BudP said:
Bud,
When you talk of latency, are you referring to transient response?
And wouldn't distortion tests and CSD plots be showing what you term the threshold of latency?
I'm not sure I follow. How can adding mass improve the transient response?
Cheers
P.S. The test drivers have arrived at RAW Acoustics, and full documentation of testing, and testing itself has begun.
Theoretically, what you want to do is prevent a wave from reaching the edge at the same time. Normally we want the wave to reach the edge at a time that diffiffers gradually at each angled point. But tests have also proven that using felt on baffle, rounding baffle edges also provide some degree of improvement. If the EnABL pattern is used, the height needs to be investigated, but I think the pattern will change the time the surface wave reaches the edge, which is in line with what we whish to accomplish. Whether this is the most effective approach still needs to be determined. But we should accept that we have not seen any tests related with the baffle application that shows which is the most effective approach. I personally like to offset the driver as first consideration.Carlp said:OK guys, help me out here. Re: baffle edge diffraction, I keep hearing that EnABL can't have an effect b/c it's before the edge. But...
The bending of the waves at the edge would seem to create a changed pressure gradient, and that change seems like it COULD have an impact BEFORE the edge. (Wave diffraction along a shoreline certainly seems to have an effect before the bend in the shoreline).
Now I'm not saying John K or Dave or anyone else is wrong about the effects of EnABL (why would I subject myself to such merciless barrages), and I'm certainly not saying it would be audible. But for theory's sake, I have to wonder if it's actually true that diffraction can't be impacted by anything on the baffle BEFORE the edge. Doesn't make intuitive sense to me.
Carl
Daygloworange,
This will be the third time through this loop for me. I could not find a way to use classical physics to explain that small confusion either. Hence the thought that I was not improving any driver function but that the pattern was eliminating a short term resonance, that was interfering with the leading edge of all notes carried to the driver periphery, by transverse waves. That sort of made sense for frequencies whose wave length was shorter or close to the same length as the diaphragm was long, radially. And then all of the boundary layer thoughts were added. Now I am assured that it is a mass loading function and that adding mass does produce all of EnABL's functionality.
That the removal of latency and the lower levels for coherent decay are the results of an EnABL pattern is the first thing you will notice, upon hearing a treated speaker. While transient response is involved, as in short rise time signals, there is another rise time, for sine signals, that is also aided. So I use latency as the term to describe this. The second or third thing you will notice is that those transient and frequency balance characteristics that are noticeable, on axis, of an untreated driver, are available across the swept angle of the diaphragm, be it cone, dome or flat panel. It is these on axis signals that provide the limits of subjective performance, when the driver is properly EnABL'd
I initially thought so, but became confused by the scale of the event. My distortion tests showed some changes, some good some bad, but I had difficulty assigning what I found to the drivers involved. Some very skilled designers, in a subjective exposure, have pegged the drop in distortion of all kinds to a 20 dB level. My tests never showed anything approaching that level of reduction.
I suppose that threshold of latency is not a very good term for describing the decay of signals. Perhaps threshold of coherency might be better aimed, if no more findable in the CSD plots.
I find John K's blink comparison for a metal driver useful because it shows activity alterations across the frequency range tested. The scale of the changes worried me at first, but I have been assured that they are enough, or rather, all there is available. So, I just have to be patient and wait for other types of drivers to be tested, to see if they also show these characteristics, at these levels or greater, that I see on that blink CSD.
Unfortunately, sans techno babble and arm waving, these are the best answers I can provide.
Bud
This will be the third time through this loop for me. I could not find a way to use classical physics to explain that small confusion either. Hence the thought that I was not improving any driver function but that the pattern was eliminating a short term resonance, that was interfering with the leading edge of all notes carried to the driver periphery, by transverse waves. That sort of made sense for frequencies whose wave length was shorter or close to the same length as the diaphragm was long, radially. And then all of the boundary layer thoughts were added. Now I am assured that it is a mass loading function and that adding mass does produce all of EnABL's functionality.
That the removal of latency and the lower levels for coherent decay are the results of an EnABL pattern is the first thing you will notice, upon hearing a treated speaker. While transient response is involved, as in short rise time signals, there is another rise time, for sine signals, that is also aided. So I use latency as the term to describe this. The second or third thing you will notice is that those transient and frequency balance characteristics that are noticeable, on axis, of an untreated driver, are available across the swept angle of the diaphragm, be it cone, dome or flat panel. It is these on axis signals that provide the limits of subjective performance, when the driver is properly EnABL'd
I initially thought so, but became confused by the scale of the event. My distortion tests showed some changes, some good some bad, but I had difficulty assigning what I found to the drivers involved. Some very skilled designers, in a subjective exposure, have pegged the drop in distortion of all kinds to a 20 dB level. My tests never showed anything approaching that level of reduction.
I suppose that threshold of latency is not a very good term for describing the decay of signals. Perhaps threshold of coherency might be better aimed, if no more findable in the CSD plots.
I find John K's blink comparison for a metal driver useful because it shows activity alterations across the frequency range tested. The scale of the changes worried me at first, but I have been assured that they are enough, or rather, all there is available. So, I just have to be patient and wait for other types of drivers to be tested, to see if they also show these characteristics, at these levels or greater, that I see on that blink CSD.
Unfortunately, sans techno babble and arm waving, these are the best answers I can provide.
Bud
John K,
We call moisture in Seattle, air. You quickly learn that you were designed to be wet and ignore the condition.
I will post a color picture of my speaker system, with dimensions, tomorrow. But as a teaser, the front baffle face is a right angle onto the sides. The box sides have a 3/4" round over, from one to another, but not the baffle to the sides. I didn't make the box, just adapted it to my needs.
Bud
We call moisture in Seattle, air. You quickly learn that you were designed to be wet and ignore the condition.
I will post a color picture of my speaker system, with dimensions, tomorrow. But as a teaser, the front baffle face is a right angle onto the sides. The box sides have a 3/4" round over, from one to another, but not the baffle to the sides. I didn't make the box, just adapted it to my needs.
Bud
BudP said:
Actually Bud, based on what you state in the last paragraph quoted, flat frequency response and driver/system linearity over a wide dynamic range are exactly what you are looking for, in addition to the lack of transient smear in the crossover (it there is one). There are several features of percussion instruments that often get overlooked and lead to distorted results. Notes from pianos, drums, cymbals, bells, etc. really have two very separate aspects. First is the transfer of energy to the system by striking it. The seconds is the natural decay of the response of instrument with all its resonances and nuances. It is not at all that different from looking at the impulse response of a speaker system. The point being that if the impulse response of the speaker system isn't as close to perfect as possible then it will always impart its own colorations on the reproduction (regardless of the input). Severe breakup adds significant coloration to initial transients. Change that behavior and I can see why there could be a perception of a change in coherence. But it is the result of irregular frequency response. Speakers with elevated and irregular high frequency response typically are very dramatic sounding when auditioned with percussion instruments because that initial impact contains much high frequency energy and the sustained part of the note is typically composed of a lower fundamental and harmonics. The elevated high frequency response make the initial strike sound very impressive with typical listen response being that the system is very dynamic when in truth it is noting more than very unbalanced. This is not uncommon with full range driver systems where the desire for direct connection to an amplifier results in ignoring even rudimentary correction for the baffle step. If you want coherence, focus flat frequency response and on impulse or step response and make them look like the ideal impulse for a system with the same low frequency cut off. Then worry about linearity over as wide a dynamic range as is required.
some basics
There is another thread forming which may serve in getting us on the "same page":
Geddes et al on measuring loudspeakers
There is another thread forming which may serve in getting us on the "same page":
Geddes et al on measuring loudspeakers
Re: some basics
Highly unlikely. I was going to suggest this thread when it started there, but if Geddes stays highly involved (and I hope he does), much if not most of that thread will be contrary to positions of many here. Some key points Geddes makes early on:
"Once you have the impulse response then interpretation becomes the issue. If you can't get an impulse response that is believable, repeatable and clean, then thats the first goal."
"In theory any and all signals MUST yield the same impulse response. "
"Don't worry about the signal type, use whatever gets you an impulse response."
Considering that there was so much questioning of John's use of the impulse response and its implications coupled with the "need" to use some other form of measurement posed by a few (that continue to this point), that thread will not be considered useful by many here. Certainly not by most of the subjectivists, as Geddes also is adamant about the need for a broad set of on- and off-axis measurements. If I'm proved wrong, that would be very pleasing.
Dave
Ed LaFontaine said:
Highly unlikely. I was going to suggest this thread when it started there, but if Geddes stays highly involved (and I hope he does), much if not most of that thread will be contrary to positions of many here. Some key points Geddes makes early on:
"Once you have the impulse response then interpretation becomes the issue. If you can't get an impulse response that is believable, repeatable and clean, then thats the first goal."
"In theory any and all signals MUST yield the same impulse response. "
"Don't worry about the signal type, use whatever gets you an impulse response."
Considering that there was so much questioning of John's use of the impulse response and its implications coupled with the "need" to use some other form of measurement posed by a few (that continue to this point), that thread will not be considered useful by many here. Certainly not by most of the subjectivists, as Geddes also is adamant about the need for a broad set of on- and off-axis measurements. If I'm proved wrong, that would be very pleasing.
Dave
Re: Re: some basics
Based on my experience, different signals do make a difference in measurements under certain conditions, but if one only indulges in the idealized mathematical aspects of a non-ideal world conditions, then maybe this will be true. I have never seen any mathematical prediction that exactly matches real world measurements. The most we can hope for is not so much error that it will invalidate a design concept.dlr said:
Re: Re: Re: some basics
You make my point. We are not likely to "get on the same page". Your post is nothing but a couple of straw-man arguments that really don't mean anything. Geddes did qualify his points by saying "in theory", he's still in the real world. No one has suggested "indulging" in the idealized. Maybe you should now drop out of the measurement thread, since you have such obvious disdain for the theory behind measurements and the insights that measurements provide. I'm surprised that you measure at all.
As an aside, consider that the system you use (SoundEasy) is based on theory, the MLS. I believe that the MLS mathematical basis was determined before the first MLS measurement system (MLSSA) was designed and implemented. Theory did come first. It was not stumbled across by accident.
Sorry, Ed, it didn't take long, did it?
Dave
soongsc said:
You make my point. We are not likely to "get on the same page". Your post is nothing but a couple of straw-man arguments that really don't mean anything. Geddes did qualify his points by saying "in theory", he's still in the real world. No one has suggested "indulging" in the idealized. Maybe you should now drop out of the measurement thread, since you have such obvious disdain for the theory behind measurements and the insights that measurements provide. I'm surprised that you measure at all.
As an aside, consider that the system you use (SoundEasy) is based on theory, the MLS. I believe that the MLS mathematical basis was determined before the first MLS measurement system (MLSSA) was designed and implemented. Theory did come first. It was not stumbled across by accident.
Sorry, Ed, it didn't take long, did it?
Dave
Oh, let's just face it. We can never do the same thing exactly over again. I mean the state of the universe is constantly changing so since everything is an initial value problem and since the initial conditions are always different on some level, nothing ever happens the same way twice. What the #, I'm going to Disney Land. 
I don't follow this thread any longer, so not sure if this has already been posted but Mark K has an interesting view point of the Enable thang up on his website:
http://www.audioheuristics.org/
http://www.audioheuristics.org/
There are just different ways of proving things. Like oriental medicine. The only proof that was needed for thousands of years was that "it worked", and the explanation is made using the oriental way. For western medicine systems, you have to know why it works in a western way of understanding. Not many people in western medicine will agree with oriental medicine reasoning method, but once they find a way to verify it using their own way, then they understand and agree.
So really, I think if EnABL works, then the people that are interested just have to find thier own way of understanding why and how to make the most out of it.
So really, I think if EnABL works, then the people that are interested just have to find thier own way of understanding why and how to make the most out of it.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.