here is interesting:
https://audioxpress.com/article/voi...udio-475bepb-8-1-beryllium-compression-driver
Last edited:
Well let’s hope so considering that he started this.
He started what? The theories of higher order modes?
Are you asking me why Geddes put foam in his horn?
The issue with home starts with the driver, so if we want to extend the discussion past diaphragm materials, we need this as part of the discussion as well. However, what happens further down a horn should be food for another thread.