planet10 said:
To be an empirical a model it must represent a correlation between the variation of some observed result and the variation of some other parameter of set of parameters. If there is anything empirical in the application of the pattern it might be that bigger drivers require a bigger pattern. His experience may have shown that the pattern must be scaled up or down to work on larger/smaller drivers. If that is the case then he has made a correlation that the pattern size scales with driver size. The hypothesis of how the pattern works in not a model. It does not tell us anything about how the audible effects vary as the pattern varies. It is just an (incorrect) attempt to describe a physical process.
dlr said:
Bud can take a new driver, unseen before, and use his model to accuratly predict how to treat it to getthe results he wants. Today there are over 30 years of previous drivers. That certainly seems to safisfy your requirements.
But i'm sure you'll find a way to twist that too since it likely does not fit your model of the world.
dave
Re: Re: empirical vs. anecdote
So let's see.. what may affect results
variable #1 unkown driver parameters pre/post/during treatment
variable #2 other "treatments" applied besides EnABL
variable #3 phase plug installed / not installed
variable #4 only 1 driver for testing, no assesment of sigma
variable #5 who knows?
any or all of these variables may interact to cause confounding of the expected results; since there are no quantifiable criteria defined to decide if the results are significant (other than "it sure does sound better") your statements are self contradicting regarding any meaningful result. Cause and effect cannot be established under these conditions. That's a scientific reality.
I know this is "only" a hobby, no ones' life depends on it, corporate profits aren't at stake, etc., but you make absolute statements implying the results are valid, which they aren't for anyone but the individual making such statements (you).
John L.
planet10 said:
So let's see.. what may affect results
variable #1 unkown driver parameters pre/post/during treatment
variable #2 other "treatments" applied besides EnABL
variable #3 phase plug installed / not installed
variable #4 only 1 driver for testing, no assesment of sigma
variable #5 who knows?
any or all of these variables may interact to cause confounding of the expected results; since there are no quantifiable criteria defined to decide if the results are significant (other than "it sure does sound better") your statements are self contradicting regarding any meaningful result. Cause and effect cannot be established under these conditions. That's a scientific reality.
I know this is "only" a hobby, no ones' life depends on it, corporate profits aren't at stake, etc., but you make absolute statements implying the results are valid, which they aren't for anyone but the individual making such statements (you).
John L.
Hi John K,
Thanks for those details, looks a very useful tool.
Can this simulator show a sine starting from zero and runing for say two cycles at that 3.5kHz peak frequency and then off again. Yes I know that sines do not suddenly start and stop in music, but waveform modification might be illustrated at this frequency.
Can I also suggest that the dust-cap be reversed so that it becomes concave with the same spherical cap shape. If this reduces the peak - then does the simulator explain why, or would this be something we must hypothesise about ?
Cheers ..... Graham.
Thanks for those details, looks a very useful tool.
Can this simulator show a sine starting from zero and runing for say two cycles at that 3.5kHz peak frequency and then off again. Yes I know that sines do not suddenly start and stop in music, but waveform modification might be illustrated at this frequency.
Can I also suggest that the dust-cap be reversed so that it becomes concave with the same spherical cap shape. If this reduces the peak - then does the simulator explain why, or would this be something we must hypothesise about ?
Cheers ..... Graham.
Re: Neighbour's subwoofer gets EnABL'd
What model is that? If you mean he can scale his pattern, then, as I said, he has made an empirical correlation between his observations and the scaling of the pattern. The discussion about how the pattern is suppose to work says nothing about the size or placement of the pattern. And, again, like a broken record, all data clearly shown that Bud discription of how enable works is incorrect.
What the heck, based on not ever listening to any driver or every having seen this discussion, given Bud pdf file of the template anyone could apply that pattern to any driver if they just use common sense. There is no model. All there is is, do this. It works.
Say what you want, its not worth arguing about what constitutes a model. I don't need to claify what I said. I think those who understand get the point. If they don't nothing I can say will alter that.
Becareful Alex. If you report that it works all will be good. If you report a negative result your idea will be treated with comtempt as you have not followed the directions for the correct application.
I wonder if nail polish with metalic glitter in it would work better than plain nail polish. Those metalic flecks should increase the added mass. Plus, the reflections of the transvers waves off the glitter will be at random angles and will be more effective (I'm kigdding guys, but I won't be surprised if someone goes off and tries glitter nail polish and claims even better results.)
planet10 said:
What model is that? If you mean he can scale his pattern, then, as I said, he has made an empirical correlation between his observations and the scaling of the pattern. The discussion about how the pattern is suppose to work says nothing about the size or placement of the pattern. And, again, like a broken record, all data clearly shown that Bud discription of how enable works is incorrect.
What the heck, based on not ever listening to any driver or every having seen this discussion, given Bud pdf file of the template anyone could apply that pattern to any driver if they just use common sense. There is no model. All there is is, do this. It works.
Say what you want, its not worth arguing about what constitutes a model. I don't need to claify what I said. I think those who understand get the point. If they don't nothing I can say will alter that.
Alex from Oz said:
Becareful Alex. If you report that it works all will be good. If you report a negative result your idea will be treated with comtempt as you have not followed the directions for the correct application.
I wonder if nail polish with metalic glitter in it would work better than plain nail polish. Those metalic flecks should increase the added mass. Plus, the reflections of the transvers waves off the glitter will be at random angles and will be more effective (I'm kigdding guys, but I won't be surprised if someone goes off and tries glitter nail polish and claims even better results.)
Graham Maynard said:
Let's separate the model from the simulation tool developed around it. Such a model could be set up to do what ever you like. The limiting factor is typical the computational effort (computer memory, CPU time) required to generate a solution. This particular implementation is set up for predicting the frequency response and for application on a PC.
Re: Re: Re: empirical vs. anecdote
All of those are a given, and known going in. This particular driver set is an opportunity to explore things and get a handle on what to test for (a necessary step in most scientific endeavor). Ron does have his own stock FE166. He could even install the plugs in them. They are behind the shadow of the whizzer cone.
dave
auplater said:
All of those are a given, and known going in. This particular driver set is an opportunity to explore things and get a handle on what to test for (a necessary step in most scientific endeavor). Ron does have his own stock FE166. He could even install the plugs in them. They are behind the shadow of the whizzer cone.
dave
Re: Re: Re: Re: empirical vs. anecdote
yup.. you're right... all are given going in... so where are they for this particular driver? There isn't any need to "explore things and get a handle..blah,blah" There are specific screening designs wherein variables are purposely manipulated between values to ascertain if such changes actually affect the results. That hasn't been done here. You've never used ANOVA or other experimental designs, I take it. You ought to at least try to understand such things before making statements about test results... from WIKI...
"Group A is given vodka, Group B is given gin, and Group C is given a placebo. All groups are then tested with a memory task. A one-way ANOVA can be used to assess the effect of the various treatments (that is, the vodka, gin, and placebo).
Group A is given vodka and tested on a memory task. The same group is allowed a rest period of five days and then the experiment is repeated with gin. The procedure is repeated using a placebo. A one-way ANOVA with repeated measures can be used to assess the effect of the vodka versus the impact of the placebo.
In an experiment testing the effects of expectations, subjects are randomly assigned to four groups:
expect vodka-receive vodka
expect vodka-receive placebo
expect placebo-receive vodka
expect placebo-receive placebo (the last group is used as the control group)
Each group is then tested on a memory task. The advantage of this design is that multiple variables can be tested at the same time instead of running two different experiments. Also, the experiment can determine whether one variable affects the other variable (known as interaction effects). A factorial ANOVA (2×2) can be used to assess the effect of expecting vodka or the placebo and the actual reception of either."
http://en.wikipedia.org/wiki/Analysis_of_variance
http://en.wikipedia.org/wiki/MANOVA
http://en.wikipedia.org/wiki/Independent_variables
They are very powerful techniques when used correctly, and can provide alot of insight into what is happening vs. what is perceived as happening.
John L.
planet10 said:
yup.. you're right... all are given going in... so where are they for this particular driver? There isn't any need to "explore things and get a handle..blah,blah" There are specific screening designs wherein variables are purposely manipulated between values to ascertain if such changes actually affect the results. That hasn't been done here. You've never used ANOVA or other experimental designs, I take it. You ought to at least try to understand such things before making statements about test results... from WIKI...
"Group A is given vodka, Group B is given gin, and Group C is given a placebo. All groups are then tested with a memory task. A one-way ANOVA can be used to assess the effect of the various treatments (that is, the vodka, gin, and placebo).
Group A is given vodka and tested on a memory task. The same group is allowed a rest period of five days and then the experiment is repeated with gin. The procedure is repeated using a placebo. A one-way ANOVA with repeated measures can be used to assess the effect of the vodka versus the impact of the placebo.
In an experiment testing the effects of expectations, subjects are randomly assigned to four groups:
expect vodka-receive vodka
expect vodka-receive placebo
expect placebo-receive vodka
expect placebo-receive placebo (the last group is used as the control group)
Each group is then tested on a memory task. The advantage of this design is that multiple variables can be tested at the same time instead of running two different experiments. Also, the experiment can determine whether one variable affects the other variable (known as interaction effects). A factorial ANOVA (2×2) can be used to assess the effect of expecting vodka or the placebo and the actual reception of either."
http://en.wikipedia.org/wiki/Analysis_of_variance
http://en.wikipedia.org/wiki/MANOVA
http://en.wikipedia.org/wiki/Independent_variables
They are very powerful techniques when used correctly, and can provide alot of insight into what is happening vs. what is perceived as happening.
John L.
I will be going to my SC home in around 2 weeks for a break, then i can pick up a stock driver for A/B testing. Yall realize i am shutting down the SC system while i am doing the test back in Texas and my SO is going to make my life miserable as she enjoys her music. (What we give up for science!)
I eliminated one axis movement and just angled the driver in its fixture to arrive at a near equal T/R distance across the cone. Had a tech run a test scan at 500 Hz. The 500 hz signal was filtered out and the result was there was still some low level energies at varing distance from the center to the outside of the cone. He rotated the cone 90 degrees and the position of the energies across the surface changed. If there is an effect it is a variable and SWAG suggests that it is due to the varing height/mass of the dots.
Until i can do a true A/B at a varity of frequencies i cannot come up with conclusive data.
ron
added note: Just looking at the results of the first trial it would appear that if its a mass effect ( i doubt it) then greater consideration should be applied to the equal mass of the individual dots. If its a height effect on the BL then the dots height will need to be more precise. Placement, i have no idea.
I eliminated one axis movement and just angled the driver in its fixture to arrive at a near equal T/R distance across the cone. Had a tech run a test scan at 500 Hz. The 500 hz signal was filtered out and the result was there was still some low level energies at varing distance from the center to the outside of the cone. He rotated the cone 90 degrees and the position of the energies across the surface changed. If there is an effect it is a variable and SWAG suggests that it is due to the varing height/mass of the dots.
Until i can do a true A/B at a varity of frequencies i cannot come up with conclusive data.
ron
added note: Just looking at the results of the first trial it would appear that if its a mass effect ( i doubt it) then greater consideration should be applied to the equal mass of the individual dots. If its a height effect on the BL then the dots height will need to be more precise. Placement, i have no idea.
Thats just another way of stating" i dont know".
Not clear about your reply. Emergent phenomena occur in open systems with some measure of chaotic behaviour. A local reduction in entropy.
I hope that helps.
Or, if you're stating the obvious (i.e. I don't know - that's why I'm asking with all the apologies) I must assume you are trying to discourage either speculative posts or my posting. If that's the case I'll certainly oblige.
No problem my friend. I have found in just about all cases that an answer is simple. If you start including chaos theory then the answer becomes more complicated. I have spent my life giving answers to the CEO or a board of directors who cannot understand the complexities of a system and could care less. I recently had a debate with a Sigma 6, black belt or whatever who wanted to run a "fish bone" paper on our process. I simply stated " we know the problems, we just have to correct them".
Analysis has its place, but the end answer is does the product work and does it meet the specifications and can it be delivered on time.
ron
(the hardest thing you can do is to make something simple)
Not clear about your reply. Emergent phenomena occur in open systems with some measure of chaotic behaviour. A local reduction in entropy.
I hope that helps.
Or, if you're stating the obvious (i.e. I don't know - that's why I'm asking with all the apologies) I must assume you are trying to discourage either speculative posts or my posting. If that's the case I'll certainly oblige.
No problem my friend. I have found in just about all cases that an answer is simple. If you start including chaos theory then the answer becomes more complicated. I have spent my life giving answers to the CEO or a board of directors who cannot understand the complexities of a system and could care less. I recently had a debate with a Sigma 6, black belt or whatever who wanted to run a "fish bone" paper on our process. I simply stated " we know the problems, we just have to correct them".
Analysis has its place, but the end answer is does the product work and does it meet the specifications and can it be delivered on time.
ron
(the hardest thing you can do is to make something simple)
Well said. But it wouldn't change this thread until both camps adopt your approach. Rather than smugly shoot down each and every attempt at a description or theory with absolute statements (which in reality are nothing more than theories that have passed the statistical tests and hence ALWAYS contain a potential of error), if the "skeptics" would instead use phrases like:
"Interesting idea, Bud, but my experience and studies just don't support that because x, y, and z... Could it be that what you're experiencing is a, b, or c..."
In truth, though, there are only a few individuals who engage in the more confrontational dialog, though they would seem to be driving this increasingly tiresome thread.
Carl
I hope this tool gets more powerfull than FineCone before it's released. Looking at the data right now gives me a feeling that it's similar but might take more time to compute due to the 3D nature. With the size of the patterns and possibly the gid size of the model necessary, I would say it might take quite a few hours to compute with grid sizes 1mm or below. But the most important thing is it needs to be able to import 3D models from existing CAD programs.john k... said:
John,
Can you explain for me what you presented in post #2851? You've used measured data from drivers to develop a model (predictive) which you've then used to simulate a driver's response, and hope to use that to simulate EnABL? And the graph shows the simulation of an actual driver (specs fed into the model) vs actual measurements? And what kind of predictive accuracies have you found when using the model and simulation on drivers that weren't used in the model development? (R2, F-value and it's significance level, etc, etc)
Thanks,
Carl
Originally posted by john k... - Post #2868
Becareful Alex. If you report that it works all will be good. If you report a negative result your idea will be treated with comtempt as you have not followed the directions for the correct application.
I wonder if nail polish with metalic glitter in it would work better than plain nail polish. Those metalic flecks should increase the added mass. Plus, the reflections of the transvers waves off the glitter will be at random angles and will be more effective (I'm kigdding guys, but I won't be surprised if someone goes off and tries glitter nail polish and claims even better results.)
LOL
I'm in uncharted waters when it comes to EnABL on ports and baffles!
There is no ideal or correct application for ports or baffles as yet.
It’s more of an intuitive process using the EnABL pattern and block sizes as a starting point for exploration.
Things I have observed so far about EnABL on ports and baffles:
1. Added mass does not apply (obviously)
2. Materials other than paint produce audible results
3. Block scaling seems to be more flexible (compared to drivers)
4. Alteration of the basic EnABL pattern is not advisable
5. Room interaction with bass frequencies changes significantly when you EnABL ports
6. EnABL on baffle edges appears to have some impact on diffraction.
I suspect people would be more willing to try sticking some magic tape to their port or baffle rather than paint their drivers.
I’m hoping others will play with this also and contribute their own experiences.
Cheers,
Alex
B&W has dimples on ports of some of their models.Alex from Oz said:
LOL
I'm in uncharted waters when it comes to EnABL on ports and baffles!
There is no ideal or correct application for ports or baffles as yet.
It’s more of an intuitive process using the EnABL pattern and block sizes as a starting point for exploration.
Things I have observed so far about EnABL on ports and baffles:
1. Added mass does not apply (obviously)
2. Materials other than paint produce audible results
3. Block scaling seems to be more flexible (compared to drivers)
4. Alteration of the basic EnABL pattern is not advisable
5. Room interaction with bass frequencies changes significantly when you EnABL ports
6. EnABL on baffle edges appears to have some impact on diffraction.
I suspect people would be more willing to try sticking some magic tape to their port or baffle rather than paint their drivers.
I’m hoping others will play with this also and contribute their own experiences.
Cheers,
Alex
Again, perusing this thread has lead me to comment. The way I see it, science is a tool. A darn handy tool at that, but no more. It's main power, as many point out, is its universality to those who speak the language and its consequent acceptance.
One should not be seduced by semantics re. the word "model". The different proponents of the use of this word are in fact referring to different things. The word does not belong to any of the different factions that use it. And one bunch appropriating the word of the other bunch leads to needless confusion.
Scientific models are valid within some very clearly specified boundaries. Empirical models can be just as valid. Take, for instance, the discipline of acupuncture. Years of study has (have?) gone into this and it has been refined to an extent not to be discounted. Its probably more accepted than not. Western science evangelists might as well get used to this. "China is sleeping. When China awakes, the world will tremble."
I think our problem here arises from the fact that the different "camps" I am referring to here (and it's not that clear cut, obviously) are trying to add the power of the other (camp) to their capabilities. Bud wants the power of formalised science (its generality and predictive powers) to bear upon his inventions and the scientists want their methodologies to assimilate or disrupt the power of the numbers of people claiming positive results of the treatment. Since both stand to potentially gain from this interaction, if performed with an open mind, one might as well cease the semantic babble and get on with it. The former is much easier but worth much less in the long run. I, for one, am not much interested in muscle flexing.
Phew! Glad that's off my chest.
Now, could anyone say "phase" please?
One should not be seduced by semantics re. the word "model". The different proponents of the use of this word are in fact referring to different things. The word does not belong to any of the different factions that use it. And one bunch appropriating the word of the other bunch leads to needless confusion.
Scientific models are valid within some very clearly specified boundaries. Empirical models can be just as valid. Take, for instance, the discipline of acupuncture. Years of study has (have?) gone into this and it has been refined to an extent not to be discounted. Its probably more accepted than not. Western science evangelists might as well get used to this. "China is sleeping. When China awakes, the world will tremble."
I think our problem here arises from the fact that the different "camps" I am referring to here (and it's not that clear cut, obviously) are trying to add the power of the other (camp) to their capabilities. Bud wants the power of formalised science (its generality and predictive powers) to bear upon his inventions and the scientists want their methodologies to assimilate or disrupt the power of the numbers of people claiming positive results of the treatment. Since both stand to potentially gain from this interaction, if performed with an open mind, one might as well cease the semantic babble and get on with it. The former is much easier but worth much less in the long run. I, for one, am not much interested in muscle flexing.
Phew! Glad that's off my chest.
Now, could anyone say "phase" please?
- Status
- Not open for further replies.