heres another freak show" to ponder...
Had some cheap 8 inch poly cone radio shack woffers once . Low level detail was far from good. As well bass and mid bass were soft and undynamic no mater what box they were in.
At the time I was doing some workon the effect of a drivers suspensions mechanical properties and wanted to see the effect of a more "linear spring" property. So I cut the spiders out and replaced them with a cross hair suspension of just 2 nylon threds with adjustable spring loading to get Fs back to exact same spot as with spider.
The results were imediatly noticable and very similar to your experience. Livley dynamic bass. No other mod was made. even went back and added a ring of hot melt glue close to VC to simulate mass lost from removal of spider.
FR looked very close to same once Fs adjusted. How do we explain this??
At first I thought the new suspension was simply holding the VC in its gap better. Problem is this "sound" was precieved at low volumes as well.
Had some cheap 8 inch poly cone radio shack woffers once . Low level detail was far from good. As well bass and mid bass were soft and undynamic no mater what box they were in.
At the time I was doing some workon the effect of a drivers suspensions mechanical properties and wanted to see the effect of a more "linear spring" property. So I cut the spiders out and replaced them with a cross hair suspension of just 2 nylon threds with adjustable spring loading to get Fs back to exact same spot as with spider.
The results were imediatly noticable and very similar to your experience. Livley dynamic bass. No other mod was made. even went back and added a ring of hot melt glue close to VC to simulate mass lost from removal of spider.
FR looked very close to same once Fs adjusted. How do we explain this??
At first I thought the new suspension was simply holding the VC in its gap better. Problem is this "sound" was precieved at low volumes as well.
Last edited:
Well, maybe the new spider has the same stiffness to result in the same Fs, but is more progressive?
I know from tape recordings that certain compression effects can result in dynamics being perceived as being better than the original, giving a tighter bass sound, but for all we know it was more linear
I'm guessing all we can do is make a modification and keep a set of original drivers to compare measurements against.
Besides being time consuming, it's also pretty expensive, especially for these kind of far-out experiments so currently it's more bla bla than scientific.
The way I try to make it work for the coming months is compare a modified jbl 2235 to a standard 2245. If it's about equivalent in quality and quantity going further financially is no problem as the difference between those 2 unmodified is bigger than expected (18" wins hands down imho).
For now it's the only way to keep a set that plays and have as reference while not investing too much in prototyping and experimenting.
How exactly did you build the nylon suspension? Did you try and dampen or change the spring?
I know from tape recordings that certain compression effects can result in dynamics being perceived as being better than the original, giving a tighter bass sound, but for all we know it was more linear
I'm guessing all we can do is make a modification and keep a set of original drivers to compare measurements against.
Besides being time consuming, it's also pretty expensive, especially for these kind of far-out experiments so currently it's more bla bla than scientific.
The way I try to make it work for the coming months is compare a modified jbl 2235 to a standard 2245. If it's about equivalent in quality and quantity going further financially is no problem as the difference between those 2 unmodified is bigger than expected (18" wins hands down imho).
For now it's the only way to keep a set that plays and have as reference while not investing too much in prototyping and experimenting.
How exactly did you build the nylon suspension? Did you try and dampen or change the spring?
That's my take on this whole conversation. No data has been presented, only "guesses" as to why something "sounds better", whatever that means.
I would agree that larger loudspeaker system have a huge advantage, but it's a necessary condition, not a sufficient one.
I don't worry too much about how woofers work around their resonance - this all gets swamped out by the room in practice. That's why I said that it is the changes in the crossover that are most audible with amplifier output impedance.
Is this not one of the biggest points in active systems - no passive components to ruin the output impedance
Hello,
The cabinet volume should change drastically the loudpeaker driver electrical and mechanical impedance curves.
Are you aware of any study talking about the room volume effect on those curves please ?
Is this not one of the biggest points in active systems - no passive components to ruin the output impedance
Passive components impedance is not a problem if accounted/designed for.
What is a problem is designing for one impedance and the amp having something different.
You missed my point: The room will dominate the room response making any changes in the resonance impedance curve irrelevant.
heres some more guessing....
Another woofer, this time a 15 inch cheap poly cone unit. Again with bass that sounded, well, slow and flabby in a bunch of different loading's from seal to ported to IB and no matter what crossover point or slope AND even active.
The surround was anything but linear spring in nature and I suspect chosen to get a reasonably well terminated edge for midrange use.
Applyed a thin layer of silicone calk to surround giving it a more spring like property.
Fs went from 20hz to high 20s. Same exact sonic change as the 8 inch cone with tensioned string spider.
Look I get it. I work in the scientific community as a design engineer. I understand the value of painstakingly collected data. Its a very very long hard road to really prove some things out. Been their many times and the "guessing" of those less involved and informed can be nothing but a frustration, especially when its the boss
However I have seen this phenom in a bunch of transducers over the years. Not just cones but with large area planers as well.
I in no way suggest that I know whats going on, only that I have seen this happen so many times over so many years and within so many different designs that it seems worth mentioning.
BTW the most dynamic planer I have ever heard by far was one I built that had absolutely NO tension at all and it was HEAVY compared to the typical designs. VERY low sensitivity. Had to turn it up quite a bit to get it to shine but Damn it was so convincing on drum transients. Truly special. Its CSD was very different than a light tensioned film design.
Another woofer, this time a 15 inch cheap poly cone unit. Again with bass that sounded, well, slow and flabby in a bunch of different loading's from seal to ported to IB and no matter what crossover point or slope AND even active.
The surround was anything but linear spring in nature and I suspect chosen to get a reasonably well terminated edge for midrange use.
Applyed a thin layer of silicone calk to surround giving it a more spring like property.
Fs went from 20hz to high 20s. Same exact sonic change as the 8 inch cone with tensioned string spider.
Look I get it. I work in the scientific community as a design engineer. I understand the value of painstakingly collected data. Its a very very long hard road to really prove some things out. Been their many times and the "guessing" of those less involved and informed can be nothing but a frustration, especially when its the boss
However I have seen this phenom in a bunch of transducers over the years. Not just cones but with large area planers as well.
I in no way suggest that I know whats going on, only that I have seen this happen so many times over so many years and within so many different designs that it seems worth mentioning.
BTW the most dynamic planer I have ever heard by far was one I built that had absolutely NO tension at all and it was HEAVY compared to the typical designs. VERY low sensitivity. Had to turn it up quite a bit to get it to shine but Damn it was so convincing on drum transients. Truly special. Its CSD was very different than a light tensioned film design.
These days there is no excuse for not taking data. Without data you just do not know that anything is changing regardless of what you think you hear. The ear is a terrible measuring instrument - well the ear itself is fine, it's the fact that it takes a brain to interpret what the ear sends that is the problem.
But, i'm curious to see if the room "domination" measurability impact the membrane and its internal air volume damping and if it can have a silght impact ont the electrical impedance curve.
agree,
back in those days I had only my ears and my wife as a check. Her lack of interest/knowladge/bias as a check on my perception. Armed with no info she was to listen for a bit ( always same material, same position , same room layot etc) and simply tell me if there was anything that stood out to her good bad or ugly. Quite consistantly IF there was somthing solidly different she would make a coment consistant with my perception. Are agreenment was to never ever talk about the results and her thoughts were to be written down for my later study, not in her presence.
That looks objective, despite the tone, but let's see.
So one guy comes up to you and changes 1 part in his HiFi set, he gets his feet tapping, he's on his instant dance floor all night all of a sudden and you call him out for having not done his measurements straight away?
That's rude and uncalled for.
Maybe next time a musician tunes his instrument you can go up to him and tell him he needs a Clio, some low budget electret microphone and proof to you he's worthy of your listening?
Putting it down with the words that it's all "guesses" and "data has not been presented" omits the most important part of research which is getting to the justification to start the time consuming and expensive proofing stuff, many times before that decision you sharpen your theoretical model by means of discussion.
This is diyaudio, this is where people, mostly beginning with a low budget or inexperienced theoretical framework start their thought process, their experiments and could use some help in getting an idea of the physics that are behind this. If listening to changes in a parameter means nothing to you then I'm guessing you're missing something important, besides accepting that a listening session can be an intermediate step.
There are no worthwile lengthy discussions with your wife about this field, students don't have the experience. What you can do when you don't trust your ears is take a few musicians home, I'm betting they can tell you things your dut didn't show on your graphs. I can tell and it makes you look at different angles to those so called "objective" measurements.
All commercially available speakers have been through enough measurements and many sound just good enough to listen on higher levels for about half an hour or so.
Would they have sounded so much better in the dynamics department when the engineer would have taken 100 hours more for measurements? Nope.
You can't even accurately predict what a speaker will sound like from any set of measurement data if you leave out the size and type of drivers of that speaker, only hope your modifications will get you a better result than the one you had last time. It's extremely relative, edging subjectivity calling out for data on the net.
It's a many sided blade. In the end all one can accept as final proof is the way it works on your brain. It does beat measurements and has absolute meaning, it just shouldn't be your only means evaluating a design.
Absolutely. A speaker could be seen as a marriage or life itself for that matter;. So much easier with enjoyable characters in it, though it's no guarantee you'll do well, actually enjoy it or be good in what you like. In short: it is a synergetic system and needs a lot of attention to detail, time and effort.
Just an example:
A JBL 2446J (large format compression driver) has an impedance change of around 100% in it's usable frequency range. When passive filtering is applied by means of transformer or l-pad, the variation the impedance variation the amplifier sees is many factors lower in it's usable bandwith from around 1KHz to 20 KHz and e.g. it's frequency range goes down by redistribution of power on both ends of the spectrum because of impedance going up on those ends.
None of these factors change much or at all as a result of output impedance by the amplifier.
However, compared to a direct ac coupled active version the speaker impedance fluctuations can result in substantial amplitude variations as a result of changes in amplifier output impedance and/or dc resistance of cables. Also, on the outer edges of its bandwidth the amplitude goes up instead of down and it is dependant on the output impedance.
For woofers you have, in general, the woofer's own resonance and the resonance frequency of the port to deal with.
Broad generalisation that output impedance only has substantial influence on passive loudspeakers is not applicable here, it depends on many factors and it is extremely easy to measure.
What's your use case to state it isn't?
I would not expect a huge change in the impedance curve due to the room. It would likely only happen in a very low damped room at the lowest modal frequencies. In a typical room the damping will likely be high enough that the rooms influence on the impedance curve would be small. But I have never seen this actually tested so I guess that I am "just guessing".
Any procedural attempt at curbing a subjective impression bias in listening tests is going to be a vast improvement. Not sharing opinions with each other is critical since I have witnessed two people convincing each other of things that just were not true. Had they not conversed there is a much better likelihood that one of them might have noticed the truth, but even that is not foolproof.
This is the same old argument: "If it sounds good to me, then it is good."
I don't accept that as the way to achieve optimal sound reproduction.
I wasn't talking about compression drivers.
Saying: "For woofers you have, in general, the woofer's own resonance and the resonance frequency of the port to deal with." - isn't very informative.
"Broad generalisation that output impedance only has substantial influence on passive loudspeakers is not applicable here, it depends on many factors and it is extremely easy to measure." It is applicable and I'll stand by my claim.
Regarding Mterbekke vs Gedlee disagreement: I think you are pursuing different goals.
One goal is to *build* a *unique* system, sounding *subjectively* better than anything a particular *individual* heretofore heard.
Another goal is to develop a *theory*, serving as a solid foundation for *designing* systems maximizing quality to price ratio in *mass producton*, as judged by *collective* purchasing decisions of tens of thousands to millions of people.
I’m saddenned by the bitterness Dr. Geddes still exhibits due to the lack of wider understanding and acceptance of his contributions to the art and science of loudspeaker design.
Apologies in advance for my next statement, as it may seem to imply that I’m wiser than Dr. Geddes. I wish I were I’m just more familiar with certain things.
I think Mr. Geddes doesn’t realize that such reaction is typical when society encounters a paradigm shift.
It usually takes a generation - about 26 years - for the society as a whole to embrace it. It takes about half a generation for the brightest minds to start appreciating it.
I think that the recent resurgence of interest in, and introductions of more properly designed waiveguide-based products, highly praised for their sound quality yet moderately priced, e.g. from JBL and Apple, are reflections of the 13-year first wave of appreciation of the work done by Dr. Geddes and others in this area.
Perhaps the trick for Dr. Geddes is to survive for at least the next 13 years, while keeping his mind sharp, to be ready to deliver that speech at a lifetime achievements award ceremony?
I’m not joking. The approaches and math in his published works are beautiful. They have that finality quality to them, which will be regarded as classic by more and more people as time goes by.
One goal is to *build* a *unique* system, sounding *subjectively* better than anything a particular *individual* heretofore heard.
Another goal is to develop a *theory*, serving as a solid foundation for *designing* systems maximizing quality to price ratio in *mass producton*, as judged by *collective* purchasing decisions of tens of thousands to millions of people.
I’m saddenned by the bitterness Dr. Geddes still exhibits due to the lack of wider understanding and acceptance of his contributions to the art and science of loudspeaker design.
Apologies in advance for my next statement, as it may seem to imply that I’m wiser than Dr. Geddes. I wish I were
I’m just more familiar with certain things.I think Mr. Geddes doesn’t realize that such reaction is typical when society encounters a paradigm shift.
It usually takes a generation - about 26 years - for the society as a whole to embrace it. It takes about half a generation for the brightest minds to start appreciating it.
I think that the recent resurgence of interest in, and introductions of more properly designed waiveguide-based products, highly praised for their sound quality yet moderately priced, e.g. from JBL and Apple, are reflections of the 13-year first wave of appreciation of the work done by Dr. Geddes and others in this area.
Perhaps the trick for Dr. Geddes is to survive for at least the next 13 years, while keeping his mind sharp, to be ready to deliver that speech at a lifetime achievements award ceremony?
I’m not joking. The approaches and math in his published works are beautiful. They have that finality quality to them, which will be regarded as classic by more and more people as time goes by.
Balanced - you must have signed up just to make that post!! But thanks.
I do feel frustrated by the continued insistence on things that are simply not true, but perhaps it comes across as bitter. That the masses take a long time to accept a paradigm shift is well known to me.
You will be happy to know that after some 20 years, a software company is going to release measurement software that has the GedLee Metric (Gm) measurement in it, so what you say is so true. I can only hope that I will be around when people accept my work on waveguides as a paradigm shift.
I do feel frustrated by the continued insistence on things that are simply not true, but perhaps it comes across as bitter. That the masses take a long time to accept a paradigm shift is well known to me.
You will be happy to know that after some 20 years, a software company is going to release measurement software that has the GedLee Metric (Gm) measurement in it, so what you say is so true. I can only hope that I will be around when people accept my work on waveguides as a paradigm shift.
I would be very interested in this software as well and the reference material too. Not originating from the audio industry, I tend to approach audio from measurement and my own perception trying to find a correlation. The basic work is focused on the playback system as a whole which started from the speaker because data seemed to show it was where most improvement can be made. Since then, many parts of the playback system seemed flawed from my field of engineering point of view, so I started tracing the effects deeper into each aspect as well as upstream to the DAC. While I had suspected DAC to have the least effect, recent measurements seem to tell a different story as well. So it seems to me, the general measurement practices are not able to tell the whole story. While in the process of setting up some additional test rigs, I am seeing possibly the computer operation system have differences as well, while this not final whether the differences are from the computer or OS difference, information seems to lead to the OS, but I may be wrong. While trying to sort this out since it will effect my measurements, I am sure I will discover more. But the main point is there has to be a metrics to address various characteristics of the audio nature. Distortion, noise, natural response as well as forced response. Not all of these can be addressed currently at different component levels, so there is a lot of work to be done to revolutionize how measurement is conducted at various levels. Every difference audiophiles tweak and hear are true to different orders of magnitude, the audio industry had just ignored such due to immediate financial benefits. This is sad for any industry because only with such incremental improvements will a jump in recognition occur.
I agree fully. Nobody stated explicitly listening is the only way to go, but indeed nobody mentioned taking measurements straight and all the way either. Later in the thread it was mentioned measurements should be the next step. I can't tell people how they diy and the positive effects on using measurement equipment are common knowledge. Either way, calculating cross over parts is always a good place to start and everybody can go on from there the way they seem fit or enjoyable.
But following this law of ears can't tell: how can anybody proof nobody can hear lower orders of high harmonic distortion if the only proof is found by listening to the system? This is really hard to follow, if not impossible.
Look, I'm really not trying to be a nuisance here. I'm happy you're here and I respect the work you have done.
I understand that you worked countless years figuring out how to get drivers to integrate as much as possible in the frequency and/or time domain and with as much as possible similar polar response throughout etc etc. Or whichever way you why went. I respect that.
But I don't understand at all where you're coming from when someone changes 1 parameter and
gives a thumbs up by means of a listening session and you consider it invalid, while to me you're doing the same thing. Oke, the population is larger, but you might want to explain more about what is solid or scientific proof to you.
You must know there are not a lot of people here that have 20 students to evaluate every single adjustment, if that's what makes it solid.
I used a bass driver as an example to make an easy case. Almost no system has impedance linearized in that frequency range.
I'm not trying to shoot holes in your view of system design, but in general speaker design done here and by the rest. It's another point of view which seems hard to get across and seems to be a major point of disagreement between us, while it really shouldn't be. I assume.
We both use compression drivers so the question was interesting, to me at least.
Please state the use case, or the fault in the calculations because there's not much that proves otherwise.
Link again, to your convenience:
Damping factor
Thanks for your time and effort, either way I'll get out of your hair