Re: Comments on data
That's just human nature. I'm willing to wager that 95% of the "audiophiles" are mistaken when it comes to WHY their speakers sound good. They might know that they like them, and they may have some IDEAS why, but those ideas are typically off-base.
Here's an example of this. If you go to CES, you can listen to dozens of HORRIBLE sounding expensive speakers. Yet you can also find a handful of inexpensive ones that sound good. In my opinion, this is due to the fact that many of the expensive speakers are designed by retired dentists and lawyers. Ironically, the CHEAP speakers are typically designed by professional loudspeaker engineers.
The bottom line is that a real engineer can make a better speaker with cheap parts than a retired dentist with an unlimited budget can.
I'd estimate that at least half of the megabuck audiophile speakers are expensive "vanity" projects made by guys who retired from another profession (law, dentistry, accountants, etc...)
gedlee said:
That's just human nature. I'm willing to wager that 95% of the "audiophiles" are mistaken when it comes to WHY their speakers sound good. They might know that they like them, and they may have some IDEAS why, but those ideas are typically off-base.
Here's an example of this. If you go to CES, you can listen to dozens of HORRIBLE sounding expensive speakers. Yet you can also find a handful of inexpensive ones that sound good. In my opinion, this is due to the fact that many of the expensive speakers are designed by retired dentists and lawyers. Ironically, the CHEAP speakers are typically designed by professional loudspeaker engineers.
The bottom line is that a real engineer can make a better speaker with cheap parts than a retired dentist with an unlimited budget can.
I'd estimate that at least half of the megabuck audiophile speakers are expensive "vanity" projects made by guys who retired from another profession (law, dentistry, accountants, etc...)
Re: Comments on data
I'm guessing you didn't read my post very closely. I was suggesting that the LACK of power compression in the compression driver, as compared to the midwoofers, causes an upward spectral tilt at high volumes which makes the system sound bright or harsh.
I don't doubt that. I was alleging that in many (most?) speakers that use a horn, the dominant factor in the change of the tonality of the system from low to high volumes is the uneven power compression in the system. Recall that I "am left wondering how significant the level dependence of the perception of HOM can be" not if they are audible or if the perception is level dependent.
Umm? Power compression is measurable and does occur in the situation described. Upward tilted responses sound harsh. I'm a little lost as to what is inconsistent with existing proven facts. Perhaps you can point me to a comprehensive study of the relative magnitudes of the various factors affecting the difference in the perception of the tonality of typical multiway systems that incorporate a horn.
Again, I am not saying HOM aren't relevant or don't occur or any of that. I am alleging that the majority of the difference in tonality from low to high levels for many systems incorporating a horn are due to uneven power compression in the system and the corresponding upward spectral tilt. At the VERY minimum, this is certainly a major contributing factor (1-2 or more dB spectral tilt is quite audible).
gedlee said:
I'm guessing you didn't read my post very closely. I was suggesting that the LACK of power compression in the compression driver, as compared to the midwoofers, causes an upward spectral tilt at high volumes which makes the system sound bright or harsh.
gedlee said:
I don't doubt that. I was alleging that in many (most?) speakers that use a horn, the dominant factor in the change of the tonality of the system from low to high volumes is the uneven power compression in the system. Recall that I "am left wondering how significant the level dependence of the perception of HOM can be" not if they are audible or if the perception is level dependent.
gedlee said:
Umm? Power compression is measurable and does occur in the situation described. Upward tilted responses sound harsh. I'm a little lost as to what is inconsistent with existing proven facts. Perhaps you can point me to a comprehensive study of the relative magnitudes of the various factors affecting the difference in the perception of the tonality of typical multiway systems that incorporate a horn.
Again, I am not saying HOM aren't relevant or don't occur or any of that. I am alleging that the majority of the difference in tonality from low to high levels for many systems incorporating a horn are due to uneven power compression in the system and the corresponding upward spectral tilt. At the VERY minimum, this is certainly a major contributing factor (1-2 or more dB spectral tilt is quite audible).
Re: Re: Comments on data
Of course that's audible - that's why most horn audiophiles don't take some thing like the summa seriously. I'm not saying it sounds 'bad', I never heard it, what I'm saying is a true horn system with horn loaded midrange and bass will not have this dynamic skew.
Of course this system lacks efficiency compared to a horn system too. When I see how far the 'HOM' waveguide system is padded down with the foam and eq I think why not just use a good dome or ribbon?
Rybaudio said:
Of course that's audible - that's why most horn audiophiles don't take some thing like the summa seriously. I'm not saying it sounds 'bad', I never heard it, what I'm saying is a true horn system with horn loaded midrange and bass will not have this dynamic skew.
Of course this system lacks efficiency compared to a horn system too. When I see how far the 'HOM' waveguide system is padded down with the foam and eq I think why not just use a good dome or ribbon?
Re: Re: Comments on data
I did not say that your statements did or did not contradict the facts. I did not actually read your claims in detail, mostly because they were based on non-scientifc tests which I don't put much credance in. So I am not saying that your claims are right or wrong, only that it does appear that they could contradict the available scientific data. Like the fact that power compression WAS NOT audible in a test of some thirty people.
I think that you are guessing here and it is my belief that your guess is incorrect. I have measured the power compression and it is more consistant than you imply AND it takes an excessively high SPL for this to occur at all. So the real point is that the HOM are a factor at a much lower SPL than the power compression is, if it is at all.
Rybaudio said:
I did not say that your statements did or did not contradict the facts. I did not actually read your claims in detail, mostly because they were based on non-scientifc tests which I don't put much credance in. So I am not saying that your claims are right or wrong, only that it does appear that they could contradict the available scientific data. Like the fact that power compression WAS NOT audible in a test of some thirty people.
I think that you are guessing here and it is my belief that your guess is incorrect. I have measured the power compression and it is more consistant than you imply AND it takes an excessively high SPL for this to occur at all. So the real point is that the HOM are a factor at a much lower SPL than the power compression is, if it is at all.
Re: Re: Re: Comments on data
Here too I think that you need real data to support your opinion as it contradicts the real data that I have.
And you have your comparison of efficiency and power handling of a "padded down with foam" (3 dB!?) waveguide to dome tweeter way off. The "padded down" waveguide still has about 6 dB more efficiency than the dome and about 20 dB more power handling. And domes and ribbons ARE NOT constant directivity. Please try and get your facts right.
Magnetar said:
Here too I think that you need real data to support your opinion as it contradicts the real data that I have.
And you have your comparison of efficiency and power handling of a "padded down with foam" (3 dB!?) waveguide to dome tweeter way off. The "padded down" waveguide still has about 6 dB more efficiency than the dome and about 20 dB more power handling. And domes and ribbons ARE NOT constant directivity. Please try and get your facts right.
Re: Re: Re: Re: Comments on data
I don't have the summa or HOM horns you do so you provide the data. .What is the power compression of the 15 inch drivers in your HOM horns at 120db at 110 and 100? Now what is the the power compression of the HOM horn at the same levels?
All the horn loaded compression drivers I have experience with are over 110 db/w . if you mate your HOM horn with a 97 db woofer you are losing 12 db or more of sensitivity. I like to mate my compression drivers with 108 db bass horns, it's far more efficient and has less compression and has lower distortion. It also eliminates the padding you have to do to make your 15 linear below baffle reinforcement. How much sensitivity do you loose there? Another 3 db? That would make your speakers around 94 db/w if they are linear in the low mid and bass! In that case you are padding the HOM around 15 db! That's not very efficient.
Vifa makes a dome as well as others (I have an old EV here from the 80's) that is around the same or more sensitive then your HOM'd horns as well as there being plenty of ribbons with around 100 db sensitivity. IME they also have about the same power compression as a typical direct radiator 15 (such as in your HOM horns) so the whole system would be more linear as it's pushed compared to HOM's horns.
Your horn is not truly constant directivity either is it!?!
BTW- I'm not saying any one methed is better than another I'm saying I see no benefit of using your HOM horns in such a low sensitivity speaker.
gedlee said:
I don't have the summa or HOM horns you do so you provide the data. .What is the power compression of the 15 inch drivers in your HOM horns at 120db at 110 and 100? Now what is the the power compression of the HOM horn at the same levels?
All the horn loaded compression drivers I have experience with are over 110 db/w . if you mate your HOM horn with a 97 db woofer you are losing 12 db or more of sensitivity. I like to mate my compression drivers with 108 db bass horns, it's far more efficient and has less compression and has lower distortion. It also eliminates the padding you have to do to make your 15 linear below baffle reinforcement. How much sensitivity do you loose there? Another 3 db? That would make your speakers around 94 db/w if they are linear in the low mid and bass! In that case you are padding the HOM around 15 db! That's not very efficient.
Vifa makes a dome as well as others (I have an old EV here from the 80's) that is around the same or more sensitive then your HOM'd horns as well as there being plenty of ribbons with around 100 db sensitivity. IME they also have about the same power compression as a typical direct radiator 15 (such as in your HOM horns) so the whole system would be more linear as it's pushed compared to HOM's horns.
Your horn is not truly constant directivity either is it!?!
BTW- I'm not saying any one methed is better than another I'm saying I see no benefit of using your HOM horns in such a low sensitivity speaker.
Re: Re: Re: Re: Re: Comments on data
Your numbers are quite incorrect and your tone (and previous experiences with you) means that I choose not to continue this discussion.
"such a low sensitivity speaker" - you have got to be kidding!!
Magnetar said:
I don't have the summa or HOM horns you do so you provide the data. .What is the power compression of the 15 inch drivers in your HOM horns at 120db at 110 and 100? Now what is the the power compression of the HOM horn at the same levels?
All the horn loaded compression drivers I have experience with are over 110 db/w . if you mate your HOM horn with a 97 db woofer you are losing 12 db or more of sensitivity. I like to mate my compression drivers with 108 db bass horns, it's far more efficient and has less compression and has lower distortion. It also eliminates the padding you have to do to make your 15 linear below baffle reinforcement. How much sensitivity do you loose there? Another 3 db? That would make your speakers around 94 db/w if they are linear in the low mid and bass! In that case you are padding the HOM around 15 db! That's not very efficient.
Vifa makes a dome as well as others (I have an old EV here from the 80's) that is around the same or more sensitive then your HOM'd horns as well as there being plenty of ribbons with around 100 db sensitivity. IME they also have about the same power compression as a typical direct radiator 15 (such as in your HOM horns) so the whole system would be more linear as it's pushed compared to HOM's horns.
Your horn is not truly constant directivity either is it!?!
BTW- I'm not saying any one methed is better than another I'm saying I see no benefit of using your HOM horns in such a low sensitivity speaker.
Your numbers are quite incorrect and your tone (and previous experiences with you) means that I choose not to continue this discussion.
"such a low sensitivity speaker" - you have got to be kidding!!
Re: Re: Re: Comments on data
It wouldn't surprise me that in your system the power compression is fairly consistent, given the following:
1) You have a relatively low sensitivity horn, being 90 degree axisymmetric as opposed to 90x40 or 90x60 nominal and having a rising vertical directivity to keep the on-axis response relatively flat. Also, the plug eats up power.
2) The horn covers a lot of bandwidth- an octave more in the heart of things than if you, say, mated it to a 10 at 1.8k
3) You have a relatively sensitive midbass- not just the driver, but your baffle is wide as well keeping the low end up.
4) Your midbass probably has a 3" or 4" voice coil, as opposed to maybe a 2" or 2.5" on a smaller midrange that might be used in some other setups.
5) Your midbass driver does not do bass duties, at least I'm guessing that based on the measurements of the speaker (rolls off below 80)
I would contend that many of these factors are not present in a lot of horn systems, and that that leads to the uneven power compression. I did look into this a lot more than what you might guess based upon my rough description above; I can't sleep anyways, so let me dig up what I have for data and fire up Matlab to run a few of the old simulations I wrote when I was first messing with this stuff.
gedlee said:
It wouldn't surprise me that in your system the power compression is fairly consistent, given the following:
1) You have a relatively low sensitivity horn, being 90 degree axisymmetric as opposed to 90x40 or 90x60 nominal and having a rising vertical directivity to keep the on-axis response relatively flat. Also, the plug eats up power.
2) The horn covers a lot of bandwidth- an octave more in the heart of things than if you, say, mated it to a 10 at 1.8k
3) You have a relatively sensitive midbass- not just the driver, but your baffle is wide as well keeping the low end up.
4) Your midbass probably has a 3" or 4" voice coil, as opposed to maybe a 2" or 2.5" on a smaller midrange that might be used in some other setups.
5) Your midbass driver does not do bass duties, at least I'm guessing that based on the measurements of the speaker (rolls off below 80)
I would contend that many of these factors are not present in a lot of horn systems, and that that leads to the uneven power compression. I did look into this a lot more than what you might guess based upon my rough description above; I can't sleep anyways, so let me dig up what I have for data and fire up Matlab to run a few of the old simulations I wrote when I was first messing with this stuff.
Re: Re: Re: Re: Re: Re: Comments on data
My numbers are dead on and you know it- That's why you won't publish your power compression figures you claim to have.
You have no previous experience with me. We have NEVER communicated before other when I asked to buy some of your efficiency killing foam.
Your speakers are quite low sensitivity by my standards.
BYE!!
gedlee said:
My numbers are dead on and you know it- That's why you won't publish your power compression figures you claim to have.
You have no previous experience with me. We have NEVER communicated before other when I asked to buy some of your efficiency killing foam.
Your speakers are quite low sensitivity by my standards.
BYE!!
Re: Re: Re: Re: Comments on data
I am not talking about "other" horns or systems I am talking about mine.
The bass response goes down to about 50 Hz. I said that some of the data was invalid at the LF. One plot that I posted had LF nearfield measurements matched into the HF far field data for a full range measurement. This shows the correct LF response.
I do not see what "horn sensitivity" or "plug eats up power" has to do with power compression. If anything these factors would make things worse not better. (Lower efficiency systems tend to power compress more than higher efficiency ones.) The system is about 98 dB /2.83 volts which is very high (contrary to some other posters opinion). This means that the drivers do not come anywhere near their power compression point until about 118 dB, maybe a little less. But at any rate, this is an outrageous sound level - one that no one could stand for any length of time. If the system does compress here - and it does - who cares? At normal listening levels power compression is not a significant factor - albeit it is one that I am aware of and am looking into.
My point is simply that you have not shown any data to say that what we found; that power compression in compression drivers is not audible, is incorrect. In our study of the sound quality of compression drivers, we were at the thermal limit of the drivers (1.5 inch pro units) and there was no audible compression, although the THD and power compression were substantial. All of the data that I have seen says that power compression IS NOT the reason that horns sound bad at high output levels. It is the limiting factor at very high SPLs, but its not a significant audible problem at normal listening levels (unless your talking about discos, etc.) - in our systems that is. In a direct radiator system it more than likely is a major factor at normal listening levels.
At 18 inches, I consider the baffle to be pretty narrow.
Rybaudio said:
I am not talking about "other" horns or systems I am talking about mine.
The bass response goes down to about 50 Hz. I said that some of the data was invalid at the LF. One plot that I posted had LF nearfield measurements matched into the HF far field data for a full range measurement. This shows the correct LF response.
I do not see what "horn sensitivity" or "plug eats up power" has to do with power compression. If anything these factors would make things worse not better. (Lower efficiency systems tend to power compress more than higher efficiency ones.) The system is about 98 dB /2.83 volts which is very high (contrary to some other posters opinion). This means that the drivers do not come anywhere near their power compression point until about 118 dB, maybe a little less. But at any rate, this is an outrageous sound level - one that no one could stand for any length of time. If the system does compress here - and it does - who cares? At normal listening levels power compression is not a significant factor - albeit it is one that I am aware of and am looking into.
My point is simply that you have not shown any data to say that what we found; that power compression in compression drivers is not audible, is incorrect. In our study of the sound quality of compression drivers, we were at the thermal limit of the drivers (1.5 inch pro units) and there was no audible compression, although the THD and power compression were substantial. All of the data that I have seen says that power compression IS NOT the reason that horns sound bad at high output levels. It is the limiting factor at very high SPLs, but its not a significant audible problem at normal listening levels (unless your talking about discos, etc.) - in our systems that is. In a direct radiator system it more than likely is a major factor at normal listening levels.
At 18 inches, I consider the baffle to be pretty narrow.
Re: Re: Re: Re: Re: Comments on data
Then perhaps you should examine my original post, or better yet the one I quoted in that first post. The conversation started in the context of "conventional horn loudspeakers" and their tonality changes from low to high volume. My post was in response to Patrick Patrick Bateman's comment regarding the level dependence on the perception of HOM.
I stand corrected. Sorry.
I'm sorry that I haven't been clear on this. Let me restate again. I am suggesting that in a conventional horn system, the compression driver NOT exhibiting any power compression is part of reason for the level dependence of the tonality because at high volumes the woofer exhibits power compression and the compression driver doesn't so the response of the system is upward tilted and sounds bright/harsh. The reasons I listed above are reasons why your system is LESS succeptable to this imbalance. In the case of the horn, those factors would give more power compression (not that it would be significant- just an idea), bringing the horn closer to or below the midbass at high volume. Not that power compression is good, but I'd rather have a downward tilt than an upward one.
This is what I was getting at in the comment about your woofer's high sensitivity. If you look at some pro drivers with nominal sensitivities in the 94-95 range and then throw in baffle step, you run into a sensitivity where power compression is an issue at moderate to high, but still reasonable, listening volumes. This is not the case in your speaker because of the high sensitivity. That is what I was arguing.
I agree! That is here- in a normal horn/woofer system the compression driver is not compressing power and the woofer is which leads to an upward tilted response at high output levels. And like you said, in direct radiator systems good buy dynamics. I'll keep my domes for computer speakers
I was under the impression that it was wider, like 24, but again I stand corrected.
I don't think we are disagreeing here; for the most part I think we are just not communicating well. I'll try to be more clear in the future.
Between this post and the other one I have been typing for some time. I don't know that I'll get all the information up tonight that I was talking about before. Certainly though none of it is claiming that your speaker has a problem with power compression. I'm leaving town in a couple hours and won't be back until Sunday so it will be a few days.
gedlee said:
Then perhaps you should examine my original post, or better yet the one I quoted in that first post. The conversation started in the context of "conventional horn loudspeakers" and their tonality changes from low to high volume. My post was in response to Patrick Patrick Bateman's comment regarding the level dependence on the perception of HOM.
gedlee said:
I stand corrected. Sorry.
gedlee said:
I'm sorry that I haven't been clear on this. Let me restate again. I am suggesting that in a conventional horn system, the compression driver NOT exhibiting any power compression is part of reason for the level dependence of the tonality because at high volumes the woofer exhibits power compression and the compression driver doesn't so the response of the system is upward tilted and sounds bright/harsh. The reasons I listed above are reasons why your system is LESS succeptable to this imbalance. In the case of the horn, those factors would give more power compression (not that it would be significant- just an idea), bringing the horn closer to or below the midbass at high volume. Not that power compression is good, but I'd rather have a downward tilt than an upward one.
gedlee said:
This is what I was getting at in the comment about your woofer's high sensitivity. If you look at some pro drivers with nominal sensitivities in the 94-95 range and then throw in baffle step, you run into a sensitivity where power compression is an issue at moderate to high, but still reasonable, listening volumes. This is not the case in your speaker because of the high sensitivity. That is what I was arguing.
gedlee said:
I agree! That is here- in a normal horn/woofer system the compression driver is not compressing power and the woofer is which leads to an upward tilted response at high output levels. And like you said, in direct radiator systems good buy dynamics. I'll keep my domes for computer speakers
gedlee said:
I was under the impression that it was wider, like 24, but again I stand corrected.
I don't think we are disagreeing here; for the most part I think we are just not communicating well. I'll try to be more clear in the future.
Between this post and the other one I have been typing for some time. I don't know that I'll get all the information up tonight that I was talking about before. Certainly though none of it is claiming that your speaker has a problem with power compression. I'm leaving town in a couple hours and won't be back until Sunday so it will be a few days.
Re: Re: Re: Re: Re: Re: Comments on data
Oh brother.....
Duh. How hard is that? I got it from your original post and thought it a very good hypothesis. Seemed pretty clear to me - and in line with what I've heard. Woofer compresses, horn doesn't, FR tilts up.
So why all the misunderstanding?
Maybe it's time Doc Geddes got Glasses?
Oh brother.....
Rybaudio said:
Duh. How hard is that? I got it from your original post and thought it a very good hypothesis. Seemed pretty clear to me - and in line with what I've heard. Woofer compresses, horn doesn't, FR tilts up.
So why all the misunderstanding?
Maybe it's time Doc Geddes got Glasses?
"Duh. How hard is that? I got it from your original post and thought it a very good hypothesis. Seemed pretty clear to me - and in line with what I've heard. Woofer compresses, horn doesn't, FR tilts up."
What levels do you listen at??? I would be hard pressed to ever run a reasonably efficient system to a point where power compression becomes significant in a home setting.
Rob
What levels do you listen at??? I would be hard pressed to ever run a reasonably efficient system to a point where power compression becomes significant in a home setting.
Rob
Robh3606 said:
Hee-hee. Good point!
But it's the peaks that count.When I went from some rather nice dual 6.5s to a pro 15", the difference was obvious. I just don't hear the 15 compress. Maybe it does, but it doesn't sound like it. (Not likely at the low, low powers I use). That's one of the things I like so much about it.
Anyhow, the FR tilt from a compressing direct radiator vs horn should be super easy to measure, right? Just do an FR plot at 1/2 watt and then at some much higher - but realistic - level.
If you wanted to get fancy, you could use an average signal level of -18dB with much louder bursts to represent the loud passages.
At any rate, the point is Geddes is going on and on about the power compression not being noticeable in horns and CDs, when that's exactly what Rybaudio was saying. Much ado about nothing, really.
panomaniac said:
Hee-hee. Good point!
I do see where I misunderstood the claims and there is not a big disagreement except that in my system the power compression for either the woofer or tweeter would not occur at any typical volume setting. So it is not an issue. But the HOM begin to be audible at about 90 dB SPL and above - quite a moderate level.
Remember that peaks in the music signal have little thermal effect as the thermal compression is more of a long term power issue.
jzagaja said:
Quite interesting. I am not sure what meaning a waterfall of group delay has, but it is interesting and it may be the HOM that are causing the "ringing" (if you can call it that). I'd have to think about this.
My thought was to substract out the minimum phase component and the time delay phase leaving only the non-minimum phase portion. Then through a Hilbert transform find the magnitude of this phase. Maybe thats not valid either. I have not looked into this too deeply.
group buy
Hello,
here is an update on the OS waveguide group buy list:
If the three who declared interest would also take 2 each the total count is 17,
pretty close to 20, far away from 100 -- but it's hollyday season in the us and europe.
The rules for a group buy so far:
"I should mention that at $250 I include an exactly matched foam plug and a driver
mounting plate. These are the same waveguides as used in the Summa. They are a perfect
match for the B&C 1" drivers." - Earl Geddes, Post #137
"As far as a "group buy", I'd only consider this if I got one payment and shipped to one
address. I could do $200 each for 20 and $150 each for 100." - Earl Geddes, Post #142
As I said before, I will not organize a group buy, somebody in the us should do that.
Also I wont be online alot in the next three weeks so I can't take care of the counting.
I'm not even sure if I can take part, a broken dishwasher and a broken car eat up my
potential diy Summas last week
martin
Hello,
here is an update on the OS waveguide group buy list:
PHP:
USERNAME WGs Location
Jazr ? us
Patrick Bateman ? us
pooge 2 us
Brett 2 au
Pallas 3 eu
cdwait 2 us
Rybaudio ? us
Total 11+If the three who declared interest would also take 2 each the total count is 17,
pretty close to 20, far away from 100 -- but it's hollyday season in the us and europe.
The rules for a group buy so far:
"I should mention that at $250 I include an exactly matched foam plug and a driver
mounting plate. These are the same waveguides as used in the Summa. They are a perfect
match for the B&C 1" drivers." - Earl Geddes, Post #137
"As far as a "group buy", I'd only consider this if I got one payment and shipped to one
address. I could do $200 each for 20 and $150 each for 100." - Earl Geddes, Post #142
As I said before, I will not organize a group buy, somebody in the us should do that.
Also I wont be online alot in the next three weeks so I can't take care of the counting.
I'm not even sure if I can take part, a broken dishwasher and a broken car eat up my
potential diy Summas last week
martin
gedlee said:
Is that measured at 1M from the mouth? And noticeable to someone close as 1M , or farther away? Sorry, you probably cover that somewhere. Feel free to point out the documentation again.
Just wondering about the throat. If the SPL is 90dB at 1M, then the sound pressure in the throat must be pretty high. How high should depend on the driver and waveguide.
How does what's going on in the throat relate to what we hear? Do big HOMs there translate into audible HOM in the listening space? In other words, is the the throat a significant influence on the sound because of the high pressures and small spaces there? Is the throat the major contributor to the HOM?
