Ben,
As an OEM (original equipment manufacturer), I could buy the IPAL system, which would be completely impervious to the "DANGER from plastic straws", and far better than anything I could possibly come up with doing completely by myself.
The IPAL is also measurably better than any DIY MF system you have posted, or I have seen thus far.
But, as I have stated before, the little bit better response that it affords over good B&C drivers is not worth the cost to me.
Cheers,
Art
There is quite a bit of explanation and proof on Powersoft's website:
IpalMod - Powersoft Audio
IIRC, Martin Audio makes an IPAL sub-woofer (among several others that don't come immediately to mind), you could look at their measured responses.
Just dozed off at the computer, I'll let you find the proof on your own, there is plenty available on the interwebs if you search and read.
Cheers,
Art
You made the claim so it is up to you to defend it or to apologize to esgibt and to all forum members for posting rubbish if your remark is indefensible.
Just because you see no application of MF speakers in your commercial sound reinforcement business doesn't give you license to damn the entire R&D activity as you do.
Ben
Last edited:
I agree that weltersys has to come up with the proof, if out there at all.
I won't deny that LS systems based on more conventional techniques can achieve brilliant results in reproducing bass... I've heard enough examples over the years. But they can't match a properly designed and build MF system, especially not if you compare cabinet sizes, bandwidth and distortion.
Ben,
I have pointed out that the IPAL system is presently seeing a fair amount of traction in the (not my own personal) commercial sound reinforcement and playback business, which in no way damns anything.
Pointing out a product and suggesting that you read about it, since you appeared to have far more interest in the subject than I, in no way requires me to defend it, or apologize for it's introduction to you years ago.
I simply pointed you in the direction of the most "state of the art" MF system commercially available, and you can pooh-pooh it or learn about it, you have made your choice apparent, so "my work here is done :^)".
Cheers,
Art
Last edited:
Hi esgigt,
I don't think that the IPAL system falls under "LS systems based on more conventional techniques"?
Following the link that weltersys posted in Post #64 I found AES Convention Paper 8501 under downloads. You may already have read it. I found it an interesting read. I find the differential pressure sensor approach as a feedback sensor interesting.
Regards,
I don't think that the IPAL system falls under "LS systems based on more conventional techniques"?
Following the link that weltersys posted in Post #64 I found AES Convention Paper 8501 under downloads. You may already have read it. I found it an interesting read. I find the differential pressure sensor approach as a feedback sensor interesting.
Regards,
Is that what I stated??? No...
C'mon... do I have to search a website for the "proof" Weltersys fails to point at directly? Don't think so... he's not my boss. And I've got more useful things to do.
As do I, presently have three low frequency designs using 9 drivers in the build que to service my shop, home stereo, and live sound reinforcement systems. Then there are two different virtual point source designs to finish after that, and a stage monitor design to give hearing to the deaf...
To more important things:
Cheers,
Art
Yes you do ! ^^
But not because of knowledge competition, but because of your own curiousity, and your generosity for those like me that want to know too ^^
Presently at work, got two young child, a wife (!), a dog, lot of work on house and gardening, a 3 way unity project, a tube amp project, and a bass guitar project, and fresh beer waiting me at home.Yes. The commercial success of IPAL speaks for how effective it is. The burden of proof is yours to show it doesn't work as claimed.
only if I stated that it didn't work as claimed... Well, I didn't. So if you're interested in knowing, research it for yourself
I’ve had a careful read of the IPAL AES paper describing differential pressure feedback and a negative impedance output Class D amp, and I offer the following thoughts.
Anybody who eyeballs the benefit of MF in the tone bursts of the article, will immediately say, “Does my sub play tone bursts as awful as THAT? Gotta have MF”.
I suspect the differential pressure gizmo is of no additional value as compared to other means of tracking output. It is a naive approach to try to measure raw acoustic output. For this system, I think the instantaneous DSP (their term and obviously nonsense) is finagling the time course of the music signal and the feedback signal according to freq and that free-dependant DSP time adjustment allows IPAL to get around the problem of Nyquist phase alignment when the feedback signal is delayed in the air and the phase is all screwed up. (And that is why I am looking into using an Arduino or a Raspberry Pi microprocessor in my next MF trials.)
Long ago at Bell Labs, I worked with tone bursts and impulses. MF really looks great with that kind of testing, as we see in this article. But what is challenging for quants like me is that tone bursts make very impressive pictures but can’t be readily quantified. So I’ve been looking for some means of quantifying the benefits of MF, besides, as in this paper, posting pictures. And any R&D is easier when you have quantitative progress.
Another approach to quantifying MF is to plot the negative output impedance of the amp. Hard to assign a measure of quality to that, but pretty impressive to have negative output impedance, eh.
The tone burst data show how awful bass reflex boxes are. Every picture, from low to high tone burst frequency shows ringing at the box resonance, although greatly controlled when their MF is applied. There is no over all freq response published. Anybody find that an odd omission? I am not surprised since MF does not linearize BR output, although anyone can see the dramatic improvement in the tone bursts. (The authors mention that the scale of the tone burst pictures wanders around meaning there is no improvement in flatness or what is fare worse compared to other MF boxes, no bass extension.)
My guess is that whoever is bankrolling this R&D told the researchers, “You guys have to make it work for BR boxes like you find in dance clubs and other rich customers”.
If you examine the tone bursts, you’ll find that accurately capture the weird roller-coaster cone activity you get with BR boxes. You can see the peculiar way the speaker output relates to the music signal. They’ve chosen the three key BR freq as well as showing a few higher freq points which are contaminated with the BR box resonance freq. See for yourself.
The authors mention that with a Rice-Kellogg driver, the whole game is the BL product (motor strength) that - in relation to moving mass - is what you want to optimize. Folks who play with sims late into the night think BL product is just another factor that you plug into your sim and explore the value in different drivers till the output seems right.
Well, as a guy who drove a Lotus for a few years, motor horses in relation to vehicle weight is the core parameter for motion. You can buy a Lotus and have a modest-sized engine and a fabulously light body or your can buy a Corvette and have a brutish old fashioned engine pushing a very heavy body. Either can be quite speedy but I sure know which makes automotive and aesthetic sense to me. Likewise for cone drivers.
The paper and aspects of the design are presented in the usual consciously dishonest fashion of patent applications. Really stinks badly of commercial motivations and is all but unreadable as well.
Ben
Anybody who eyeballs the benefit of MF in the tone bursts of the article, will immediately say, “Does my sub play tone bursts as awful as THAT? Gotta have MF”.
I suspect the differential pressure gizmo is of no additional value as compared to other means of tracking output. It is a naive approach to try to measure raw acoustic output. For this system, I think the instantaneous DSP (their term and obviously nonsense) is finagling the time course of the music signal and the feedback signal according to freq and that free-dependant DSP time adjustment allows IPAL to get around the problem of Nyquist phase alignment when the feedback signal is delayed in the air and the phase is all screwed up. (And that is why I am looking into using an Arduino or a Raspberry Pi microprocessor in my next MF trials.)
Long ago at Bell Labs, I worked with tone bursts and impulses. MF really looks great with that kind of testing, as we see in this article. But what is challenging for quants like me is that tone bursts make very impressive pictures but can’t be readily quantified. So I’ve been looking for some means of quantifying the benefits of MF, besides, as in this paper, posting pictures. And any R&D is easier when you have quantitative progress.
Another approach to quantifying MF is to plot the negative output impedance of the amp. Hard to assign a measure of quality to that, but pretty impressive to have negative output impedance, eh.
The tone burst data show how awful bass reflex boxes are. Every picture, from low to high tone burst frequency shows ringing at the box resonance, although greatly controlled when their MF is applied. There is no over all freq response published. Anybody find that an odd omission? I am not surprised since MF does not linearize BR output, although anyone can see the dramatic improvement in the tone bursts. (The authors mention that the scale of the tone burst pictures wanders around meaning there is no improvement in flatness or what is fare worse compared to other MF boxes, no bass extension.)
My guess is that whoever is bankrolling this R&D told the researchers, “You guys have to make it work for BR boxes like you find in dance clubs and other rich customers”.
If you examine the tone bursts, you’ll find that accurately capture the weird roller-coaster cone activity you get with BR boxes. You can see the peculiar way the speaker output relates to the music signal. They’ve chosen the three key BR freq as well as showing a few higher freq points which are contaminated with the BR box resonance freq. See for yourself.
The authors mention that with a Rice-Kellogg driver, the whole game is the BL product (motor strength) that - in relation to moving mass - is what you want to optimize. Folks who play with sims late into the night think BL product is just another factor that you plug into your sim and explore the value in different drivers till the output seems right.
Well, as a guy who drove a Lotus for a few years, motor horses in relation to vehicle weight is the core parameter for motion. You can buy a Lotus and have a modest-sized engine and a fabulously light body or your can buy a Corvette and have a brutish old fashioned engine pushing a very heavy body. Either can be quite speedy but I sure know which makes automotive and aesthetic sense to me. Likewise for cone drivers.
The paper and aspects of the design are presented in the usual consciously dishonest fashion of patent applications. Really stinks badly of commercial motivations and is all but unreadable as well.
Ben
Last edited:
Well now that sounds like a poke at sim users and specifically at me, because who else played around with Bl recently with different drivers until the output seemed right?
If you are in fact referring to the large coil lossy inductance study I performed, here's the facts. It can (and has) been statistically proven that these drivers with large amounts of lossy inductance don't sim properly while drivers that don't have large amounts of lossy inductance sim quite accurately.
Lossy inductance is not a new idea, this is why Wright and Leach developed complex inductance parameters to more accurately simulate the effect. This wasn't a big deal though until recently with the uprising of several manufacturers producing ultra high excursion drivers with massive motors and voice coils that are particularly susceptible to this lossy inductance issue. With these large coil drivers it is clear to see that the lossy inductance decreases the apparent motor strength by around 30 percent compared to what the simple t/s parameters would suggest.
How do you propose to account for that? Using complex t/s parameters might work (or it might not, I haven't tried) but these parameters are never published and none of the best simulators are set up to use them as inputs. So I devised my own method. How do you account for decreased motor strength? The only parameters that you could adjust to compensate are Re and/or Bl, and Bl gets right to the root of the problem with little or no side effects.
My large coil lossy inductance adjustment works, it's been empirically proven. it works 100 percent of the time for drivers of this type regardless of enclosure type.
Here's just one example, the latest example. This horn was designed by LTD02 using the principles that recognize and correct for the large coil lossy inductance. The horn was built and measured by asarose. The sim matches the measurement very well. A normal sim that does not account for this issue does not match the measurement well at all. And I have 30 more examples just like this one.
If you have any proof that this issue doesn't exist, or that there is a better way to deal with it, or even that the way I've dealt with it doesn't work, please provide it instead of posting backhanded insults. Most of us here prefer to deal with things in scientific fashion and take things a bit deeper than the "audiophile" buzzwords and catchphrases that you are fond of using in lieu of actual data to describe issues that you don't fully comprehend.
As for the rest of your ramblings, 100 percent of your subs past and present do NOT employ motional feedback and suffer from all the problems that you constantly rail against, yet you have been very satisfied with them. The issues (like BR ringing) that you consider incredibly important are not audible for the most part. As I've told you, most of what we hear is frequency response, if you design a BR with the same frequency response as a sealed box, it will sound like a sealed box. Do you really think the ringing of a BR is a major audible problem compared to the absolute horror show that the room effects are going to have on both your frequency response and time smearing? Room modes ring a lot longer and stronger than BR ports.
Cleaning up your tone bursts is not going to be nearly as audibly effective as doubling up on your drivers and power, that will do a lot more to cut audible distortion down than MF ever will.
Regardless of how you think things should be done (with cars or audio products) there are products available that work very well. Not surprisingly the things that work do their job well precisely because they are NOT engineered the way you think they should be.
Last edited:
Ben,
Your reading does not seem to be very "careful", for if it were you may have understood the following:
1) I read the article and did not say either. Certain aspects of a speaker's response, such as tone bursts, can "look" awful, but sound just fine.
2) Publishing a frequency response for a system that inherently can achieve any frequency response desired (within the limitations of the driver's linear response) would be redundant.
3) Powersoft, as primarily an amplifier manufacturer, found it both in their, and their customer's best interest to provide an Adaptive (the "A" in IPAL) platform that will work equally well with any speaker design, though the optimum TS parameters for working "adaptively" are quite different from "run of the mill" woofers. In this case, since the customers are OEM (original equipment manufacturers) it would have been design suicide to create a product that was limited to one cabinet (or driver manufacturer) design format.
Since the release of the IPAL system, Powersoft has developed and patented the M-Force, a moving-magnet low frequency device that uses a similar drive system, though obviously set up for the vastly different TS parameters entailed by a moving magnet compared to a moving coil.
Though the moving-magnet device is impressive, and capable of moving from 22" to 40" diaphragms with great authority, and has received rather high praises in many reviews, it has not cut (much) into the market share of the drivers produced by B&C for the original IPAL pairing.
Cheers,
Art
Attention REW and all those using microphones and objective test methods: throw away your lab equipment.
B.
Last edited:
A little bit of research and practical experience will give a user some practical knowledge about what the lab equipment is showing you and what is important and what is not. We've been trying to tell you this for years now.
A perfectly shaped tone burst is not necessary. Ability for a speaker to accurately reproduce a square wave signal is not necessary. There are a lot of things that can sound very good but fall well short of the mark in zoomed in lab measurements.
It's up to the user to determine what's important and what is not. This will require at least a bare minimum of research and experience.
A perfectly shaped tone burst is not necessary. Ability for a speaker to accurately reproduce a square wave signal is not necessary. There are a lot of things that can sound very good but fall well short of the mark in zoomed in lab measurements.
It's up to the user to determine what's important and what is not. This will require at least a bare minimum of research and experience.