That is why we closed the other thread and pointed to this one. To do something plausible. Controlled listening is a valid scientific experiment too BTW. So not to leave any side uncovered.
Hi Rodeodave,
Things may be difficult to measure some times, but that's the only reliable metric humanity has to be honest. The equipment George is talking about is extremely sensitive, and useful at very low signal levels as well. I'm sure he has more than good enough gear to do baseband frequencies as well.
SY is also a very careful experimenter. He knows how to set up reproducible tests and control other factors that may affect the outcome. So if SY finds anything, I'm pretty sure his results will be reproducible in any decent lab setting. The documentation will be up to anyone's standards.
-Chris
You have that right!
Things may be difficult to measure some times, but that's the only reliable metric humanity has to be honest. The equipment George is talking about is extremely sensitive, and useful at very low signal levels as well. I'm sure he has more than good enough gear to do baseband frequencies as well.
SY is also a very careful experimenter. He knows how to set up reproducible tests and control other factors that may affect the outcome. So if SY finds anything, I'm pretty sure his results will be reproducible in any decent lab setting. The documentation will be up to anyone's standards.
-Chris
I think this is a terribly important point in removing any hidden biases. If it is known ahead of time that the original listening familiarization is either modified or unmodified then listeners will be able to deduce information accordingly from the double blind test as to which system has been modified.
It is imperative that there is no bias as to whether the system familiarization is done on a (one only) unmodified or modified system.
Hi exeric,
That is why testing is so very important. The numbers are what they are, no bias either way - ever. Not even a subconscious bias can exist. I would think that anyone that only has subjective tools at their disposal would welcome any testing that can eliminate bias.
Every particle and assembly of same must follow the rules of physics, no matter how strange the claims may be. IF testing finds something, anything, then you know you aren't crazy and you actually may be hearing a real effect. That's got to put your mind at ease!!
-Chris
That is why testing is so very important. The numbers are what they are, no bias either way - ever. Not even a subconscious bias can exist. I would think that anyone that only has subjective tools at their disposal would welcome any testing that can eliminate bias.
Every particle and assembly of same must follow the rules of physics, no matter how strange the claims may be. IF testing finds something, anything, then you know you aren't crazy and you actually may be hearing a real effect. That's got to put your mind at ease!!
-Chris
I gather from the last post that you don't actually have any experience with double blind sensory testing or know how those are set up? That's OK, there are others who do. I suspect that Panomaniac, at least, will report his experimental results in this thread, and I hope others will as well.
testing, testing 1...2...3...
Part of experimental design involves criteria decisions to determine what exactly represents acceptance/rejection of testing results (hopefully some form of multifactorial). I'm sure Sy fully comprehends this and will appropriately descibe the conditions and results comprehensively.
I'm not so sure many other proponents and thread participants here fully understand the significance of this design consideration.
John L.
Part of experimental design involves criteria decisions to determine what exactly represents acceptance/rejection of testing results (hopefully some form of multifactorial). I'm sure Sy fully comprehends this and will appropriately descibe the conditions and results comprehensively.
I'm not so sure many other proponents and thread participants here fully understand the significance of this design consideration.
John L.
What you described bears no relation to a double blind test and you've made it clear that you won't be contributing to that aspect of testing. But other people are willing to step in and do the work, so don't worry about it.
Well, you have hit upon it exactly Chris. When the only tool one has is a hammer then everything becomes a nail. It used to be thought that total harmonic distortion was all that mattered. This was especially true when transistors came about. Maximum feedback occured to bring the distortion down but no notice was taken of how feedback just creates higher frequency distortion products that actually sound worse. (Maybe not so much with early transistor designs. They had to do it just to make it tolerable.)
It took a long time to realize the subtleties of the effects of massive amounts of feedback. We didn't know how the ear accepts different levels of distortion from different harmonics of the fundamental. Many of the higher order of odd harmonics came directly from the feedback process itself. Why do you assume that the situation is any different now? People eventually learned to respect that what they heard in high feedback designs was real and was causing them to sound bad. I think the individuals who actually listen to the QPs will eventually cause everyone to decide that there is something fundamental that we aren't measuring but that the ear can hear. And even if we can't discover it for a long time if enough people actually listen to the bybees and like them it will be just like before. We will realize there are things that we now don't know how to measure that effects sound.
Ok, I will help you with constructing a double blind listening test, if you want me to. But first you have to help me understand where I've gone wrong in my understanding. Also I'd like to understand how the method I described would cause the results to be ambiguous, or worse, skewed. I think that is owed me certainly because so far I'm the only one representing the opposite side of the Bybee question.
There are no opposite sides in this thread for real. Its a test thread. Is there any opposition in test standards till now? Its two things to be put to the test here. Noise and controlled listening. It claims noise cleaning and tell tale aural perception of the effect. $1 Million DC to RF lab availability and a pro operator has been offered to recheck SY's findings after he finishes, and a high sensitivity speaker guy has offered to connect the devices on his speakers and recheck in controlled listening for effects. Its either it cuts it or not. Simple.
Use a 0.025 Ohm resistor as comparison in the Loudspeaker test. Very small differences in the volume of the tweeter can have drastic effects on sound. Read Linkwitz Lab - Loudspeaker Design about what Siegfried found out recently.
Hi exeric,
All I can ask is when was the last time you really looked at the engineering world? For that matter, how about the test equipment industry? Audio design? I have involvement in all these areas, and what you have described as the audio engineering landscape existed in the early 70's at the latest. Our collective knowledge has increased by leaps and bounds, the information successfully disseminated throughout a high percentage of practicing audio engineers. Actually, I hate the term "audio engineer". This is really an analog signals engineer that happens to focus on audio signals. That is the reality. Sure, you will find people stuck in the past with old ideas and understanding, but that cuts both ways.
You want to see how much is understood these days? Just have a look at any high grade oscillator or distribution amp. We are looking at jitter in the fs range - and reducing it. We know how critical power supply noise is and how much it can pollute a clean clock. Any circuit noise creates a degraded clock signal. Temperature changes create instability and noise. Mechanical stress has long been shown to increase the tendency for noise. All these things coming from frequency standards and reference frequency distribution to test instruments - never mind the clock on the wall. We measure frequency down to 10 exp(-13) and lower. Never mind phase in the audio band.
Not everything is quantified for it's audible effects, but at least we know enough to listen to the product and how to avoid things we know causes distortion, or poor sound quality to you. Signals are being measured to tiny levels I doubt you can imagine, but the up-shoot of all this is that the old view of an engineer who only watches a THD meter has been dead long before tubes went out. They had IM meters built with tubes for heaven's sake!
Now for the clincher. Even I have far more tools than a hammer. I do have to admit to owning more than one type of hammer, and that they can be really useful sometimes. Currently, there are 13 different instruments on my bench, not including hand held instruments (more than just meters), 9 or 10 other instruments, a computer with a good sound card setup. Also, an audio test system consisting of a separate amp, preamp and other items I can swap out including tube sound equipment.
I'm pretty sure that there is an appropriate tool in there somewhere for the job at hand. Then you have real engineers and labs that are far better equipped than I am.
The old, tired refrain of we can't measure what we can hear should be retired, as should generalities applied to all signal engineers. They aren't all equal, but the process and newer equipment darn well is. Even the devices we use have improved dramatically over what you have been talking about. It really is a brave, new world where people really do know what they are doing. They also know enough to listen.
-Chris 🙂
All I can ask is when was the last time you really looked at the engineering world? For that matter, how about the test equipment industry? Audio design? I have involvement in all these areas, and what you have described as the audio engineering landscape existed in the early 70's at the latest. Our collective knowledge has increased by leaps and bounds, the information successfully disseminated throughout a high percentage of practicing audio engineers. Actually, I hate the term "audio engineer". This is really an analog signals engineer that happens to focus on audio signals. That is the reality. Sure, you will find people stuck in the past with old ideas and understanding, but that cuts both ways.
You want to see how much is understood these days? Just have a look at any high grade oscillator or distribution amp. We are looking at jitter in the fs range - and reducing it. We know how critical power supply noise is and how much it can pollute a clean clock. Any circuit noise creates a degraded clock signal. Temperature changes create instability and noise. Mechanical stress has long been shown to increase the tendency for noise. All these things coming from frequency standards and reference frequency distribution to test instruments - never mind the clock on the wall. We measure frequency down to 10 exp(-13) and lower. Never mind phase in the audio band.
Not everything is quantified for it's audible effects, but at least we know enough to listen to the product and how to avoid things we know causes distortion, or poor sound quality to you. Signals are being measured to tiny levels I doubt you can imagine, but the up-shoot of all this is that the old view of an engineer who only watches a THD meter has been dead long before tubes went out. They had IM meters built with tubes for heaven's sake!
Now for the clincher. Even I have far more tools than a hammer. I do have to admit to owning more than one type of hammer, and that they can be really useful sometimes. Currently, there are 13 different instruments on my bench, not including hand held instruments (more than just meters), 9 or 10 other instruments, a computer with a good sound card setup. Also, an audio test system consisting of a separate amp, preamp and other items I can swap out including tube sound equipment.
I'm pretty sure that there is an appropriate tool in there somewhere for the job at hand. Then you have real engineers and labs that are far better equipped than I am.
The old, tired refrain of we can't measure what we can hear should be retired, as should generalities applied to all signal engineers. They aren't all equal, but the process and newer equipment darn well is. Even the devices we use have improved dramatically over what you have been talking about. It really is a brave, new world where people really do know what they are doing. They also know enough to listen.
-Chris 🙂
Already have one ordered from Mouser. (Tip o' the hat to the kind individual who disassembled a Bybee and told me what he found)
Parasitic resistance will have to be measured exactly and substituted in any loudspeaker test. Very good point Joachim.
I certainly wasn't trying impose an old paradigm but just that the THD was just one example, but was the clearest example I can think of. Of course I know that it is an old example. But it illlustrates the limitations of human frailties in which perceptions are guided by what we already know. We actually ignore what we hear when we can't explain it. It has happened time and time again in audio. When something new comes about that doesn't fit into that paradigm instead of trying to figure out why it is occuring we instead deny its very existance. That is what I was trying to illustrate.
Sure, I understand very well that things have improved greatly from the 60s and 70s. I have 20 years working in avionics that ended in 1999 so I'm not a neophyte either. But human frailties and the desire to explain away real mysteries is still with us. I really do hope that we can get to the bottom of this mystery but I'm also afraid. I'm afraid because I know that a vacuum of scientific knowledge about a phenomenon usually results in denying the phenomena. At least for a long time. I don't want that fear to taint the listening experiment in any way.
I'm not worried about the measurement experiment too much because I'm expecting it to largely be inconclusive. If I'm wrong about that I will be very happy.
I don't think that if the guys can not measure differences but can hear differences will have a problem reporting it.
Does that 3468A have the ability for kelvin resistance measurements SY? Still, 0.025 ohms is a challenge to measure accurately, maybe a comparison bridge might help there. Then there is the question of temperature rise and how the resistance reacts.
To me, 0.025 ohms might probably be inherent in any construction. That might make it a red herring, and I don't like fish.
-Chris
To me, 0.025 ohms might probably be inherent in any construction. That might make it a red herring, and I don't like fish.
-Chris
- Status
- Not open for further replies.