I have the following hypothesis. Let's call it such, because not yet objectively proven by my part. I'll try to be brief.
1. The speaker outputs acoustic waves across the listening room.
2. All components to an extent have acoustically resonating mechanical components, examples of which can be.
a) Metal sheet covers.
b) Printed circuit boards.
c) Heatsinks
3. These components can be excited by speaker outputs and modes across the listening room and can output resonating sound of their own.
4. I believe they can also be excited via mechanical connections to the speakers via the cables, hence the reason different cable geometries, weight and conductors can be audible, where they transmit speakers cabinet vibrations to the rest of the components and vice-versa.
5. My claim is that this "audiophile" definitions such as a solid state sound and tube sound are dominantly related to the choice of materials of the build up and not electrical signals.
6. Although (for the moment) subjectively, also with unbiased people, I have tested many different components as dampeners/modifiers attached mechanically to sensitive parts of an audio system, such as speaker walls and driver baskets. Some examples of components I have tested are:
a) Vacuum tubes
b) Capacitors
c) Devices made of different liquids, alloys and metals
7. I have learned to build projects such as speakers and amplifiers to sound very SET like, while they employ solid state devices, just alone by the choice of materials and degree of dampening. I started to use liquids in my projects, such as oils, greases and waxes.
8. In case someone is curious and willing to try out an experiment.
a) Take two equal jars.
b) Fill them with an equal amount of cooking oil
c) Immerse a large bundle of multistranded wire inside, where preferably, the end immersed in the jar oil is shaggy similar to tree branches, in order to contact with the most surface area as possible.
d) Connect the other end of your wire to the speaker basket, then listen / measure. Or you can record A / B via microphone and listen to the recordings.
To measure this, I think one should compare the raw acoustic output from the listening position via microphone and look for difference in frequency response and resonance decay. As a supporting measurement, one could attach accelerometers to various surfaces of components and compare different frequency responses and harmonics.
Unfortunately, my time is being swallowed by building audio transformers and family time, so I'll need, a few years probably, to put all my findings into measurable data. But I do not mean just a few measurements, but a full book of findings and tests. I'm pretty confident that many audible differences, considered as snake oil by the audio community, have an objective explanation by the acoustic properties of the materials used and not electrical signals. I'm also aware that this hypothesis will clash with a huge quantity of already known audiophile dogma, hence a generous amount of free time for explanations, preparations and data will be needed from my part.
I've always been a fan of Delrin for top plates, with a wood base. Makes for a very natural sound, reducing ringing from a steel or aluminum chassis. To me the difference is quite audible. A friend mine who is a very competent DIYer turned me on to Delrin top-plates or sub-plates many years ago, and he builds some of the best-sounding amps in my experience.
What's a cheaper and more available substitute for Delrin? What plastic is closest, and what thickness is recommended?
Wiki says it's Polyoxymethylene (POM), also known as acetal, polyacetal, and polyformaldehyde. Acetal POM seems to be the common name. How different if it from nylon?
Wiki says it's Polyoxymethylene (POM), also known as acetal, polyacetal, and polyformaldehyde. Acetal POM seems to be the common name. How different if it from nylon?
Last edited:
In case you're following my philosophy, keep in mind it gets complicated and there is no simple solution to a problem. Your subjective preferences, speaker choice, room will also play the part. For example, you could choose steel or aluminum top plates that ring, however you could balance things out with oil capacitors and/or add more weight. Or you could choose a ringing surface, but drill many holes into it to break resonances (check drilled cymbals).
For example, I recently created a solid-state class AB system, which sounds very dark and tube like, I kind of overdampened everything, from speaker basket, crossover networks to cables, heatsinks and PCBs. Not everyone likes such sound.
I have never tried acetal, which is similar to Delrin. Delrin has good rigidity and shape retention, and is also extremely easy to machine. A 12" x 12" 1/4" thick sheet will not sag much, if at all, under a load of heavy transformers, and it does not warp over time. Acetal is not much cheaper so I've never bothered to try it. Here's a pic of one of my current Williamson monoblocks:
Not sure that dogma is a correct choice of words - many orders of magnitude of any possible effects below signal level is a serious burden for any hypothesis to overcome. The very first step to winnowing all hypotheses is to eliminate all those with many orders of magnitude special effects. Sometimes it's true (Einstein 1915, 1919) but that's famous because it turned out to be true (exclamation point!). Most don't, because that's just the way knowledge grows, and most hypotheses turn out to be wrong. And in the long enough term, most (ultimately all?) theories turn out to be approximations.
All good fortune,
Chris
I can agree completely. My concern is that we are an "unreliable narrator" to ourselves. We live in a world of our own making, completely constructed inside our own heads, running at about 200mS behind real-world clock time, generating a 3D rectilinear world in color (which, by the way doesn't actually exist - "color" is only inside our heads) from a spherical retina, which is seamlessly synched to our (lower latency) hearing so well that we don't even know it's happening. Just like we don't see our blinks, or our noses, without special effort.
We lie to ourselves more that we lie to other people. To deny this part of us is a tactical mistake. We are untrustworthy enough to make us unreliable observers, without some extra effort. Few folk are willing to make the extra effort (blind testing), but without that extra effort, published judgements are not worth the electrons to type them. Nobody is special enough to transcend our human limitatations.
All good fortune,
Chris
We lie to ourselves more that we lie to other people. To deny this part of us is a tactical mistake. We are untrustworthy enough to make us unreliable observers, without some extra effort. Few folk are willing to make the extra effort (blind testing), but without that extra effort, published judgements are not worth the electrons to type them. Nobody is special enough to transcend our human limitatations.
All good fortune,
Chris
I'm not going to argue against science - reliable and repeatable experiments and evidence based progress.
I also agree that human judgements are unreliable - plenty of psychology shows this. But I also believe blind testing isn't the only way to make progress through listening. For a start it's impractical and on a DIY forum I see few of us using it.
The alternative is a) to be aware of the possibility of bias and factor that into the listening experience - read the psychology and be familiar with it. And respect the nul hypothesis. And b) to be a very educated listener with carefully chosen listening material like voice, piano and acoustic instruments which a number of us are very familiar with in a live form from playing them or being a frequent live music concert-goer. And then becoming super familiar with the test tracks over months and years of testing so there is an awareness of small details. I remember reading some experiments showing that "better listeners" as above got more repeatable results but I can't remember the references.
I also agree that human judgements are unreliable - plenty of psychology shows this. But I also believe blind testing isn't the only way to make progress through listening. For a start it's impractical and on a DIY forum I see few of us using it.
The alternative is a) to be aware of the possibility of bias and factor that into the listening experience - read the psychology and be familiar with it. And respect the nul hypothesis. And b) to be a very educated listener with carefully chosen listening material like voice, piano and acoustic instruments which a number of us are very familiar with in a live form from playing them or being a frequent live music concert-goer. And then becoming super familiar with the test tracks over months and years of testing so there is an awareness of small details. I remember reading some experiments showing that "better listeners" as above got more repeatable results but I can't remember the references.
Also come from a real sound background, although opera, which is pretty extreme. Don't let yourself be positioned two meters downstream of a real soprano - 'nuf said. My concern re: sound reproduction is our human hubris. We believe our our bull$hit. To let ourselves be blinded to our inherent limitations is a classical fallacy, and to position ourselves above or outside of the world of other animals is hubris.
We are all just giant robots built by our DNA in their best attempt to reproduce the genes they carry (Dawkins 1976). Our perceptions are both flawed and easily malliable, and yet we are firmly convinced of our righteousness. We can't help it, or be otherwise.
I'll shut up now,
Chris
We are all just giant robots built by our DNA in their best attempt to reproduce the genes they carry (Dawkins 1976). Our perceptions are both flawed and easily malliable, and yet we are firmly convinced of our righteousness. We can't help it, or be otherwise.
I'll shut up now,
Chris
Hard to argue with most of what you sat, Chris. So you're a singer?
Judging from the appalling state of the world and the failure to take any meaningful action on climate change, it's hard to make a convincing case for humans as part of the animal kingdom. We can nevertheless give humans credit for art and science and sometimes charity, if for little else. I'm also a Dawkins reader from long ago, and a fan of Darwinian evolution. It has at least ensured human survival, though whether we are the "fittest" to be custodians of the planet is a bit of a philosophical issue.
Judging from the appalling state of the world and the failure to take any meaningful action on climate change, it's hard to make a convincing case for humans as part of the animal kingdom. We can nevertheless give humans credit for art and science and sometimes charity, if for little else. I'm also a Dawkins reader from long ago, and a fan of Darwinian evolution. It has at least ensured human survival, though whether we are the "fittest" to be custodians of the planet is a bit of a philosophical issue.
One thing you guys didn't mention when it comes to not trusting your hearing is to have some friends who are also skilled at listening to standard test tracks you all know well. Have them listen on their own and tell you what they hear. If they don't hear what you think you hear then you have an indicator of potential error on somebody's part that needs to be resolved. IOW, DBT is far from the only practical tool we have to check our perceptions. In fact DBT is often impractical because of cost and complexity. OTOH, descriptive analysis is well known classical tool used in perceptual science. The important thing is to have friends who will tell you the truth, and then for you to have the skill to tease out who is wrong or right. For myself, if I'm fooling myself that's bad, so I really want to know if that's the case.
EDIT: to make an analogy, what if you thought you smelled smoke, but weren't sure? Would get someone else to take a sniff and opine on whether they smell smoke too? And what does it smell like exactly, burning insulation, barbeque next door, something else? Two people might help each other find the correct answer more reliably that one person could. However, maybe they are suggesting different smoke smells to each other, so its possible an mutual error could occur if the smoke smell is at a very low level, and the people are suggestable, and if they really want to reach a final conclusion. However, what they should do in that case is not jump to yes/no conclusions. If they don't find a clear answer then just say "we don't know at the moment" and let that be the conclusion (and maybe stay on the alert for awhile just in case).
EDIT: to make an analogy, what if you thought you smelled smoke, but weren't sure? Would get someone else to take a sniff and opine on whether they smell smoke too? And what does it smell like exactly, burning insulation, barbeque next door, something else? Two people might help each other find the correct answer more reliably that one person could. However, maybe they are suggesting different smoke smells to each other, so its possible an mutual error could occur if the smoke smell is at a very low level, and the people are suggestable, and if they really want to reach a final conclusion. However, what they should do in that case is not jump to yes/no conclusions. If they don't find a clear answer then just say "we don't know at the moment" and let that be the conclusion (and maybe stay on the alert for awhile just in case).
Last edited:
Good points, Mark.
The problem I usually find when describing A-B comparisons is finding words small enough to properly describe subtle and only just audible differences. Phrases like "better treble" are poor descriptors unless one says by how much. You end up saying things like "the bass was the tiniest bit more solid" or "the highs had just a fraction of a touch more grain". Language gets to be a problem with tiny differences because language is typically used to describe much more obvious things.
The problem I usually find when describing A-B comparisons is finding words small enough to properly describe subtle and only just audible differences. Phrases like "better treble" are poor descriptors unless one says by how much. You end up saying things like "the bass was the tiniest bit more solid" or "the highs had just a fraction of a touch more grain". Language gets to be a problem with tiny differences because language is typically used to describe much more obvious things.
Its better to compare something specific such as, does the ride cymbal sound from 1:24 to 1:38 seem different in some particular way in A versus B? If someone then says, "the cymbal decay seems more clear and well defined in A." "Also I can hear more of the ping of the stick striking the bell. It sounds more real." You can also ask, are you sure? Do you want to listen again? Then take note of any response to that.
If you can collect comments from a few different people who come in and listen at different times (so they are not influencing each other), and if the comments are pretty well aligned with each other, then there is likely something audibly different between A and B.
And in that case it doesn't have to be HD, nor have to be noise floor causing it, although it could be either or both, and or there could be other causal factors. The first thing is to establish a likelihood of difference exists, then go hunting for causal factors.
EDIT: Also, new people may have to be trained in order to develop skill. Who but a trained musician listens to details of a ride cymbal while there is a melody and other instruments to listen to at the same time? There is a lot going on and short-time details can go by fast. It takes training for the brain to process those events as signals to be noticed.
If you can collect comments from a few different people who come in and listen at different times (so they are not influencing each other), and if the comments are pretty well aligned with each other, then there is likely something audibly different between A and B.
And in that case it doesn't have to be HD, nor have to be noise floor causing it, although it could be either or both, and or there could be other causal factors. The first thing is to establish a likelihood of difference exists, then go hunting for causal factors.
EDIT: Also, new people may have to be trained in order to develop skill. Who but a trained musician listens to details of a ride cymbal while there is a melody and other instruments to listen to at the same time? There is a lot going on and short-time details can go by fast. It takes training for the brain to process those events as signals to be noticed.
Last edited:
The first step is to investigate. For decades 'gurus' were eviscerated by 'realists' over claims that speaker binding posts could degrade fidelity. Recently ASR posted measurements of two variations of the same Buckeye amp. The only reported difference was replacing the ferromagnetic speaker binding posts. The ASR measurements below show a nearly 20 dB improvement in distortion performance from the change at power.
Since this is a human activity the 'realists' of course shifted to debates over audibility. I believe too many years of team public emotional/social investment to do otherwise. The 'gurus' scoring a point undercuts a world view. Yet they still appear willing to ignore every complex assumption, weighting and algorithmic process between the raw instrumentation input and the final output graph. What was the crest factor of the released ferromagnetic energy? What was the timing? The graph implies Fourier behavior and graphs must be science. Hard measurement Purfi's recent report on the origins of listener dissatisfaction with their amps - traced to ferromagnetic storage in the output filters - I believe showed it was closer to delayed pseudo-random noise than Fourier structured harmonics.
Eventually a mechanism analogous to ferromagnetic energy may be discovered in audio's other energy storage device; the capacitor. I stopped worrying and follow measurements where they're available and my ears where they end. Maybe it's a comforting illusion but life's too short and I'm too old to build for team theoretical approval.
Attachments
Hey, hey you are opening a can worms. I have tried many years back to talk about the one and only definitive solution but mostly got bad reactions. My dear friend Lorenzo Russo has solved almost all the problems with his system. The turntable dates back to the mid 80ies. Starting from the turntable, then isolating and following precise criteria through out the entire system and finally the room treatment. Isolation is either (thin membrane) air spring (made by himself) or elastic wire while the object being isolated is as rigid as possible within it's working band. In this case I can truly say this is the one and only REFERENCE stereo system. He is a brilliant engineer who has worked for many years at the Italian patent office....
From left to right: Eldorado (the big one) and RotoMoss turntables, power amplifier, Indian Fathers loudspeakers (in the first picture one can see also one of the elements used for room treatment), Xover, listening position. The room treatment is mostly behind the speakers. On the listening side it's just those few elements around the main listening position but one can really stand anywhere and get a different 3D holographic perspective of the sound stage.
From left to right: Eldorado (the big one) and RotoMoss turntables, power amplifier, Indian Fathers loudspeakers (in the first picture one can see also one of the elements used for room treatment), Xover, listening position. The room treatment is mostly behind the speakers. On the listening side it's just those few elements around the main listening position but one can really stand anywhere and get a different 3D holographic perspective of the sound stage.
Attachments
Double Blindfold Listening Tests are a good thing.
But properly conducted tests like that are extremely hard to carry out.
One of my such tests, at VSAC 2008 had the right idea. Almost everything was done right.
As soon as I set up the test in front of a large audience, I realized the one flaw in the test.
Even with the flaw, it still was entertaining and useful.
It is a Mean thing to say that Karl Friedrich Gauss was just an Average man at the Peak of his career, and at the Center of the Gaussian Curve.
But properly conducted tests like that are extremely hard to carry out.
One of my such tests, at VSAC 2008 had the right idea. Almost everything was done right.
As soon as I set up the test in front of a large audience, I realized the one flaw in the test.
Even with the flaw, it still was entertaining and useful.
It is a Mean thing to say that Karl Friedrich Gauss was just an Average man at the Peak of his career, and at the Center of the Gaussian Curve.