Ideally, the point of the thread is to provide a focal point for contributions relating to the topic. This time around, some real pay dirt. Lynn Olsen's casual aside about the microphonics, which lent credence to some manufacturers who feel that high winding tension helps. The Bates article, which I refused to buy on principle because of the (seemingly) exorbitant cost of back issues, now available for a reasonable sum. (The discussion of foil vs. metallized was particularly interesting to me. ) The Conrad Hoffman test results; a very nice addition.
So, all in all, a very worthwhile thread.
Not understanding the myriad of ways that non-auditory cuing can happen is OK. But you do need to understand that, like it or not, understand it or not, they happen. In order for a listening test to be valid, they have to be controlled for, otherwise what you have is a nice little story for the internet and nothing more.
I've participated in some informal testing with Wolf's box. The difference between NPE and Poly was instantly and easily obvious, without prior knowledge of which was in the circuit. The NPE and the Poly could easily be identified and my preference was clearly the poly. Not a happy situation when one routinely puts 100-200uf caps in woofer circuits.
I think I heard differences between poly's which were much more subtle. More testing would have to be done to determine if I could reliably identify a change, much less a particular cap.
Couple of thoughts.
Ultimately, for an individual DIY'er, it is matter of preference and budget.
Switchboxes are may be useful in an A/B/X format for establishing whether differences can on average be heard. Not so much for individual preferences which require listening longer and to more music.
I'm an individual differences psychologist, and know that an average is boring description of central tendency. The juice is in the variance.
Listening is a learned skill. Individual differences in hearing acuity, attentional focus, and experience will influence results. In a group A/B test, there will be much more variance among listeners than among caps. What individual differences in cap and listener lead to individual preferences? Nobody really knows I don't think.
A protocol that would firmly establish the individual differences bases of preferences among similar spec'd caps would probably be well beyond any threshold of economic value to manufacturers. The masses don't much care.
I'll make no assertions about "better"; and for me, my personal preferences need no inferences as to cause.
I think I heard differences between poly's which were much more subtle. More testing would have to be done to determine if I could reliably identify a change, much less a particular cap.
Couple of thoughts.
Ultimately, for an individual DIY'er, it is matter of preference and budget.
Switchboxes are may be useful in an A/B/X format for establishing whether differences can on average be heard. Not so much for individual preferences which require listening longer and to more music.
I'm an individual differences psychologist, and know that an average is boring description of central tendency. The juice is in the variance.
Listening is a learned skill. Individual differences in hearing acuity, attentional focus, and experience will influence results. In a group A/B test, there will be much more variance among listeners than among caps. What individual differences in cap and listener lead to individual preferences? Nobody really knows I don't think.
A protocol that would firmly establish the individual differences bases of preferences among similar spec'd caps would probably be well beyond any threshold of economic value to manufacturers. The masses don't much care.
I'll make no assertions about "better"; and for me, my personal preferences need no inferences as to cause.
Your findings seem very similar to those published in the CC AES paper I referenced in an earlier post. I fully agree that when comparing only film types, a skilled listening panel is necessary to descriminate between samples and secondly, variance among panelists will be greater than among film caps.
Wolf, you could have been 100 miles away and remotely hitting a switch and people here would have still found flaws with your method.
Here's one of my recent experiences, but 1st, disclaimers. 1. I'm no Golden-Eared Audiofile. More often, I feel like a tin-eared audiofool. However, I discovered years ago that I can indeed hear some differences, for instance among interconnect cables. 2. I believe that EVERY change to a music-reproduction system causes the sound to change, no matter how infintessimally. 3. I believe that the engineers/scientists who claim that if they can't measure a difference, that difference doesn't exist are highly arrogant and not nearly smart enough to admit that they don't know everything.
My experience is with removing a 55uF bank of film caps from the frontend of my Audio Physic Avanti midrange crossovers. These 55s were hi-pass filters thru which ALL of the midrange passed. I jumped them and 'replaced' them with a different hi-pass-filter cap in the poweramps. I heard less grittiness/edgyness and less softness of information, that is, more detail. Years ago I did the same thing with a pair of Eminent Technology ET-8s with the same result--better-sounding music.
No one needs a switchbox or double-blind testing to convince me that capacitors are audible.
My experience is with removing a 55uF bank of film caps from the frontend of my Audio Physic Avanti midrange crossovers. These 55s were hi-pass filters thru which ALL of the midrange passed. I jumped them and 'replaced' them with a different hi-pass-filter cap in the poweramps. I heard less grittiness/edgyness and less softness of information, that is, more detail. Years ago I did the same thing with a pair of Eminent Technology ET-8s with the same result--better-sounding music.
No one needs a switchbox or double-blind testing to convince me that capacitors are audible.
No problems with proceeding, but now you know some things to be careful about. Any test like this you present to the forum will have a very large audience and therefore come under a lot of scrutiny.
The better you can make the test, the more confidant you can be of the results. You'll never have a perfect test, but at least you now have a better idea of what traps to beware of.
The better you can make the test, the more confidant you can be of the results. You'll never have a perfect test, but at least you now have a better idea of what traps to beware of.
They may not know everything but they know the importance of controlling any test that has a high probability of very subjective (uncontrolled) conclusions. Your conclusions below could be a PERFECT example.
Case in point, your conclusions could be very inaccurate because you did not control your listening test. You have to understand that your brain takes a whole bunch of variables along with the signal from your ears and makes conclusions. Your brain does not use the signal from your ears alone. You have to remove all other variables so that the signal from your ears is pretty well all that is left. Then and only then will your conclusions be accurate.
More detail differences can be shown in measurements, do you know how to do measurements at all? did you try and find the differences in the measurements?
If you have never spent time double checking any conclusions you have from your senses you will never understand what Im posting.
You definitely do not need anything to make conclusions day to day but when posting and making bold claims on a forum that has a little science in it maybe you should atleast present conclusions with a little bit more science behind it.
Last edited:
Yep, all the controlled tests I have seen and done have the same conclusions.
Makes you wonder about those "Other" variables in uncontrolled tests
I think we should be ignoring "Bad quality" caps in any test.
Well, this has been proved over and over in the science community so until you have something that proves people in general make accurate conclusions without controlling the hidden variables we will stick with the science.
Im amazed you think its a weakness in the science community. Ignorance is bliss. From BOSE owners to the 1 billion $$ bogus weight loss market and 1 billion $$ bogus natural suppliment market...yeah, people definitely have self control over their bias
Im just happy no one like you works in the drug industry. "Don't worry Joe about doing a controlled test, People are very self aware and they would never give us the wrong conclusions. Release the drug next week, its time to make money!!!" Yikes!!
Last edited:
Don't believe in false Gods
Beware of what you say. Inquisition 's here,and it 's here to stay.
B.L
Hello Wolf teeth
Have you tried Charge Coupling with capacitors?? JBL has been doing it for about 20 years on their TOTL line speaker systems. What you do is run two capacitors in series with a 2-3 meg resistor tied to the node between the capacitors. The other side of the resistor goes to a 9 volt battery with the battery negative to ground. You effectively DC bias the capacitors.
Something to try if you have not. I use it combined with aguaplassed damped compression drivers and the combination of adding a dampening compound on the diaphrams and biasing the capacitors is a real winner IMHO.
Bypassed and Biased Capacitors
Rob
Have you tried Charge Coupling with capacitors?? JBL has been doing it for about 20 years on their TOTL line speaker systems. What you do is run two capacitors in series with a 2-3 meg resistor tied to the node between the capacitors. The other side of the resistor goes to a 9 volt battery with the battery negative to ground. You effectively DC bias the capacitors.
Something to try if you have not. I use it combined with aguaplassed damped compression drivers and the combination of adding a dampening compound on the diaphrams and biasing the capacitors is a real winner IMHO.
Bypassed and Biased Capacitors
Rob
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.