A Study of DMLs as a Full Range Speaker

Hey DMLBES,
Personally I think trying to guide the sound from DML’s is counter productive to the enjoyment of how the sound is emitted from a panel in open space. If that is want you want then that’s what you should do, but the intriguing thing about DML’s is their ability to project sound in all directions so using a wave guide would disturb this benefit. I say forget trying to produce anything below 100HZ (give or take) as mechanical vibration issues will become increasingly more of a problem, especially if you have any spine or frame attachments. Suspending a pair of panels freely in space will never ‘sell’ as a commercial idea so of course the design must be modified to look more conventional and that requires various methods of framing and support for the exciter.
I wish I had more time and in particular, space, for experimentation as I might change my mind, but so far I cannot see why the added complication of frames, spines and waveguides
would bring any benefits.
You obviously have a much deeper understanding on the topic so I do take on board what you say and everyone else of course, but I really like what I am hearing even with this very simple set-up. £20.00 for two exciters and a couple of second hand ceiling tiles should not produce sound this good. It’s weird and marvellous at the same time.
 
So I had a bit of time this afternoon and thought I would try to measure the distortion from my panels. It was an attempt to answer a question from some time back.

First point is this was a very basic attempt. I measured the panels in situ in my study in my house in London where there is ambient noise to contend with. Secondly I used an iPad App to measure the THD+n levels with the signal coming from a web based sound generator. So this is just a rough guide rather than anything definitive.
The measurements were made at 0.5M as the App directed.

The results did surprise me so I have averaged the results from a number of runs.


f THD+n
125hz. 0 .96%
250hz. 0.67%
500hz. 0.35%
1000hz. 1.15%

I didn’t bother testing lower than 125hz as I cross over to subs. I didn’t go higher as the App didn’t have a setting for that. So a surprising result and no doubt caused by my inability to measure correctly ; )
 
DMLs will frequency cancel at the edges if you don't place an open baffle style frame around them. That may or may not be good depending on your goals.

DMLs can reproduce low frequencies but that's not their comparative advantage. Their advantage is above the Schroeder transition frequency. In addition, humans can't localize low frequencies so there's no need to reproduce low frequencies in the mains.

There's no perfect DML or right way to build a DML. When you read research papers on DMLs you want to keep their goals in mind. Are they designing a DML for a large space like a stadium? If so, some of their methods won't benefit you in a small room like a house. You also want to keep an eye on the dates of the research paper. Where they written before new technology came along? Some older papers try to defeat problems that were difficult when the paper was written but are easily solved with signal processing today.

Personally, I chose to go the brute force route (individually amplified MAP arrays with DSP) because I built a one-off. I'm not manufacturing thousands of units.

If you know your goals and you know the comparative advantage DMLs offer you can build an excellent set of speakers. Don't get caught up in the idea that there is only one perfect way to build these things.
 
Hey DMLBES,
Personally I think trying to guide the sound from DML’s is counter productive to the enjoyment of how the sound is emitted from a panel in open space. If that is want you want then that’s what you should do, but the intriguing thing about DML’s is their ability to project sound in all directions so using a wave guide would disturb this benefit. I say forget trying to produce anything below 100HZ (give or take) as mechanical vibration issues will become increasingly more of a problem, especially if you have any spine or frame attachments. Suspending a pair of panels freely in space will never ‘sell’ as a commercial idea so of course the design must be modified to look more conventional and that requires various methods of framing and support for the exciter.
I wish I had more time and in particular, space, for experimentation as I might change my mind, but so far I cannot see why the added complication of frames, spines and waveguides
would bring any benefits.
You obviously have a much deeper understanding on the topic so I do take on board what you say and everyone else of course, but I really like what I am hearing even with this very simple set-up. £20.00 for two exciters and a couple of second hand ceiling tiles should not produce sound this good. It’s weird and marvellous at the same time.

How do you know if its counter productive if you have never tried and or experience it? That's my whole point. I am not arguing about personal preference. I am arguing about actual hands on experience combined with facts to back it up. How does a frame and a spine make mechanical vibrational issues more of a problem especially when you have NOT tried it yet? What are you basing it off of? Again no legit logical reasoning to back up such statements. How do you know panels supported by strings will never sell??? It seems a lot of people who watch Tech ingredients vids buy into his design and hang them and if Tech ingredients wanted to sell his panels I bet a lot of people would buy them commercially as he would be the first commercialized business to ever hang panels from a string. No other commercialized company who build DML panels ever hung panels from a string for the mere fact it is not the ideal way to get decent audio sound quality nor is it asthetically pleasing. It seems you prefer to go the opposite way of all the FACTS and BENIFITS presented to you, not sure why. I mean if you gave me a legit logical explanation with actual experience to back it up I can see where you are coming from but to just listen to Tech Ingrdients because he said to hang them by strings with no legit logical reasoning behind it I just cant fathom that type of logic. Its like some people saying they don't like that particular speaker even though they have NEVER HEARD those speakers play, or they don't like that particular type of food even though they have never tasted it in there life. How would one know whats true/works and what is false/don't work if one has never experienced it themselves first hand? My main pet peeve is when someone makes bold statements/claims and tries to argue something they have never experienced first hand and ignores all the factual benefits presented to them.
 
The results did surprise me so I have averaged the results from a number of runs.


f THD+n
125hz. 0 .96%
250hz. 0.67%
500hz. 0.35%
1000hz. 1.15%

I didn’t bother testing lower than 125hz as I cross over to subs. I didn’t go higher as the App didn’t have a setting for that. So a surprising result and no doubt caused by my inability to measure correctly ; )

How was this surprising? What did you expect and why? Your results are in the range of what I measured for my DML, so seem reasonable to me.

Eric
 
Veleric I was not expecting the figures to be as good as this. I would have been happy to have seen figures similar to a good conventional speaker but these figures appear to better all but high end systems like B&W’s 800 series which is a real surprise to me. For a panel
of plywood and a cheap exciter it seems like magic. Still getting used to the new paradigm, but very happy with the result.
 
DMLBS,
I think no you are taking this a bit too seriously. Anything I say about DML’s is based on my own limited personal experience, and, as I have also said, I wish I had more time
(and space) to experiment with the various techniques that are on offer here by you and others. But I don’t so......
This is DIY Audio (emphasis DIY) and I am thoroughly enjoying enjoying what I am hearing from the set up I have at present. Please re-read my last post to you as I think you misunderstand what I am talking about. I’m stating the blindingly obvious fact (unless you have never heard a DML in the round) that if you try to waveguide a DML you are spoiling the inherent qualities of a DML.
When you say I choose to go against all the ‘factual benefits’ that are presented to me, which so called ‘facts’ are you talking about? KISS is pertinent to this discussion. In my world at least.
 
Always people in these discussions become adamant that their way is the only way. Better to share ideas guys... And talk of wave guides. Hmm, usually these are relevant to tweeters as dimensions are related to wavelengths. Note the curvature they have? Putting sides on a DML is not a waveguide.
 
Exactly.......I’m not saying that what I have is superior to anything else, I’m just enjoying what I’ve got. Each to their own. We all have limits of experience and facilities and money so have to make do with what we have at hand. If it sounds good compared to what you are used too, then it is a good thing, regardless of the ‘facts’ involved. Let’s not forget that sound ‘quality’ is fairly close to 100% subjective in nature. It’s a frame of reference thing....
 
DMLBS,
I think no you are taking this a bit too seriously. Anything I say about DML’s is based on my own limited personal experience, and, as I have also said, I wish I had more time
(and space) to experiment with the various techniques that are on offer here by you and others. But I don’t so......
This is DIY Audio (emphasis DIY) and I am thoroughly enjoying enjoying what I am hearing from the set up I have at present. Please re-read my last post to you as I think you misunderstand what I am talking about. I’m stating the blindingly obvious fact (unless you have never heard a DML in the round) that if you try to waveguide a DML you are spoiling the inherent qualities of a DML.
When you say I choose to go against all the ‘factual benefits’ that are presented to me, which so called ‘facts’ are you talking about? KISS is pertinent to this discussion. In my world at least.

Again I am not arguing about personal preference as you can do what ever you want I don't care. What I do care about is when you make a bold statement like and I quote> . "I’m stating the blindingly obvious fact (unless you have never heard a DML in the round) that if you try to waveguide a DML you are spoiling the inherent qualities of a DML" …...How would you know this if you have NEVER TRIED IT? What are you basing this off of?

Factual benefits of using a frame and a spine PREVENTS voice coil sag this is a factual FACT not an opinion. Ask any speaker manufacturer about voice coil sag and they will all say that it is a big issue.

Anything that can help make ones panel sound better is pertinent to this discussion.....Kiss being pertinent is only in your world.
 
Always people in these discussions become adamant that their way is the only way. Better to share ideas guys... And talk of wave guides. Hmm, usually these are relevant to tweeters as dimensions are related to wavelengths. Note the curvature they have? Putting sides on a DML is not a waveguide.

No that is not my approach about my way is the only way. There are scientific standard foundations for producing high quality sound. For example if one says they are going to build a conventional cone 3 way speaker but use no filters crossover will it sound as good compared to another 3 way with a good 3 way crossover? NO IT WONT.

Now you are talking semantics about wave guides. What do you call putting sides on a DML? Whats the official term?
 
Last edited:
It’s really not worth entering into an argument with you about this is there? The law of diminishing returns is also another rule to observe, especially if your resources or time or space are limited. (as has already been pointed out)
In what way would a waveguide help a highly omnidirectional speaker system? Is it a bold statement to say that it would be ‘spoiling the benefit’ of a DML? I don’t think so!
Voice coil sag? Yeah, I get it. But if this was such a major issue then exciters would be designed differently, wouldn’t they?
 
It’s really not worth entering into an argument with you about this is there? The law of diminishing returns is also another rule to observe, especially if your resources or time or space are limited. (as has already been pointed out)
In what way would a waveguide help a highly omnidirectional speaker system? Is it a bold statement to say that it would be ‘spoiling the benefit’ of a DML? I don’t think so!
Voice coil sag? Yeah, I get it. But if this was such a major issue then exciters would be designed differently, wouldn’t they?

Well you are the one arguing with nothing to back up your statements. The law of your diminishing returns has nothing really to do with the argument as that is your personal problem.

Any time you don't have any factual proof/evidence/experience to back up and support your claim(it would be spoiling the benefit of a DML) then it becomes a BOLD statement.

Yes voice coil sag. Do you? No, exciters have been design this way since the 50's or earlier it hasn't changed. The only thing changes is the design. All the commercially available designs use a frame and a spine specifically for voice coil sag. I just don't understand how you cant see it. ANY WAY I am done trying convince you. DO WHAT EVER YOU WANT.
 
Now you are talking semantics about wave guides. What do you call putting sides on a DML? Whats the official term?


Well it could be called a baffle or a tuned line depending on which way these sides run. If it's putting the dml in an extended frame its like an h-frame enclosure for a woofer. No effect at high frequencies as a wave guide would, but it will introduce a resonance in the few hundred hertz region, which you probably don't want.
 
Well it could be called a baffle or a tuned line depending on which way these sides run. If it's putting the dml in an extended frame its like an h-frame enclosure for a woofer. No effect at high frequencies as a wave guide would, but it will introduce a resonance in the few hundred hertz region, which you probably don't want.

Its like GR Research Super V's open baffle speakers and or open baffle subs. Its like a H frame without the H. Its just a extended square frame. Since there is no official name I will call it a extended square frame. Are you satisfied now with the new term? 🙄....I never mentioned anything about the high frequencies as I kept on repeating myself about the bass response so that could of gave you a clue of what I was talking about. If there is any or if needed I can EQ the 100hz resonance or I can use damping material around the frame like GR Research uses with there NO REZ foam. The fact of the matter is it works or else people wouldn't be using it in there open baffle designs to increase bass performance.
 
Last edited:
DMLBES,
using capitals usually means you’re shouting. So on that basis, and the fact that you don’t appear to recognise that I have given way to and accepted you ‘superior’ knowledge about DML’s and would, if I had the time, space and money to commit to further in-depth experimentation most likely try some of the methods you put forward. End of......
 
Morning all and back to the music.

Just a simple observation. I am currently listening to La Nozze de Figaro by Mozart played by the London Philharmonic conducted by Sir George Solti. I am sitting on the sofa more or less at right angles to the central axis of the conventional stereo sweet spot right beside one panel with the other 2 meters away. I can still hear a clearly defined sound stage with the elements of the orchestra and singers clearly defined in space. Vocals are crystal clear, strings are resonant and textured, brass is nicely raspy and rich.

Frankly its like being in the concert hall, a performance I have strived to achieve for years but have have never quite attained.

DML's, however you build them, are bloody amazing.
 
Last edited: