A recent project of is a pair of small bass reflex speaker cabinets for an old 4-wheel-drive pickup in which the previous speakers were trashed. I intend to use the cabinets on the floor on the passenger side, with the speakers in an up-firing arrangement. The necessary small size of the cabinets made a small driver also necessary. I chose the Markaudio CHN70, not only because it is inexpensive, but also because I have had previous experience with Markaudio drivers and like the sound they produce. The boxes were built with offcuts of plywood, which made certain compromises in box size necessary. I ended up with two boxes of 8.5 liters each, with rectangular speaker baffles of 8 inches by 16 inches. The driver center is at one-third of the length of the rectangle and centered on the width. The baffle edges are square (no bevels). The drivers received no break-in time.
To test the speakers, I placed them on a pair of old woofer boxes, the front baffles of which are each 27 inches high by 21 inches wide. This placed the centers of the CHN70s at 38 inches high, just below my seated ear level. The woofer boxes as well as the CHN70 boxes were toed in so the drivers were on axis with my ears. I cued up a stereo copy of the Beatles' "Sergeant Pepper's Lonely Hearts Club Band" to which I have listened for many years and know intimately. In this arrangement the CHN70s sounded far too hot in the upper midrange, so I then placed them on the floor in an up-firing position to audition them in the way that I intend to use them. I placed them in front of the toed-in woofer boxes with the CHN70 end of the rectangle pointed toward my seated listening position. The CHN70 drivers themselves were about 12 inches horizontally from the vertical panels of the woofer baffles. The centers of the woofer panels are about four feet from each other, and the CHN70 drivers were about three feet from each other. I was seated centrally between the speakers, with about three feet on a direct line between ear and driver.
Listening in this position, the upper midrange was tamed as I expected, since I was listening somewhat off axis. The bass improved also (no baffle step effect). These improvements acquired secondary importance for me, however, because with this arrangement I was surprised to discover that the speakers produced a phantom image that was centered in the prior on-axis position that I first used to audition the speakers. The sound did not come from near my right or left foot, but rather was coming from ear height at the left front and right front. I have never before heard this effect. I have heard of "holographic imaging", and perhaps it means other things to other people, but it is the best term I can produce to describe what I heard.
I considered possibilities as to the causes of this experience. I thought that perhaps the elevated upper midrange frequencies that I heard, to which the human ear is most sensitive, were reflecting off the vertical panels of the woofer baffles and were interfering with the direct sound of the CHN70 drivers, altering the phase. I also considered the room: it is a smallish room, rectangular in shape, 8 feet in width with 8 feet high ceiling. The center of the toed-in woofer boxes stand about 2 feet out from the back wall along the width of 8 feet. The CHN70 drivers were about 3 feet in front of the wall, and I sat facing this wall, perhaps 4 feet from it. The room is filled with diffusive elements such as shelves, books, and CDs, as well as some furniture. There is not much open floor space.
I don't know what, if anything, of value can be garnered from this experience, but I found it to be arrestingly interesting. I assume that others more knowledgeable than I have experienced this phenomenon, and perhaps even designed speakers using it. So, I am asking the experts: What did I hear?
To test the speakers, I placed them on a pair of old woofer boxes, the front baffles of which are each 27 inches high by 21 inches wide. This placed the centers of the CHN70s at 38 inches high, just below my seated ear level. The woofer boxes as well as the CHN70 boxes were toed in so the drivers were on axis with my ears. I cued up a stereo copy of the Beatles' "Sergeant Pepper's Lonely Hearts Club Band" to which I have listened for many years and know intimately. In this arrangement the CHN70s sounded far too hot in the upper midrange, so I then placed them on the floor in an up-firing position to audition them in the way that I intend to use them. I placed them in front of the toed-in woofer boxes with the CHN70 end of the rectangle pointed toward my seated listening position. The CHN70 drivers themselves were about 12 inches horizontally from the vertical panels of the woofer baffles. The centers of the woofer panels are about four feet from each other, and the CHN70 drivers were about three feet from each other. I was seated centrally between the speakers, with about three feet on a direct line between ear and driver.
Listening in this position, the upper midrange was tamed as I expected, since I was listening somewhat off axis. The bass improved also (no baffle step effect). These improvements acquired secondary importance for me, however, because with this arrangement I was surprised to discover that the speakers produced a phantom image that was centered in the prior on-axis position that I first used to audition the speakers. The sound did not come from near my right or left foot, but rather was coming from ear height at the left front and right front. I have never before heard this effect. I have heard of "holographic imaging", and perhaps it means other things to other people, but it is the best term I can produce to describe what I heard.
I considered possibilities as to the causes of this experience. I thought that perhaps the elevated upper midrange frequencies that I heard, to which the human ear is most sensitive, were reflecting off the vertical panels of the woofer baffles and were interfering with the direct sound of the CHN70 drivers, altering the phase. I also considered the room: it is a smallish room, rectangular in shape, 8 feet in width with 8 feet high ceiling. The center of the toed-in woofer boxes stand about 2 feet out from the back wall along the width of 8 feet. The CHN70 drivers were about 3 feet in front of the wall, and I sat facing this wall, perhaps 4 feet from it. The room is filled with diffusive elements such as shelves, books, and CDs, as well as some furniture. There is not much open floor space.
I don't know what, if anything, of value can be garnered from this experience, but I found it to be arrestingly interesting. I assume that others more knowledgeable than I have experienced this phenomenon, and perhaps even designed speakers using it. So, I am asking the experts: What did I hear?
I'm not an expert, but the same happens with any midbass even with LP @ 300hz, you can have them on floor firing normal like woofer not upward, and the stage is going to be in ear hight 
That effect is good for Desktop if you don't have place for bigger box but you want it louder, just place widerangers on desk and midbass on floor under the table, some toe up would be benefit, but again it's not even near optimum.
That effect is good for Desktop if you don't have place for bigger box but you want it louder, just place widerangers on desk and midbass on floor under the table, some toe up would be benefit, but again it's not even near optimum.
They sound hot because they are hot. There is a +9dB square top plateau from 600Hz to 7kHz built into the design. More info on this and other 3 in and 4 in drivers here.
An Objective Comparison of 3in - 4in Class Full Range Drivers
Try measuring the response if you can.
An Objective Comparison of 3in - 4in Class Full Range Drivers
Try measuring the response if you can.
Last edited by a moderator:
Hello Moderators,
Majerjack has already created the same thread in the Markaudio section which makes it a multiple.
xrk
Our on-baffle anechoic measurements/data are correct and were verified independently for/by the Japanese who contracted me to design this driver. Unlike you, they understand different testing set-ups produce different results for wide compliance drivers of the type Markaudio produces.
You’ve more than had your say. Its time to turn down your volume and stop turning your issue into a vendetta
Mark
Majerjack has already created the same thread in the Markaudio section which makes it a multiple.
xrk
Our on-baffle anechoic measurements/data are correct and were verified independently for/by the Japanese who contracted me to design this driver. Unlike you, they understand different testing set-ups produce different results for wide compliance drivers of the type Markaudio produces.
You’ve more than had your say. Its time to turn down your volume and stop turning your issue into a vendetta
Mark
Last edited by a moderator:
My post on this full range forum was done in an effort to discover what the possible reasons could be for the effect I experienced. My post on the Markaudio forum was done to share with others interested in the CHN70 the results of a quick and easy project and the favorable impressions I formed. I see the two posts as separate and distinct. I thought this full range post might draw more views than if it were put on the Markaudio forum, but if the Moderators think it would be more appropriate there, I have no objection. I only ask that the two posts are kept separate and unedited.
MarkAudio,
If you read my thread it is aimed at objective data which has now been confirmed to be correct for the 6 other drivers tested and now for the CHN by someone on a Japanese website who posted the same +9dB plateau from 600Hz to 7kHz. I worked really hard to verify my findings and took more data at different settings as suggested by other users. I, and many others on this forum agree that all we ask for is for factory measurements to reflect the product sold. I bought the CHN70 in hopes of finding a new full range driver that suited my needs based on the spec sheet response. Unfortunately, it turned out to be very different than advertised. I have no vendetta nor do I wish ill will on you or your company, so you need to stop attacking the tools and the users when the problem is obviously not the measurement.
If you read my thread it is aimed at objective data which has now been confirmed to be correct for the 6 other drivers tested and now for the CHN by someone on a Japanese website who posted the same +9dB plateau from 600Hz to 7kHz. I worked really hard to verify my findings and took more data at different settings as suggested by other users. I, and many others on this forum agree that all we ask for is for factory measurements to reflect the product sold. I bought the CHN70 in hopes of finding a new full range driver that suited my needs based on the spec sheet response. Unfortunately, it turned out to be very different than advertised. I have no vendetta nor do I wish ill will on you or your company, so you need to stop attacking the tools and the users when the problem is obviously not the measurement.
Did I read this right?
Vendetta?
English is not my native language. Does this word have another meaning other than I found in the dictionary?
We've both exchanged views between us on driver design/testing last year. You and I have conversed on the difference between testing which I felt was good at the time. You have to come to terms with the reality that for some driver products, where their operational design is mechanically complex, they will vary more when tested.
The Japanese web site to which you refer was part of the interaction for CHN between tests at that time (still going on). Hence why it refers the CHN's "character" and makes no negative comment about the driver itself. That site goes on to help end-users decide on box sizes and possible equalisation based on my early recommendations. Sadly google doesn't do justice to translation.
Its main reason why the CHN isn't published on the Markaudio's site......because its design and performance is open for input by Japan's forum members and Diyaudio forum members. It will go global once the collective evaluation is worked through. I realise you're treating my efforts as a commercial crime which is sad, as it undermines collaboration from the only driver who's willing to listen to end-users.
The guys in Japan talk to me without feeling the need to turn issues into a kick-in-the-pants session. Had you taken the time to email me, or message or post in the Markaudio section, the discourse would have been welcomed. We may not always agree, but you constantly forget that I'm not the "typical" manufacture. I do actually listen to what guys say and want. That's how the CHN got born. The "N" by the way stands for "Norio" (Nakajima san) Markaudio's Japan importer.
Now maybe we can get somewhere.
Mark
Let's get technical, Mark.
1. Can you please explain why XRK's measurements of 6 out of 7 drivers match the specs provided by manufacturer's but do not match the published measurement of the CHN70? Also, his measurements match that of at least one more measurement published online.
2. If you are able to, can you redo a measurement of the CHN70 in your anechoic chamber and post a graph with a scale of 50 db and 5 db increments, comparable to the Vifa TC9 spec sheet?
http://www.parts-express.com/pedocs/specs/264-1062-vifa-tc9fd-18-08-specifications.pdf
It is well known that with a calibrated mic and using software such as REW, HolmImpulse and by gating out reflections, quasi-anechoic measurements can be made in a home environment. Such measurements are highly relevant and valid. One look at the multi-way forum or the many other speaker development forums on the internet will show that many DIYers across the world are able to make such measurements. These quasi-anechoic measurements have been shown to match manufacturer published measurements on numerous occasions. The science behind the measurements and their interpretation has been well-established and several books have been written on the subject.
1. Can you please explain why XRK's measurements of 6 out of 7 drivers match the specs provided by manufacturer's but do not match the published measurement of the CHN70? Also, his measurements match that of at least one more measurement published online.
2. If you are able to, can you redo a measurement of the CHN70 in your anechoic chamber and post a graph with a scale of 50 db and 5 db increments, comparable to the Vifa TC9 spec sheet?
http://www.parts-express.com/pedocs/specs/264-1062-vifa-tc9fd-18-08-specifications.pdf
It is well known that with a calibrated mic and using software such as REW, HolmImpulse and by gating out reflections, quasi-anechoic measurements can be made in a home environment. Such measurements are highly relevant and valid. One look at the multi-way forum or the many other speaker development forums on the internet will show that many DIYers across the world are able to make such measurements. These quasi-anechoic measurements have been shown to match manufacturer published measurements on numerous occasions. The science behind the measurements and their interpretation has been well-established and several books have been written on the subject.
Ra,
I am aware of your very partisan position. We should get this fact established so that readers know you come with "baggage attached". I'm also aware of your suggestion to take unfair advantage of Madisound's return policy, having no genuine purchasing intention (post 338 on this thread):
http://www.diyaudio.com/forums/full...ison-3in-4in-class-full-range-drivers-34.html
Answering your points
1 - I already did. Read my previous comments
2 - I'd like it to do it, if all makers and independent testers will agree to standardisation.
Your last statement, give exact Independent references to research that filtered non-anechoic testing is every bit as accurate as anechoic. Comments won't cut it. Show full independent research.
While you're at it. You can explain to the likes of me and other experienced audio guys like Nelson Pass (he's on record liking Lowthers) why full-range drivers should be flat response. An when you've finished, you'll be clearly able to educate designers like me and all members, by publishing your own complete driver specifications, materials lists and design drawings. Similarly for for others who believe their driver knowledge is sacrosanct. Then we can have a proper test.
1. I'm sorry, but I could not glean from your earlier post about why the CHN70 measurements don't agree with the published specs. Could you explain again please?
2. Again, I'm not sure I understand. You could put one of the CHN70s on hand into the anechoic chamber and get a measurement. Who are "all makers" and "independent testers" and why should they agree to standardisation? All I am asking is if you can post a measurement with the scale changed to 50 db. You can even repost the measurement you have with a scale of 50 db.
Regarding references, here is one:
http://audioxpress.com/files/dappolito2959.pdf
Figure 3 could interest you. I never said the quasi-anechoic response is "every bit as accurate as anechoic." What I did say was that quasi-anechoic measurements made in the home environment are "highly reliable, relevant and valid." The only thing they lack is resolution. But resolution is not important in frequencies higher than about 500 Hz because our ear/brain's critical bands are wider than the resolution that can be achieved at home. Besides, even with the higher resolution of an anechoic chamber, you will not see much difference in the two measurements because in both measurements, the measurement is purely of the driver, the reflections have been gated out. It is true that the baffle and mounting will have some impact, but it will be small once you are above about 500 Hz.
D'Appolito also wrote the book "Testing Loudspeakers" in which he explains how to make the quasi-anechoic measurements. Certainly, there is no substitute for high resolution anechoic measurements. But the difference in the measurements posted by XRK and that in the published specs of the CHN70 cannot be accounted for by the difference in the way the two measurements were obtained.
Nelson does not sell speakers, he buys them. He can like whatever he wants. Also, if you have ever read his articles, you will notice that he does all sorts of EQ to tame the rising response of the Lowthers, putting notch filters on the peaks, and supplementing them with subs or large woofers on an OB. It looks like he is well aware with the flaws and does his best to remedy them. He also sells products designed specifically for full range drivers that will provide the users with these EQ functions (btw, these are superb products). The reason he likes full range drivers, I believe, is because of their simplicity. But he encourages users to correct their flaws. He is well aware that a +9db rising shelf is going to sound shrill and shouty.
I think everyone here appreciates that you spend time with the DIYers, me included. But what is not appreciated is a lot of hand waving and telling others about how much knowledge and experience you have. You do not know my educational background, nor do you know my professional experience. Please argue the points. You have not yet communicated why the measurements should be different.
2. Again, I'm not sure I understand. You could put one of the CHN70s on hand into the anechoic chamber and get a measurement. Who are "all makers" and "independent testers" and why should they agree to standardisation? All I am asking is if you can post a measurement with the scale changed to 50 db. You can even repost the measurement you have with a scale of 50 db.
Regarding references, here is one:
http://audioxpress.com/files/dappolito2959.pdf
Figure 3 could interest you. I never said the quasi-anechoic response is "every bit as accurate as anechoic." What I did say was that quasi-anechoic measurements made in the home environment are "highly reliable, relevant and valid." The only thing they lack is resolution. But resolution is not important in frequencies higher than about 500 Hz because our ear/brain's critical bands are wider than the resolution that can be achieved at home. Besides, even with the higher resolution of an anechoic chamber, you will not see much difference in the two measurements because in both measurements, the measurement is purely of the driver, the reflections have been gated out. It is true that the baffle and mounting will have some impact, but it will be small once you are above about 500 Hz.
D'Appolito also wrote the book "Testing Loudspeakers" in which he explains how to make the quasi-anechoic measurements. Certainly, there is no substitute for high resolution anechoic measurements. But the difference in the measurements posted by XRK and that in the published specs of the CHN70 cannot be accounted for by the difference in the way the two measurements were obtained.
Nelson does not sell speakers, he buys them. He can like whatever he wants. Also, if you have ever read his articles, you will notice that he does all sorts of EQ to tame the rising response of the Lowthers, putting notch filters on the peaks, and supplementing them with subs or large woofers on an OB. It looks like he is well aware with the flaws and does his best to remedy them. He also sells products designed specifically for full range drivers that will provide the users with these EQ functions (btw, these are superb products). The reason he likes full range drivers, I believe, is because of their simplicity. But he encourages users to correct their flaws. He is well aware that a +9db rising shelf is going to sound shrill and shouty.
I think everyone here appreciates that you spend time with the DIYers, me included. But what is not appreciated is a lot of hand waving and telling others about how much knowledge and experience you have. You do not know my educational background, nor do you know my professional experience. Please argue the points. You have not yet communicated why the measurements should be different.
Ra,
Thought I was clear in my previous post.
If you want to reference, please do it on this basis.
Quote the specific references (passages) that clearly show validated independent analysis that demonstrates gated/filtered non-anechoic testing an equal to, or greater than anechoic measurement accuracy. Generalised comments won't do.
If you want to continue posting, you'll avoid the type of personalisation you attempted in your previous (deleted) post.
Thought I was clear in my previous post.
If you want to reference, please do it on this basis.
Quote the specific references (passages) that clearly show validated independent analysis that demonstrates gated/filtered non-anechoic testing an equal to, or greater than anechoic measurement accuracy. Generalised comments won't do.
If you want to continue posting, you'll avoid the type of personalisation you attempted in your previous (deleted) post.
I'm not sure what else to do, Mark. I gave you references to a book, to a paper online that is freely accessible. Anybody with an interest in audio should be able to read the references and understand how to make quasi-anechoic measurements. I made no personal remarks and yet my post got deleted. How about some fair treatment? Could you please let my post remain and some others judge whether my post was improper?
Ra,
I genuinely want a discourse with you, I really do!
But I'm MSc Mech Eng and we are wired in a very particular way. I need hard validated evidence. Meaning actual text from any source you mention. And reference the source complete so others can check your interpretation.
Some of stuff going on in the full-range open forum right now beggars rational belief. But if say anything, I continuously get slapped down suggested I'm arrogant when I'm not.
As I said till I'm blue in the face. We're all using different test methods. Nothing more than my long-throw soft suspension drivers to measure the same all the time in any test environment. But chances are they won't.
I really want to help guys. But you're all rattling on about DDR, Polar response + a whole bunch more but none as yet (from what I can read so far) wants to understand the different operation and production criteria applied by driver makers to commercial drivers and their audiophile alternatives. You're all heading for a testing train wreck if you don't properly appreciate the differences in properties of the drivers.
Agree with you regarding Nelson's adjustments re Lowthers - QED, mine also. See my comments in CHN launch thread. We can imagine a future xrk headline "Lowthers are +12dB out here, -7bB out there, the other little flat response driver is the winner!" No wonder Low refrains from publishing frequency data.
Mark
The moderation team apologizes for the confusion in this, and related, threads.As is standard practice, two threads on the same subject were combined into one. In this case, the combined threads landed in a commercial forum where they were further split and edited. This was done in error. This thread has been restored as much as possible.
Member
Joined 2009
Paid Member
Interesting observations.
I have a single copy of this driver and haven't found it hot in the slightest, in fact it sounds rather 'plain' which for most speakers is a compliment.
I do remember having two startling holographic experiences though which may or may not be relevant,
first one was with two speaker boxes wired out of phase to each other (by error) and my FM tuner set on mono. The result was really strange, the sound floated free of and above the speakers like some kind of alien orb - it wasn't stereo of course but the effect was this single source of music that was distinctly disconnected from the speakers.
second time was a pair of MA10.3 pencils - simply outstanding stereo imaging the likes I've not heard before or since (I built them for a friend).
I have a single copy of this driver and haven't found it hot in the slightest, in fact it sounds rather 'plain' which for most speakers is a compliment.
I do remember having two startling holographic experiences though which may or may not be relevant,
first one was with two speaker boxes wired out of phase to each other (by error) and my FM tuner set on mono. The result was really strange, the sound floated free of and above the speakers like some kind of alien orb - it wasn't stereo of course but the effect was this single source of music that was distinctly disconnected from the speakers.
second time was a pair of MA10.3 pencils - simply outstanding stereo imaging the likes I've not heard before or since (I built them for a friend).
Mark, I suggest Testing Loudspeakers where he displays the technique for getting gated results and briefly discussed the math involved. Detailed math explaining the FFT can be read about in many places. I'm surprised your M.Sc in Mech Eng doesn't cover this. There are also countless examples comparing anechoic to gated results. If you'd like I can take some for you to show. It's important to understand that no one is suggesting the anechoic and gated results are comparable <1/Tgate. And the data does get smoothed somewhat. It's all in the well established math if you'd like to read about it.
If gated measurements aren't acceptable to you, any DIYer can provide you with anechoic measurements if that's your preference. It just takes more effort and a long extension cord. If you think that would be more fair to your CHN70 ID take the time to do it. Dave has mentioned he has a pair available.
If gated measurements aren't acceptable to you, any DIYer can provide you with anechoic measurements if that's your preference. It just takes more effort and a long extension cord. If you think that would be more fair to your CHN70 ID take the time to do it. Dave has mentioned he has a pair available.
Glad to see this came back.
About the height of image, I remember reading somewhere that it's related to a certain range of frequency - centered at about 8kHz (maybe wrong). This is probably related to pinna cue - that certain frequency fools our hearing. Sorry I can't trace back to the origin of this, can someone help?
By this, it could be an emphasis around the HF range in the driver, so the imaging becomes that way -- floating high above the location of source.
About the height of image, I remember reading somewhere that it's related to a certain range of frequency - centered at about 8kHz (maybe wrong). This is probably related to pinna cue - that certain frequency fools our hearing. Sorry I can't trace back to the origin of this, can someone help?
By this, it could be an emphasis around the HF range in the driver, so the imaging becomes that way -- floating high above the location of source.
Thanks to Bigun and CLS for the on-topic replies.
I have come to the conclusion that the effect I heard was a reflection off the vertical panels directly behind the up-firing CHN70 boxes. I have no way of measuring how wideband that reflection was, but I find the comments about a range of frequencies centered on 8kHz interesting, because to my ears the unbroken-in CHN70s were too hot in the upper midrange for my preference in on-axis listening. I speculate that those elevated and reflected upper midrange frequencies interfered with the phase of the direct sound and caused the phenomenon I experienced.
Given the unique circumstances, I am not sure of the usefulness of this phenomenon or its predictability. I suppose one could build some up-firing wideband speaker boxes and simply experiment with placement in the listening room. I may try that at some point in the future, after reading the thread about floor coupled up-firing speakers started by graaf.
I have come to the conclusion that the effect I heard was a reflection off the vertical panels directly behind the up-firing CHN70 boxes. I have no way of measuring how wideband that reflection was, but I find the comments about a range of frequencies centered on 8kHz interesting, because to my ears the unbroken-in CHN70s were too hot in the upper midrange for my preference in on-axis listening. I speculate that those elevated and reflected upper midrange frequencies interfered with the phase of the direct sound and caused the phenomenon I experienced.
Given the unique circumstances, I am not sure of the usefulness of this phenomenon or its predictability. I suppose one could build some up-firing wideband speaker boxes and simply experiment with placement in the listening room. I may try that at some point in the future, after reading the thread about floor coupled up-firing speakers started by graaf.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.