I don't know the logic behind this "common criterion". For me 7/8 is not good enough. At least, I have presented better numbers in last year ABX.
There are many things wrong in the way people understand Statistics. There should be no "subjectivity" in Statistics. But it will take great effort to explain logic, so I prefer to save that for another time.
While there should be no and is no "subjectivity in statistics" the constitution of a criterion in an experiment is another subject.
It´s the experimenters choice (and therefore subjective) to set up the experimental conditions and the threshold for rejection of the null hypothesis, if a NHST will be done.
Always curious to know about Tests, which one are you referring to with "last year ABX" ?
If you have good logic, you will love Statistics even more!
Mr Black has 4 daughters. The chance he will have a 5th in 10 weeks is zero. How come? Oh, he took vasectomy. How about adoption? There was unstated ASSUMPTION that adoption was irrelevant.
Mrs Black has 4 daughters. The chance he will have a 5th in 10 weeks is 3.5%. How come? Because she uses pills, and according to this book, the chance for failure of this kind of pill is 0.2%. But something wrong with the Math! No, I know Mrs Black. She wont change her mind, and I know the level of her carelessness (probability she will forget the pills).
So Statistics is about ASSUMPTIONs. It's not just numbers or Math, it's LOGIC!
you appear to have neither maths nor logic. I would step back and listen to those who do and are trying to help your education rather than show how confused you are.
Hehe 😀 That's why I said it took great effort to explain logic. I have made it so simple imho. I can try to make it simpler if you want
Perception Bias?
I think it might be an extra ordinary claim to say that cello distortion here is audible. But apparently Pavel thinks so. So I'm thinking, what would SY think is a proper proof for this kind of claim:
This is interesting, because if you pay careful attention to SY's words, he has also made the exact same extra ordinary claim when he said:
Now let's see what is their "standard" when they have to proof their extra ordinary claim. I believe subjectivists can learn from them.
I think it might be an extra ordinary claim to say that cello distortion here is audible. But apparently Pavel thinks so. So I'm thinking, what would SY think is a proper proof for this kind of claim:
This is interesting, because if you pay careful attention to SY's words, he has also made the exact same extra ordinary claim when he said:
Now let's see what is their "standard" when they have to proof their extra ordinary claim. I believe subjectivists can learn from them.
Statistics (and other parts of math and science) often doesn't appear "logical" or intuitive.
In a group of 100, wouldn't the chances be 1 out of 100 that the same person scores well the second time?
But the chances of that are low enough that I'd want to run more tests on that one person's hearing!
Sort of. The second question is different than the first, namely, is Specific Person X able to distinguish A from B? And the probabilities are the usual numbers, e.g., 95% confidence or whatever level you choose. At 95%, it's 1 in 20 that the results are due to chance from that test. Mixing the statistics can be misleading, but if all the protocols are the same, you can combine the numbers to see the significance.
"Intuitive" or not depends on the ability to use logic or to understand the "subjectivity" of the situation. Statistics is only a tool to MODEL real world situations. As opposed to a mathematical model (which is very "subjective" or less accurate because of unknown/unstated assumptions), a real world has so many variables/information that affect the probability calculation.
I have shown here by example:
So in a real world, the solution is more complicated because we have LESS assumptions (or MORE information).
But logic can be missing from a Mathematical model (like often found in Physics books) simply because the writer cannot see BEYOND the Mathematical model/formula. (Sadly, this is also true when many who claim themselves "objectivist" treats "theory" as a God and couldn't see beyond what is written)
Look at this classics example:
In Mathematical models, infinite number is an "axiom" which is often needed for modeling. It's a magical word that makes a model correct. In real world we deal with finite numbers.
The solution to such Statistical problems can be started with comparing the number of TRIALs with the number of POSSIBILITIES/OCCURRENCES.
Now for the number of trial, "infinite" is unrealistic in the real world (because in a model, infinite+1=infinite, which is wrong). And so is "all the GREAT books", it must be quantified. This is where the "fun" begins (or confusion for some) because so many assumptions can be made.
And then, why we use MONKEY in the model, not ANT? Does it matter? It DOES, in the real world. So a monkey is less intelligence than human (its up to you if you want to assume that chimpanzee is also a monkey, you have to prove it).
What is the possibility of a piano key in the middle of a big piano being hit by a monkey sitting in the middle? Is it the same with possibility of the left-most key? You can construct different models/assumptions.
So, to design a model you MUST make ASSUMPTIONs. It is these assumptions that determine the accuracy of the model. Quite often, the assumption is "wrong".
If I WERE you, what will I say? YOU should know better. (See what assumptions I have made here)
🙂
What just happened?I thought this was a distortion test....
We discussed Statistics, of which understanding is needed in judging if this distortion in the test is AUDIBLE or not. So far, there is no proof that it is audible.
Jay, I would say most things are audible if one's attention is drawn to them - we humans are very good at pushing aside aspects of what we sense, in order to" highlight" what we deem relevant at that moment. Whether something is audible I would put in the same category as "how long is a piece of string" ...
IME getting sound reproduction to work at its best, subjectively, is by lowering the level of distracting elements in the sound - if there is distortion there which you can do nothing about, it's buried deeply within the source signal, then you need to take your mind off that as much as possible - and I've found the best way of doing that is by revealing as much as possible of the actual musical event that was captured - that "image" then can override the distracting distortions - and trigger what I call convincing sound to manifest.
IME getting sound reproduction to work at its best, subjectively, is by lowering the level of distracting elements in the sound - if there is distortion there which you can do nothing about, it's buried deeply within the source signal, then you need to take your mind off that as much as possible - and I've found the best way of doing that is by revealing as much as possible of the actual musical event that was captured - that "image" then can override the distracting distortions - and trigger what I call convincing sound to manifest.
You can hear cables, Frank. Thus our opinion has no value in this engineering context 😀
Imagine the implication of a conclusion (which will be treated as axiom) that a 10% (is that right?) second order distortion is NOT audible. Then why say a bad thing about H2? Isn't it a good idea if Scan Speak tweeters were designed to have very low H3 (and above) at the expense of H2?
If you think that amps sound the same, why not just pick the cheapest that you can differentiate in an ABX and be happy with it???
There are more (and critical) implications if you think about it. I have tried to explain my theory many times in the past but it seemed it is too "complicated" to be understood.
I understand what you are saying. And this new speaker of mine is really surprising me. I have just compared 04a file with another work from Beethoven in the same file format and how can I explain that the file we use for the test here is so lousy in comparison??
A tweeter with ultra high frequency capability has a clear benefit when playing good recording and music from Chesky record and the like but with most recordings you can hear no benefit (only noises). So your approach is to prevent the noises from being induced or created or amplified in the chain. That's fine.
But remember, there's no distracting distortion in a good speaker. I prefer to call it "disturbance" instead of "distortion".
What I can hear is when a system is not firing properly - the reasons could be almost anything, 😀.
Yes, I would agree about a decent speaker having no "distracting disturbances" - ultra high frequency capability may or may not have benefit, I have heard recordings that I'm very familiar with on speakers that would supposedly go way up there, and I have heard nothing magically revealed - instead, typically, large chunks of the soundscape are missing, whole parts of the musical whole are just simply not there, too lost in the system "noisiness". I'm after detail in the musical message, I find many "audiophile" recordings quite boring, often there is a sterile, empty quality about them ...
Yes, I would agree about a decent speaker having no "distracting disturbances" - ultra high frequency capability may or may not have benefit, I have heard recordings that I'm very familiar with on speakers that would supposedly go way up there, and I have heard nothing magically revealed - instead, typically, large chunks of the soundscape are missing, whole parts of the musical whole are just simply not there, too lost in the system "noisiness". I'm after detail in the musical message, I find many "audiophile" recordings quite boring, often there is a sterile, empty quality about them ...
This has just devolved into blah, blah, blah. Nothing useful at all. You've killed a thread again. 😡
I may try another distortion profile, later, if you guys would be interested. Unfortunately Pano is right about some repetitive post with zero value and empty statements.
Well, we are talking about preferences, and distorted vs. undistorted signals. A majority of people preferred the distorted - why? Many forms of music rely on distortion as part of the musical message, as an obvious think Marshall amps here - unless one is fully familiar with the original track it may be very hard to assess which is more 'correct'. And indeed added "distortion" is very much part of mastering these days - a pristine digital take is sent through some analogue stages, or an analogue tape loop to "dirty" it up a bit ... which is the "real" recording now?
- Status
- Not open for further replies.