For me I think of all the things that DML's deliver its the soundstage that blows me away. I have the privilege of joining Mr Tom Waites in his Diner every time I want to be a Nighthawk. Mr Lou Reed's backing singers joining his walk on the wild side start off in a distant alleyway and end up sitting on my lap.
I have never heard such an immersive 3d soundstage before. Instrument definition is just as outstanding.
I have never heard such an immersive 3d soundstage before. Instrument definition is just as outstanding.
DML Attributes
I thought it might be beneficial at this point to summarise the benefits of DML’s as this information is held in a variety of posts in this thread and it certainly helps me to have all of them gathered together. I think I have included all of them but please add in any I may have missed. There is no ‘ranking’ of attributes implied in the following list, its just a list.
Full range/near full range performance
This has a number of benefits.
1. it removes the need for multiple drive units with different diaphragm dimensions to handle high frequencies and low frequencies which adds complexity and cost.
2. It removes the need for crossovers which, no matter how well designed, introduce a number of problems e.g. crossover points at a point where hearing is most sensitive ( 2000-3000hz), phase change compensation, baffle step compensation, frequency response compensation, efficiency matching, all of which adds complexity and cost ( e.g. Myths & Facts about Loudspeaker Crossovers: Identifying Legitimately High Fidelity Designs | Audioholics)
3. Conventional drivers/ speakers have a directional distribution which reduces the listening sweet spot and in addition means that early reflection from walls is significantly different in spectral composition from a natural sound field introducing artificial coloration (Early Reflections 101 - Acoustic Frontiers) DML’s avoid this problem.
Open Baffle/ Panel
1. DML’s do not need cabinets, removing a significant source of colouration and loss of dynamics ( cabinets resonate because they channel energy away from the drive units( ) ( [A Whitepaper: The Audibility Of Cabinet Panel Resonances and Pat. Pend. Method Of Reduction Of Audible Coloration - Dagogo)
2. Because there is no cabinet there is no early back reflection from the cabinet walls passing through the cone. This can be reduced but not eliminated entirely.
3. DML wave propagation is by transverse waves, not pistonic (Wave Propagation) . This reduces distortions like Doppler distortion experienced by conventional drivers . Doppler Distortion in loudspeakers
3. Typically DML’s have a much larger surface area than conventional drivers which results in better impedance matching with air ( https://en.wikipedia.org/wiki/Impedance_matching#Acoustics
Bipolar Operation
DML’s act as Bipoles, not Dipoles i.e. both front and back radiation are in phase. This has a number of benefits.
1. There is no phase cancelation between front and back waves leading to an omnidirectional sound-field
2. Unlike dipoles early reflected bass is additive, not subtractive.
3. Unlike electrostatics, where dipole operation leads to a very directional response with a narrow ’sweet-spot’, DML’s have a very wide listening area.
Just to note that there is a reduction in output to the side of DML’s but this is not due to cancelation, it is due to a lack of output from the edges of the panels. This null point is notably much smaller that that experienced with dipoles, especially electrostatics.( My comment based on 20 years experience designing an building OB and electrostatic speakers)
Bending wave/Transverse wave operation
I have noted one of the benefits of this mode of operation above but there are further benefits that should be noted.
1. DML’s do not move pistonicaly and therefore have no moving mass. This removes a constraint on loudspeaker performance regarding high frequency reproduction.
2. All frequencies down to the bass cut-off are transmitted equally across the entire surface of the panel giving the typical DML ‘window on the performance’ effect. Although in principle this also happens with electrostatics because they are dipoles and not bipoles this advantage is lost to them through phase cancelation.
3. Due to its mode of operation edge reflection is a potential problem with DML’s but is very easily dealt with compared to the complex problem of dealing with mechanical vibration in drive units and cabinets. To be clear, it IS a problem, but low order compared to other systems. For balance, electrostatics are free of these problems entirely.
Easy Drive Characteristics
This is a much overlooked benefit of DML’s but coming from a background of electrostatic speaker development where substantial cost and complexity is required to provide a decent load to an amplifier (from using expensive transformers with phase and bandwidth problems up to and including building custom direct drive amplifiers from scratch- a difficult and dangerous task) I can assure you this is a major benefit.
1. DML exciters exhibit a simple largely resistive impedance to an amplifier opening the door to the use of all sorts of purist through to exotic amplifier designs.
2. DML exciters are ridiculously inexpensive devices for their performance.
3. Although some early papers report sensitivity and efficiency failings in my experience this is overstated- DML’s play LOUD. Thats very loud and very clean.
Inherently Low Distortion Devices
I know I am going to catch some flack for this claim but its based on my listening experiences and also some of the measurements reported in this thread.
From my measurements a 3mm beech ply DML panel crossed over at 100hz to a subwoofer measures a maximum of 1% THD+n dipping to 0.3% at 500hz. For reference speakers in the Bowers & Wilkins 800 range, e.g. the 805 D3 stand mounted model measures 1% to 0.6% THD+n. (https://www.bowerswilkins.com/sites/default/files/2018-08/805-D3-info-sheet_0.pdf) The 805 D3 costs £4,500 the pair. Maybe I measured it all wrong, happy to be challenged on that, but they do sound like low distortion devices to me.
The last and most obvious point is that this technology is perfect for DIY giving stunning results for very little cost and time expenditure. It is also in its infancy relative to ‘conventional’ methods with much still to be explored and invented.
Please do shoot me down where you think I am wrong or add in points I have forgotten.
Have a good week all!
Burnt
I thought it might be beneficial at this point to summarise the benefits of DML’s as this information is held in a variety of posts in this thread and it certainly helps me to have all of them gathered together. I think I have included all of them but please add in any I may have missed. There is no ‘ranking’ of attributes implied in the following list, its just a list.
Full range/near full range performance
This has a number of benefits.
1. it removes the need for multiple drive units with different diaphragm dimensions to handle high frequencies and low frequencies which adds complexity and cost.
2. It removes the need for crossovers which, no matter how well designed, introduce a number of problems e.g. crossover points at a point where hearing is most sensitive ( 2000-3000hz), phase change compensation, baffle step compensation, frequency response compensation, efficiency matching, all of which adds complexity and cost ( e.g. Myths & Facts about Loudspeaker Crossovers: Identifying Legitimately High Fidelity Designs | Audioholics)
3. Conventional drivers/ speakers have a directional distribution which reduces the listening sweet spot and in addition means that early reflection from walls is significantly different in spectral composition from a natural sound field introducing artificial coloration (Early Reflections 101 - Acoustic Frontiers) DML’s avoid this problem.
Open Baffle/ Panel
1. DML’s do not need cabinets, removing a significant source of colouration and loss of dynamics ( cabinets resonate because they channel energy away from the drive units( ) ( [A Whitepaper: The Audibility Of Cabinet Panel Resonances and Pat. Pend. Method Of Reduction Of Audible Coloration - Dagogo)
2. Because there is no cabinet there is no early back reflection from the cabinet walls passing through the cone. This can be reduced but not eliminated entirely.
3. DML wave propagation is by transverse waves, not pistonic (Wave Propagation) . This reduces distortions like Doppler distortion experienced by conventional drivers . Doppler Distortion in loudspeakers
3. Typically DML’s have a much larger surface area than conventional drivers which results in better impedance matching with air ( https://en.wikipedia.org/wiki/Impedance_matching#Acoustics
Bipolar Operation
DML’s act as Bipoles, not Dipoles i.e. both front and back radiation are in phase. This has a number of benefits.
1. There is no phase cancelation between front and back waves leading to an omnidirectional sound-field
2. Unlike dipoles early reflected bass is additive, not subtractive.
3. Unlike electrostatics, where dipole operation leads to a very directional response with a narrow ’sweet-spot’, DML’s have a very wide listening area.
Just to note that there is a reduction in output to the side of DML’s but this is not due to cancelation, it is due to a lack of output from the edges of the panels. This null point is notably much smaller that that experienced with dipoles, especially electrostatics.( My comment based on 20 years experience designing an building OB and electrostatic speakers)
Bending wave/Transverse wave operation
I have noted one of the benefits of this mode of operation above but there are further benefits that should be noted.
1. DML’s do not move pistonicaly and therefore have no moving mass. This removes a constraint on loudspeaker performance regarding high frequency reproduction.
2. All frequencies down to the bass cut-off are transmitted equally across the entire surface of the panel giving the typical DML ‘window on the performance’ effect. Although in principle this also happens with electrostatics because they are dipoles and not bipoles this advantage is lost to them through phase cancelation.
3. Due to its mode of operation edge reflection is a potential problem with DML’s but is very easily dealt with compared to the complex problem of dealing with mechanical vibration in drive units and cabinets. To be clear, it IS a problem, but low order compared to other systems. For balance, electrostatics are free of these problems entirely.
Easy Drive Characteristics
This is a much overlooked benefit of DML’s but coming from a background of electrostatic speaker development where substantial cost and complexity is required to provide a decent load to an amplifier (from using expensive transformers with phase and bandwidth problems up to and including building custom direct drive amplifiers from scratch- a difficult and dangerous task) I can assure you this is a major benefit.
1. DML exciters exhibit a simple largely resistive impedance to an amplifier opening the door to the use of all sorts of purist through to exotic amplifier designs.
2. DML exciters are ridiculously inexpensive devices for their performance.
3. Although some early papers report sensitivity and efficiency failings in my experience this is overstated- DML’s play LOUD. Thats very loud and very clean.
Inherently Low Distortion Devices
I know I am going to catch some flack for this claim but its based on my listening experiences and also some of the measurements reported in this thread.
From my measurements a 3mm beech ply DML panel crossed over at 100hz to a subwoofer measures a maximum of 1% THD+n dipping to 0.3% at 500hz. For reference speakers in the Bowers & Wilkins 800 range, e.g. the 805 D3 stand mounted model measures 1% to 0.6% THD+n. (https://www.bowerswilkins.com/sites/default/files/2018-08/805-D3-info-sheet_0.pdf) The 805 D3 costs £4,500 the pair. Maybe I measured it all wrong, happy to be challenged on that, but they do sound like low distortion devices to me.
The last and most obvious point is that this technology is perfect for DIY giving stunning results for very little cost and time expenditure. It is also in its infancy relative to ‘conventional’ methods with much still to be explored and invented.
Please do shoot me down where you think I am wrong or add in points I have forgotten.
Have a good week all!
Burnt
Last edited:
This is a great summary and usefully devoid of certain design elements discussed above.
My poplar panels cut in at 300hz, but I could easily make it lower if needs be. However, the Monacor 15 incher is an excellent driver and fully capable at these higher frequencies. It does need tailoring though to compensate for the U frame hump around the U's resonant frequency.
I too find no efficiency issues and my TPA3116 monoblocs have no problems driving them to a more than acceptable volume.
As you say, the soundstage is huge and better than any 'point and squirt' speaker I've owned.
Today I installed a cheapo 6J1 based preamp with stock Chinese valves. Things improved still further over and above the old Arcam 7R I was using as a preamp. These Chinese 6J1 valves are known not to be the best - they are rather hot in the treble department. Therefore, some GE 5654W will be arriving midweek.
An additional aspect is the transient response of my panel over OB bass setup. Quite simply, it's much superior to any conventional coned speakers in a box I've owned.
I heard a couple of pairs of OB speakers at the UK National Audio Show in September. In my opinion, in my room and with my system, I prefer the sound of my DIY panel/OBs.
For even a barely half-competent DIYer like me, DMLs represent an economical way of trying something very different that might just have a transformative effect on one's listening experience.
Kind of you to say. The more we share with one another the more we all learn.
I have also found changes to the front end are revealed very clearly by the DMLs as well.
I found the effect of using a tube Preamp was very noticeable, in fact tubes in the chain with DML's make a lovely sound.
For me I think that the areas I am motivated to try next are exploring panel materials as they are a major component in determining sound and there are a whole stack of options to be tested ( e.g poplar v beech v ash v balsa cross grain v bamboo laminates just in the ply board options) and then exciter types, designs, mounting and positions. DMLBES was very helpful to me when I was disappointed in the performance of a Tectonics exciter I had bought and his mod made a significant difference.
I find beech ply to be an excellent alternative to carbon fibre which is my current neutrality standard. I would love to go to carbon fibre but the cost for a large panel is daunting. Beech ply is a trifle more characterful, really just a hint of colouration if you listen hard, but in VFM terms its a steal.
Time to shut down this laptop and pop across to Chicago for Solti's Beethoven 9th. First LP I ever bought.
I have also found changes to the front end are revealed very clearly by the DMLs as well.
I found the effect of using a tube Preamp was very noticeable, in fact tubes in the chain with DML's make a lovely sound.
For me I think that the areas I am motivated to try next are exploring panel materials as they are a major component in determining sound and there are a whole stack of options to be tested ( e.g poplar v beech v ash v balsa cross grain v bamboo laminates just in the ply board options) and then exciter types, designs, mounting and positions. DMLBES was very helpful to me when I was disappointed in the performance of a Tectonics exciter I had bought and his mod made a significant difference.
I find beech ply to be an excellent alternative to carbon fibre which is my current neutrality standard. I would love to go to carbon fibre but the cost for a large panel is daunting. Beech ply is a trifle more characterful, really just a hint of colouration if you listen hard, but in VFM terms its a steal.
Time to shut down this laptop and pop across to Chicago for Solti's Beethoven 9th. First LP I ever bought.
Sounds like we are singing from the same hymn sheet!
My first LP was Fritz Reiner conducting Also Sprach Zarathustra when I was about 16yo.
I feel a bit of Shostakovich coming on.
Well, Burnt Coil,
that’s an excellent summary and strongly correlates to my own very limited personal experience of DML’s, although I’d say that the reduced volume from the edge of panels really is not much of an issue. I have just turned my panels edge-on and played the opening track of Antonio Forcioni’s live album (with his band) which a stunning recording and the sound is just as good and maybe only slightly reduced in volume.
that’s an excellent summary and strongly correlates to my own very limited personal experience of DML’s, although I’d say that the reduced volume from the edge of panels really is not much of an issue. I have just turned my panels edge-on and played the opening track of Antonio Forcioni’s live album (with his band) which a stunning recording and the sound is just as good and maybe only slightly reduced in volume.
@BurntCoil - Outstanding review of what a DML can be! One other point you may want to add is that due to its almost bipolar omni directionality, the loss of SPL's as you move away from the speaker is about half that of the loss that you get from a piston type driver. I.e., there will be more volume at your listening position with a DML than there will be from a piston driver for the same speaker effieciency. This adds to the impression that DML's play louder, even if they aren't as effecient in a near field measurement.
Hi BurntCoil
I find your comments re Lou Reed's Walk on The Wild Side very interesting. I have to say my DML's are the only speakers I've ever really enjoyed listening to Lou Reed's Transformer album on. It has always sounded poorly recorded, kind of unbalanced on any other speakers I've owned. On my DMLs the album 'makes sense' recording wise now. DML's Rule!
Everyone, as a simple experiment, if you haven't done it already, wire one DML speaker out of phase with the other. I think you'll find very little difference with the sound at the listening position with them in phase or out of phase. I am no audio engineer, but maybe this is the closest a speaker can come to how sound is naturally emmited from music instruments, as far as 'no' introduced phase anomalies go, and is a large part of their magic. As I said, I am no expert, just a believer.
I find your comments re Lou Reed's Walk on The Wild Side very interesting. I have to say my DML's are the only speakers I've ever really enjoyed listening to Lou Reed's Transformer album on. It has always sounded poorly recorded, kind of unbalanced on any other speakers I've owned. On my DMLs the album 'makes sense' recording wise now. DML's Rule!
Everyone, as a simple experiment, if you haven't done it already, wire one DML speaker out of phase with the other. I think you'll find very little difference with the sound at the listening position with them in phase or out of phase. I am no audio engineer, but maybe this is the closest a speaker can come to how sound is naturally emmited from music instruments, as far as 'no' introduced phase anomalies go, and is a large part of their magic. As I said, I am no expert, just a believer.
Here's one more thing to add to the list.
I have previously investigated using full range drivers as mid/tweeters in conjunction with a bass driver. By avoiding crossovers in sensitive areas, the overall sound is supposed to be more coherent. However, the larger the mid/tweeter, the more prone it is to beaming above certain frequencies, unless there's some very fancy (and usually expensive) driver design work. And for frequencies above the typical crossover point (which could be anywhere from 80hz to 320hz) small 3 inch full range drivers just don't cut it (and yes, I've tried).
DMLs on the other hand do not suffer from these beaming problems, and provided that one gets the design and materials right, offer very acceptable treble performance without the need for additional drivers.
I have previously investigated using full range drivers as mid/tweeters in conjunction with a bass driver. By avoiding crossovers in sensitive areas, the overall sound is supposed to be more coherent. However, the larger the mid/tweeter, the more prone it is to beaming above certain frequencies, unless there's some very fancy (and usually expensive) driver design work. And for frequencies above the typical crossover point (which could be anywhere from 80hz to 320hz) small 3 inch full range drivers just don't cut it (and yes, I've tried).
DMLs on the other hand do not suffer from these beaming problems, and provided that one gets the design and materials right, offer very acceptable treble performance without the need for additional drivers.
Hi Gapmedia,
I agree there are a number of recordings I never really liked that now 'make sense'. There are also a few that I used to like but now sound artificial and 'thin' which is a shame. DML's seem to expose recording and editing flaws really well. I was listening to Led Zeppelins Tangerine a few days back and quite by chance happened to be standing close to one of the panels. You could clearly hear a tiny whispered exchange between Robert Plant and Jimmy Page and then Page counting into the start. Tiny tiny sounds not meant for the mike thatI have never heard before and to my knowledge no one has ever mentioned.
You make a very interesting point about phase that I had completely missed. I will try your experiment as well.
Snap! I listened to the very same track two days ago. I am now impatient for my GE preamp valves to arrive so that I can do some more serious listening.
A power amp section change is also on the cards. I have a couple of TDA7498E modules to try. Mids & tops are supposed to be very nice with this chip.
Of course xrk971 might suggest I try TPA3255 boards next!
Very good summary by @burntcoil on the previous page.
I went around town looking for the highest density XPS/EPS I could find at 20/30mm and apparently that is not stocked here. And since I don't want to buy an entire crate of the those...
@DMBLES
How would a framed DML differ from the unframed one in terms of construction.
The way I've seen styrofoam based DMLs suggested for building is with rounded edges and rounded corners, covered with a diluted mixture of PVA.
If I frame them, do I still want them rounded?
I assume I still want them hanging away from the wall and not on the wall as a painting.
I went around town looking for the highest density XPS/EPS I could find at 20/30mm and apparently that is not stocked here. And since I don't want to buy an entire crate of the those...
@DMBLES
How would a framed DML differ from the unframed one in terms of construction.
The way I've seen styrofoam based DMLs suggested for building is with rounded edges and rounded corners, covered with a diluted mixture of PVA.
If I frame them, do I still want them rounded?
I assume I still want them hanging away from the wall and not on the wall as a painting.
In the piston driver world you need smaller diameter drivers for higher freq's. In the DML world larger panels are actually better for HF's . The reason being that the larger panel supports more modal points, which evens out the FR. The HF already creates a large number of modal points and making the panel larger simply enhances that. Obviously, larger panels also enhance the Mid and Bass as well.
DML's do beam at higher Freq's however, they create a series of beams (or lobes if you will) that spread the sound in such a manner as to sound as if its omni...i.e., kind of a beaming omni. Again the larger the panel, the more beaming lobes you can get to smooth it all out in the far field.
I really need to stop by HD on my way home and pick up a couple of XPS panels and 2x4’s and build this baby. I am thinking of going with the foam edge suspension idea. I think the melamine scrubber sponge pads that I use for sound dampening (same as BASF Basotect) hot melt glued to the edges of the XPS and mounted in a frame might be the way to go. I am also thinking a trapezoid shape might be good to prevent regular drum head modes.