Reading research literature is one thing, while actually and practically validating a certain test may be something different.
From your evasive replies I conclude that you haven't actually and practically validated the tests you are conducting.
God is in the details.
As I wrote above, the validity of any test may fall upon not following exactly and accurately all the major and minute details of the procedure. For instance, using a switching box may invalidate such a test, though all other details count.
I can't imagine any practical test procedure for a preamp, where I'd have my own mitts on the volume control. Maybe compare a second one in series to a straight wire bypass?
The real difficulty for me would be in choosing a question worth asking. What do I really disbelieve enough about my own experience? All the questions worth the effort are by definition difficult to observe.
Thanks, as always,
Chris
From my "evasive reply," you should conclude that I believe your question to be frivolous, first because you've defined "validated" (to the extent you defined it at all) as "getting the results Joshua expects." If stuff is audible, I care about it, if stuff isn't, I don't. I personally use DBTs where they make sense. So do many of my friends whom I think have good ears but also share with me a curiosity about what's the right answer, moreso than any sort of personal or professional validation.
If it makes you feel better to refuse to look through the telescope (to use John's favorite chestnut), all power to you.
Bingo. If we took a Blowtorch and switched the outputs between "direct" and "through a simple, low distortion gain stage" padded to unity gain and couldn't hear a difference, the simple, low distortion gain stage is at least as "transparent" sonically as the Blowtorch. Then you can have the tactile and visual fun of using the Blowtorch's luxurious controls.
The validity of the test Joshua, is that if you eliminate as far as possible all of the variables that we know cause reactions in people, and try to isolate what they actually hear from all of the things they think they are hearing, then people simply cannot hear the difference between standard c.d. sampling rates and 192kHz. twenty odd bit ones, good valve amps and equally good solids state ones etc.
The difficulty with being a scientist is that you must accept the hypothesis that best fits the data, and quite often that is one that is completely contrary to your prejudice.
Never the less by using the best methodology we are able to devise they are the results, and if the results do not accord with your preconceptions that is just too bad, because they will continue to be so, no matter how much you might rage against them.
rcw
The difficulty with being a scientist is that you must accept the hypothesis that best fits the data, and quite often that is one that is completely contrary to your prejudice.
Never the less by using the best methodology we are able to devise they are the results, and if the results do not accord with your preconceptions that is just too bad, because they will continue to be so, no matter how much you might rage against them.
rcw
192 KHz/ 24 bit recording is very convenient. That means I can process the live record of an unique event later that I made with big headroom, with zero clipping possibility, and still get very good result after "amplification".
Comparing preamp with straight wire we should remember that signal source and power amp have output and input impedances, and they are non-linear. Removing buffer between such stages you may add audible distortions. But I know also that it is much easier to make superb line level preamp than microphone or power amp, and if preamps are made properly you can't hear differences between them. I am with SY here.
Comparing preamp with straight wire we should remember that signal source and power amp have output and input impedances, and they are non-linear. Removing buffer between such stages you may add audible distortions. But I know also that it is much easier to make superb line level preamp than microphone or power amp, and if preamps are made properly you can't hear differences between them. I am with SY here.
Last edited:
I'm not at all certain that we could draw such a sweeping conclusion. At best we could conclude that _with that particular complete system_ the added gain stage was transparent. For example, we couldn't also conclude that the speakers were transparent.
But for me, already believing that a line level flat gain stage can be made transparent enough for me, the larger impediment is laziness. What issue bugs me enough to go to the considerable effort of a formal test? I probably should work on that question.
Thanks, as always,
Chris
Distorting my words will get you nowhere.
I wrote in post #27403:
So, I defined "validity of a test" – the test should demonstrate that the subjects being tested do distinguish between two amps having considerable, meaningful and clearly measure differences between them, like the differences between certain SS power amps and certain tube power amps. If and when such gross differences aren't being discernable, or detected, with valid statistical numbers – that test is definitely invalid. Of course it will not reveal more subtle differences.
1. Do you mean by a valid rest, or any DBT?
2. Probably the 'musicality' of a sound setup means nothing to you.
As I wrote earlier in this post:
"Furthermore, to me, the one most important attribute of my sound setup is it's 'musicality', that is, the degree to which music reproduced by the setup is exciting, convincing and enjoyable. I don't care at all what does make my sound setup more 'musical' (to my taste). I don't care at all whether it is some 'real' sound qualities, or 'imagined' sound qualities, or any combination of those two. Hence, I don't care at all if there are things that I may imagine – as long as my setup is more 'musical', or more enjoyable and consistently so – that is all I care about. Since I listen to music at home sighted, I don't care at all how my setup may sound 'blind', or in 'blind tests'. Therefore, blind tests are absolutely useless to me. I care about the degree I enjoy listening to reproduced music, not in any 'scientific' proof of what makes it more or less 'musical', or enjoyable."
No.
As I wrote above:
"For any test to be valid by scientific criteria, it has to be proven to produce positive results when there are big and meaningful differences."
And:
"The test should demonstrate that the subjects being tested do distinguish between two amps having considerable, meaningful and clearly measure differences between them, like the differences between certain SS power amps and certain tube power amps. If and when such gross differences aren't being discernable, or detected, with valid statistical numbers – that test is definitely invalid. Of course it will not reveal more subtle differences."
You are again making the mistake of asserting that since you can measure quite large differences you can also hear them, and any test that proves otherwise must be invalid.
What you are proposing is not a valid test in any scientific sense. The reason being that your argument is a deductive one in which you assert that given both a and b are axioms that are true, then we can deduce that a proposition such as c must be true, this is valid for deductive logic but science is inductive.
In this case instead of a and b being axioms at least one of them is a proposition that you can prove is false, and if you can not prove either one false then c has not been proven false.
All valid scientific tests are of this form, i.e. something is only true in as much as you cannot prove it to be false, the first statement is just that something is true because its axioms are true and just asserts the basic principle of deductive logic.
A double blind test is just such a thing that contains a proposition that can be proved false, in this case the proposition is that since we can measure differences between a and b we can also hear them.
The fact is that if the measured differences are very large we can but after a certain threshold is reached we can't .
Exactly where that threshold is located is up for debate, but the fact that it exists is not.
rcw
What you are proposing is not a valid test in any scientific sense. The reason being that your argument is a deductive one in which you assert that given both a and b are axioms that are true, then we can deduce that a proposition such as c must be true, this is valid for deductive logic but science is inductive.
In this case instead of a and b being axioms at least one of them is a proposition that you can prove is false, and if you can not prove either one false then c has not been proven false.
All valid scientific tests are of this form, i.e. something is only true in as much as you cannot prove it to be false, the first statement is just that something is true because its axioms are true and just asserts the basic principle of deductive logic.
A double blind test is just such a thing that contains a proposition that can be proved false, in this case the proposition is that since we can measure differences between a and b we can also hear them.
The fact is that if the measured differences are very large we can but after a certain threshold is reached we can't .
Exactly where that threshold is located is up for debate, but the fact that it exists is not.
rcw
For any test to be scientifically valid, it must give positive results.
You are welcome to develop a test to your liking that will be proved to be valid in scientific methodology. You are welcome to determine what differences are audible, by any way you may choose, and develop such a test that will give positive results; that those differences will be pointed out definitely. Only such a test will be valid to explore questionable differences.
Those who are convinced that there are no audible differences between various pieces of gear for audio setup can suffice with the cheapest setup, thus saving a lot of money.
Conviction goes both ways.
Enjoying listening to reproduced music is a different thing, it has nothing to do with any conviction.
Objection:
Hypothesis presented as fact. Not uncommon for you.
The other, is that hearing ability is variable between people, regarding wiring and capacity to learn.
Lorentz lost, math does not prove reality, observation begins the process -before any point of determination, and observation remains king throughout.
Lorentz's methods where rightfully dismissed as a dead end and dogmatic. (Solvay, 1927)
The obvious relational aspect is how audiophile reviewers require that the equipment be around for a month or more, in order to really be able to flesh out the given differences, regarding their capacity to hear said differences.
This becomes critical, regarding long term enjoyment of said equipment that may be purchased and enjoyed -or not.
I'm sure that some will ignore this little bit, which turns out to be critical. It's one of those things, one of the core faults of the overly deterministic mind as tied to scientific exploration. A kind of mind that does not belong in the world of science. That is, unless someone wants science to be full of circular logic and dead ends.
Which is kinda weird to see in the face of the highly touted statement that we know maybe a few percent of what this universe may teach us. It's beyond irony, it's sick.
Last edited:
- Status
- Not open for further replies.