Hi,
I have a few topics at this forum about my DIY electrostatic panels. This one is about my latest creation:
My new stats, model #4
Yesterday a friend took some cd's that contain a lot of bass ('punchiness' / pressure at low frequencies). The music sounded rather flat on my stats. After the test we tested my 2-way Scanspeak DIY transmission line boxed loudspeakers and it sounded way different: it had a lot of 'puch': pressure. This particular music was made to be punchy / use a lot of pressure at lower frequencies.
My stats sound very fine in my opinion for the music of my personal taste (classical music, some pop music, electronic music, human voices). But this type of music lacks bass 'energy':
Mute Records • Diamond Version • EP4 - Out 6 May on 12" & Download - Mute Records
This isn't a big deal as normally I don't listen to this type of music, but from a technical point of view I'm interested in what exactly causes this difference between the esl and boxed speaker, and would commercial esl's show a similar difference?
When measuring frequency responses at listening position I can't see big differences between my esl's and boxed speakers, my esl's go quite low. When listening to high quality recordings with for instance organs, I feel some pressure and I think bass sounds quite good. Any idea about what could be the cause of this difference? And would it be just my particular esl, or would it be something all els's have? (unfortunately I don't own a set of let's say Quad els's to make comparisons...)
I have a few topics at this forum about my DIY electrostatic panels. This one is about my latest creation:
My new stats, model #4
Yesterday a friend took some cd's that contain a lot of bass ('punchiness' / pressure at low frequencies). The music sounded rather flat on my stats. After the test we tested my 2-way Scanspeak DIY transmission line boxed loudspeakers and it sounded way different: it had a lot of 'puch': pressure. This particular music was made to be punchy / use a lot of pressure at lower frequencies.
My stats sound very fine in my opinion for the music of my personal taste (classical music, some pop music, electronic music, human voices). But this type of music lacks bass 'energy':
Mute Records • Diamond Version • EP4 - Out 6 May on 12" & Download - Mute Records
This isn't a big deal as normally I don't listen to this type of music, but from a technical point of view I'm interested in what exactly causes this difference between the esl and boxed speaker, and would commercial esl's show a similar difference?
When measuring frequency responses at listening position I can't see big differences between my esl's and boxed speakers, my esl's go quite low. When listening to high quality recordings with for instance organs, I feel some pressure and I think bass sounds quite good. Any idea about what could be the cause of this difference? And would it be just my particular esl, or would it be something all els's have? (unfortunately I don't own a set of let's say Quad els's to make comparisons...)
Last edited:
First....great work, one of dream like Diy ESL i have seen.
An thanks for posting your findings.
After 45 years of working in the Audio biz an ending up here with Esls as my dream speakers, me saying what i have heard an think can be fun for some.
For sure i know, is if you dont go...you wont know!
Punchy, ML CLX has a crossover two way, King sound sound also crossover two way.
Crossover in a Esl or any speakers to my ears, dead end.
So geting any one driver of any type to do all the things we would like, physics not yet!
So keep up the good work...have fun always....good luck
all just one mans finding
An thanks for posting your findings.
After 45 years of working in the Audio biz an ending up here with Esls as my dream speakers, me saying what i have heard an think can be fun for some.
For sure i know, is if you dont go...you wont know!
Punchy, ML CLX has a crossover two way, King sound sound also crossover two way.
Crossover in a Esl or any speakers to my ears, dead end.
So geting any one driver of any type to do all the things we would like, physics not yet!
So keep up the good work...have fun always....good luck
all just one mans finding
A box speaker momentarily pressurizes the room which causes a build up in bass below the room fundamental
Make a measurement in the listening position and present it here? Use pink noise and wave the microphone in a circle in front of your head. I.e. play a PN track and use the RTA in REW.
//
//
Using large Sound Lab electrostats here. Can get huge bass, but it depends on everything. Speakers, their placement, power amp choice, dac choice, etc. That said, a friend with a similar setup added stereo Rel subs to get the very deepest bass reproduction down to 15Hz.
Anyway, point is that just because you changed speakers and the sound changed, don't necessarily assume that's all there is to it.
Anyway, point is that just because you changed speakers and the sound changed, don't necessarily assume that's all there is to it.
Last edited:
DAC would be the furthest component in the chain of replay audio electronics these days when it comes to affecting bass performance. It's so far away that it needs not to be bothered looking at.
Bass from dac is much fuller, tighter, and distinct with Topping D90 than with Benchmark DAC-3. People who bother to listen will know, people who theorize it can't matter can't know for sure.
Last edited:
People who bother to subjectively listen say all kinds of claims. This is nothing new.
You left out the measurement. I wonder why.
Silvershadelynx,
The LF response of your panels certainly looks impressive at the SPL you tested at.
The track you were playing would definitely reveal differences due to dynamic range, as well as differences due to LF room interaction between the different speaker types.
1) Both 'stats and moving coil speakers have an Xmax, a linear range of excursion. The large panel only has to move a small fraction of a millimeter to have the equivalent displacement and SPL as the moving coil speaker, but has very limited excursion capability compared to the moving coil.
2) From your description, sounds like a difference in dynamic range between the two speakers. I suspect that as level is increased, your stats low frequency range reaches an excursion limit which can't be exceeded, so responds like a LF dynamic compressor/limiter, the highs respond normally, but bass is relatively reduced.
Without testing the dynamic response of your ESLs compared to others, can't comment other than many commercial models use moving coil subs to augment their limited bass output capabilities.
You should be able to test your ESL dynamic range using pink noise, if the LF does not increase equal to the HF as volume is increased to the level that could be achieved with the "box speaker", you have the explanation of "lack of punch".
Nice work on your panels!
Art
There are a few reasons why full range ESLs tend to have less bass “punch” than boxed loudspeakers.
Some are specific to ESL operation while others are related to general acoustic principles and/or psychoacoustics.
Step-up transformer compression/limiting: The Audiostatic transformers you are using have only 60 primary turns on a small 1.8in^2 core, so they are well down the saturation curve by only 10Vrms at 30Hz. The resulting falling primary inductance combined with the winding resistance and 2.2ohm resistance you have in series with the primary results in compression of bass transients. It is a difficult problem to get full range transformers with step-up ratios of 150:1 like you are using. That is one of the reasons I chose to limit my ESL to ~70Hz. With a step-up ratio of 140:1 they produce uncompressed 40Vrms. An alternate method is to use separate transformers for bass and treble. For example, Sound Lab bass transformers can handle 50Vrm at 30Hz without compression. DIY bass transformer for ESLs ?
Dynamic Range capability As Weltersys mentioned, it may also just come down to dynamic range or peak output capability. You mentioned in another thread that you set your HV bias as high as possible. It is possible that during bass transients the air breakdown threshold is exceeded and begins to conduct, draining off charge from the diaphragm. You can test this by installing a neon lamp charge indicator and observe the flash rate as you increase volume level while playing, say, a 40Hz tone. While I am sure you are experiencing some compression/limiting due to transformer saturation, I’m not sure how much bias voltage compression(if any) you are experiencing. Some additional details on optimizing HV bias setting here: Measured my ESL attempt
Near-Field/Far-Field Effects: As you are aware, if you EQ a line source for flat response at a close distance, you will notice a falling off of bass level as you move further from the line source. Even if you EQ them for flat response at the listening position, the further distance to the walls of the room will affect the overall perception of bass impact coming from the ambience or room decay. If I understand your posts correctly, you have EQd for flat bass response at the listening position so this may not be as big a concern as if you had Eqd flat at 15cm. A good example of this issue was posted here: First ESL build - Full Range .
Dipole Power Response: For the same on-axis response, dipoles radiate 4.8dB less power into the room. Since a good portion of our perception of bass impact and quality comes from the room response, it makes sense that dipoles often sound less impactful in the bottom octaves. I’ve experienced this both with ESLs and dynamic woofer dipole subs. Applying a gentle boost at the bottom end starting at about 150Hz and ending with about +6dB at 30Hz makes dipole subs sound (to me) a lot more like a box sub, assuming the dipoles are actually capable of handling the boost without compression. There is obviously some psychoacoustics involved not just acoustic theory, and best sounding boost curve will be different for different rooms and line source heights. But, you might experiment a bit with it. I recall one of the best production dipole bass speakers I have heard was the Whisper by Legacy Audio using 4x15” per side. They included an analog bass shelving adjustment to tune the bass response to the room. It was interesting turn the knob while listening to music and hearing the bass go from tight/lean, to impactful, to full, and then bloated.
This last issue is probably part of the “circle of confusion” that Toole and others talk about, since most recordings are mixed using box speakers.
Roger Sanders explains ESL bass quality here:
Sanders Sound Systems - electrostatic | What's Best Audio and Video Forum. The Best High End Audio Forum on the planet!
Sanders Sound Systems - electrostatic | What's Best Audio and Video Forum. The Best High End Audio Forum on the planet!
The moving coil loudspeaker box excites more room modes because of its different radiation pattern and it probably distorts more. A well-known trick to make bass more noticeable is to distort it ( Missing fundamental - Wikipedia ).
Can one overcome or compensate for the high q in the bass range of esl’s with amplifier that has extremely high damping factor?
Crown for example has or had power amp with damping factor higher then 3000!
Crown for example has or had power amp with damping factor higher then 3000!
No, but you could in principle use some kind of inverse resonance filter, like a Linkwitz transform circuit.
Could that be done with a dsp? Thinking about geting med two hypex fa122 and they have dsp ”muscles”
While you can improve the sound and frequency response through equalization, you can't overcome the basic character of the high Q device. The panel would still ring when playing the resonance frequency, but at a lower level. If you are looking for the highest performance, this approach seems suboptimal.
You're likely going to have some resonance frequency drift over time as well as things relax or expand/contract with temperature. This could lead to misalignment of the correction (the more narrow and deep the filter, the more likely).
That's not to say that these are dealbreakers, just things to be aware of.
I'm no expert on this, but I believe you are correct in a minimum phase system.
With a high Q resonance like that of an ESL, I think difficulties are more likely though. Non-linear aspects of the system may interfere with the action of the filter, and I think nonlinear distortion effects related to the high Q resonance may remain even when the filter is working properly.