Thanks for going to the trouble of posting all those plots. I'm still confused. The last plot, 25 Hz, is like .4% THD and the spectrum seems to bear that out.
But there's 8% THD+Noise. I'd say 8% even "loud" up very close to the cone sounds like the usual distortion figure. Puzzling.
B.
But there's 8% THD+Noise. I'd say 8% even "loud" up very close to the cone sounds like the usual distortion figure. Puzzling.
B.
There is an easy explanation - the room starts rattling and brings the "NOISE" part of THD+N value up.
The SPL levels where the measurements took place are at the limit where the room rattling becomes noticeable.
Just watch the rising overall noise levels below the tested frequency from 10Hz to 20Hz. There is an affect on the higher part of the spectrum too although not as much.
At 25Hz the room is still quiet, at 20Hz the room gets into resonance mode (the longest dimension 35 ft brings 16Hz resonance), the resonance diminishes below 10Hz.
At and above 25Hz the room noise is very low and distortion harmonics really stand out above the noise.
The SPL levels where the measurements took place are at the limit where the room rattling becomes noticeable.
Just watch the rising overall noise levels below the tested frequency from 10Hz to 20Hz. There is an affect on the higher part of the spectrum too although not as much.
At 25Hz the room is still quiet, at 20Hz the room gets into resonance mode (the longest dimension 35 ft brings 16Hz resonance), the resonance diminishes below 10Hz.
At and above 25Hz the room noise is very low and distortion harmonics really stand out above the noise.
Last edited:
Again thanks for your cogent post.
But I remain skeptical that an 8" inexpensive woofer - Dayton Audio 8" DCS205-4 with a 32 Hz resonance - could have THD specs anywhere so vastly superior to what I am familiar seeing and esp. in the low range and loud.
B.
But I remain skeptical that an 8" inexpensive woofer - Dayton Audio 8" DCS205-4 with a 32 Hz resonance - could have THD specs anywhere so vastly superior to what I am familiar seeing and esp. in the low range and loud.
B.
Those are multiple 8" sub drivers measured at 4m in large room at listening position.
Here is my build thread "In-floor subwoofers - 8x8" drivers each, highest WAF factor, lowest THD":
In-floor subwoofer s- 8x8" drivers each, highest WAF factor
-
Here is my build thread "In-floor subwoofers - 8x8" drivers each, highest WAF factor, lowest THD":
In-floor subwoofer s- 8x8" drivers each, highest WAF factor
-
Last edited:
No, there are more then enough great implementations out there. Like with food only a few appreciate fine cuisine where most settle for mcDonalds. The DIYaudio logo top left states we are fanatics, not all are.
Unfortunately it is true. Go to the local electronics store. There's hardly any hifi left! Most of it is streamers, bluetooth speakers and wireless headphones. Apparently the majority of people is satisfied with the audio quality of those. Otherwise we would see MFB in every sub.
I can also imagine it being tricky to manufacture, because most OEM manufacturers don't offer standard MFB drivers (exceptions are there for Rythmik, Piratelogic and hopefully some more). This would instantly mean that a custom solution is required. This scares a lot of DIY people.
In post #13 I posted my basic reply to the OP question: "foundational idiocy" of sub drivers having to play outside their rational passband.
Another basic issue is the HiFi audio business which sells components rather than systems, mostly for good reasons. But MFB requires you purchase (or make) a powered sub system with a speaker, amp, and control system, not just drivers or speakers.
More puzzling is why audio enthusiasts are stupid about "investment" decisions? Although amps have been all but perfect since maybe 1985, people still fuss about them... and fuss about inaudible factors like cab edge diffraction and more.
Instead, neglected are major sources of imperfection such as cone speakers (instead of ESLs), room treatment, DSP fine-tuning of installed systems, etc.
You really got to wonder about the commitment to quality and rational decisions at a forum famous for advocacy of tapped "horns" and sixth-order tuned boxes.
B.
Another basic issue is the HiFi audio business which sells components rather than systems, mostly for good reasons. But MFB requires you purchase (or make) a powered sub system with a speaker, amp, and control system, not just drivers or speakers.
More puzzling is why audio enthusiasts are stupid about "investment" decisions? Although amps have been all but perfect since maybe 1985, people still fuss about them... and fuss about inaudible factors like cab edge diffraction and more.
Instead, neglected are major sources of imperfection such as cone speakers (instead of ESLs), room treatment, DSP fine-tuning of installed systems, etc.
You really got to wonder about the commitment to quality and rational decisions at a forum famous for advocacy of tapped "horns" and sixth-order tuned boxes.
B.
It is easy to agree with these objections. It is impossible to achieve quality using presently manufactured components.
First klippel controlled sound hardware is published but it is for clock radios and alike outside the hifi realm 🙂 Klippel and Nuvoton Unveil Smart Amplifier Solution with Integrated KCS Speaker Control Algorithm | audioXpress
Certainly brings feedback control to the mass market with ease of use.
"Developers will be able to connect and tune their own designs with this evaluation system for their target application. The initial identification of KCS speaker parameters is done almost automatically using the Klippel Analyzer 3 measurement system. The identification of a new speaker only takes a few minutes and the maximum working range of the device under test is identified automatically, providing comprehensive information about the suitability and potential quality of a speaker system."
Certainly brings feedback control to the mass market with ease of use.
"Developers will be able to connect and tune their own designs with this evaluation system for their target application. The initial identification of KCS speaker parameters is done almost automatically using the Klippel Analyzer 3 measurement system. The identification of a new speaker only takes a few minutes and the maximum working range of the device under test is identified automatically, providing comprehensive information about the suitability and potential quality of a speaker system."
Last edited:
Are we still disputing whether it is motional feedback or Genius Engineer Knows Best (after measuring a few parameter)?
Sure is still looking to me like the latter - just parameter pre-distorting.
Even though this doesn't look like authentic feedback to me, it has the usual "patent-pending" commercial goal: how to make a silk purse out of a sow's ear.
B.
Sure is still looking to me like the latter - just parameter pre-distorting.
Even though this doesn't look like authentic feedback to me, it has the usual "patent-pending" commercial goal: how to make a silk purse out of a sow's ear.
B.
There is no feedback from the moving mass. Any other form of feedback (current/voltage) is not good enough.
I've never heard servo controlled woofers and I don't know the field almost at all, certainly not sure what is good enough as a feedback from the driver. To me the KCS looks superior to mass feedback since if promises so much more but I'm not here to debate. It is nice to see some high tech entries in the very old and rather stagnant field of audio reproduction. I don't know if the KCS is a good thing with hifi subwoofer application or is it just a mean to make the cheap small speakers to sound a bit better.
Anyway, I'm fascinated by recent advancements in the field of AI so eagerly waiting applications in the field of audio reproduction. It is quite amazing how much information can be extracted from seemingly bad data, for example 😉 This Neural Network Restores Old Videos - YouTube
Anyway, I'm fascinated by recent advancements in the field of AI so eagerly waiting applications in the field of audio reproduction. It is quite amazing how much information can be extracted from seemingly bad data, for example 😉 This Neural Network Restores Old Videos - YouTube
Philips and more recently Celestion have both stated how extremely difficult it is to implement MFB. I know infinity had a servo controlled sub, although I don't recall the details. I know the technique would bring great benefits, I know eventually we will be using it, eventually! I would imagine computer management will have to be used to control the system to enable safe, adjustable and more importantly wonderful tight bass.
In the meantime we have the rather wonderful B&W PV1D and others. I admire the PV1D, it ticks all of the boxes, force cancelling driver mounting, excellent adjustments, and it sounds bloody good! Much better than my last sub around 20 years ago. (The music was better served with the sub switched off!)
In the meantime we have the rather wonderful B&W PV1D and others. I admire the PV1D, it ticks all of the boxes, force cancelling driver mounting, excellent adjustments, and it sounds bloody good! Much better than my last sub around 20 years ago. (The music was better served with the sub switched off!)
I think kipples point is that it is good enough to inform a model which predicts the excursion (amoung other parameters) once parameter identification has taken place on a sample speaker using a the kipple measurment system.
KLIPPEL Controlled Sound - Adaptive Algorithm - YouTube
As much as I dislike the Klippel approach and misleading promotion (many companies lie, some don't blush).... I suspect the parameter correction provides empirical distortion correction for that model of driver, not just basic modelling.
B.
B.