Skip to main content ↑ ↓

Notice

Please note that most of the software linked on this forum is likely to be safe to use. If you are unsure, feel free to ask in the relevant topics, or send a private message to an administrator or moderator. To help curb the problems of false positives, or in the event that you do find actual malware, you can contribute through the article linked here.
Topic: How do you listen to an ABX test? (Read 357551 times) previous topic - next topic
0 Members and 1 Guest are viewing this topic.

How do you listen to an ABX test?

Reply #1350 –
Quote from: Arnold B. Krueger on
Quote from: jkeny on
Quote from: Wombat on
Quote from: jkeny on
Quote from: Wombat on
I still don't see your positive abx result of the same files Arny did this null result.

So it's fine by you to lie about doing an ABX "test" by randomly guessing as long as there are no positive ABX results in evidence

Again, this reveals the lack of integrity, & "scientific principles" that permeates HA

He didn't lie. If you think he did it wrong just do it better and not only talk.

He didn't lie???
He went into great detail about the conditions of the test
"The comparison was of a 24/192 file titled "15 Haydn_ String Quartet In D, Op_ 76, No_ 5 - Finale" to a 256 kbps MP3 based on it prepared using Lame. This file happens to be exact same one that was recently the star of an AES paper that Amir has been heaping praise onto."

"The playback DAC+ headphone amp was an Asus U7 which is capable of handling 192 KHz sample rate files with good performance"

"The playback transducers were the highly regarded Audio Technica ATH-M50 headphones."

All the above details & not a word about the fact that he didn't actually listen but just randomly guessed.


As usual, no quotes related to the overall conditions under which the above was done. Just some deceptive out-of-context quotes which is what we expect from Keny, so no harm, no foul.


Still lying & dodging, Arny. Here's your full post containing the above quotes & where you mention nothing of having cheated the results. Not until questioned about the timing do you trot out your excuses. So you initially give all this detail about the condition of the test but leave out the fact that you guessed randomly to get your results. You were asked by Amir to run a simple test & report the results. You didn't run a "test" you simply hit buttons randomly & tried to pass it off as a "test"

Who are you trying to fool - yourself, maybe?

Quote
Quote from: Originally Posted by amirm link=msg=0 date=

That's right. Anything but running a simple test and reporting on its results.

The comparison was of a 24/192 file titled "15 Haydn_ String Quartet In D, Op_ 76, No_ 5 - Finale" to a 256 kbps MP3 based on it prepared using Lame. This file happens to be exact same one that was recently the star of an AES paper that Amir has been heaping praise onto.

The playback DAC+ headphone amp was an Asus U7 which is capable of handling 192 KHz sample rate files with good performance.

The playback transducers were the highly regarded Audio Technica ATH-M50 headphones.

The ABX log is as follows:

foo_abx 2.0 beta 4 report
foobar2000 v1.3.5
2015-01-06 21:04:53

File A: 15 Haydn_ String Quartet In D, Op_ 76, No_ 5 - Finale - Presto + cues 256kbps.mp3
SHA1: f24d8c506ae5d38fd7d3a8e7700ee8595cd5e025
File B: 15 Haydn_ String Quartet In D, Op_ 76, No_ 5 - Finale - Presto + cues.wav
SHA1: 961320fa0baa1983130304bed02df943a32cfe25

Output:
DS : Primary Sound Driver

21:04:53 : Test started.
21:05:18 : 00/01
21:05:39 : 01/02
21:06:39 : 02/03
21:06:45 : 03/04
21:06:47 : 04/05
21:06:50 : 04/06
21:06:54 : 04/07
21:06:56 : 05/08
21:06:58 : 06/09
21:06:59 : 07/10
21:07:01 : 07/11
21:07:04 : 07/12
21:07:05 : 08/13
21:07:08 : 08/14
21:07:10 : 08/15
21:07:31 : 08/16
21:07:31 : Test finished.

----------
Total: 8/16
Probability that you were guessing: 59.8%

-- signature --
b54eb2a632d09ae60dbb1c13774d4152ee32f110


In short my ears are as I have said not up to this sort of thing, but to please Amir...


It's not your ears that weren't "up to this sort of thing" - it was your motivation & morality!

How do you listen to an ABX test?

Reply #1351 –
Quote from: xnor on
How can you not see the level of hypocrisy and self-deception?

The hypocrisy & self-deception is all yours. Your hypocrisy that refuse to admit that most blind tests are of the type I gave as an example. Your attempts at deception (including self-deception) that ABX tests are exempt from needing some indicator of the quality of the test. Your refusal to admit that most blind tests are of the type I gave as an example.

How do you listen to an ABX test?

Reply #1352 –
Quote from: xnor on
Despite your latest contradictions and trying to again divert attention to one of Arny's logs (which already has been dealt with earlier in this thread, after educating you on these things I didn't think you'd make this asinine point again) this is exactly what you do.
Arny's cheated test is a perfect example of why quality measures are needed for ABX tests. The only other solution is to ALWAYS use a trusted proctor for all tests - which is what you people invoke when faced with positive results you don't agree with .

So which is it?:
- accept all ABX results without qualification - without knowing their quality/sensitivity/validity
- use a trusted proctor for all ABX tests
- use some other sort of QA mechanism

How do you listen to an ABX test?

Reply #1353 –
Quote from: jkeny on
Quote from: Arnold B. Krueger on
Quote from: jkeny on
Quote from: Wombat on
Quote from: jkeny on
Quote from: Wombat on
I still don't see your positive abx result of the same files Arny did this null result.

So it's fine by you to lie about doing an ABX "test" by randomly guessing as long as there are no positive ABX results in evidence

Again, this reveals the lack of integrity, & "scientific principles" that permeates HA

He didn't lie. If you think he did it wrong just do it better and not only talk.

He didn't lie???
He went into great detail about the conditions of the test
"The comparison was of a 24/192 file titled "15 Haydn_ String Quartet In D, Op_ 76, No_ 5 - Finale" to a 256 kbps MP3 based on it prepared using Lame. This file happens to be exact same one that was recently the star of an AES paper that Amir has been heaping praise onto."

"The playback DAC+ headphone amp was an Asus U7 which is capable of handling 192 KHz sample rate files with good performance"

"The playback transducers were the highly regarded Audio Technica ATH-M50 headphones."

All the above details & not a word about the fact that he didn't actually listen but just randomly guessed.


As usual, no quotes related to the overall conditions under which the above was done. Just some deceptive out-of-context quotes which is what we expect from Keny, so no harm, no foul.


Still lying & dodging, Arny. Here's your full post containing the above quotes & where you mention nothing of having cheated the results. Not until questioned about the timing do you trot out your excuses. So you initially give all this detail about the condition of the test but leave out the fact that you guessed randomly to get your results. You were asked by Amir to run a simple test & report the results. You didn't run a "test" you simply hit buttons randomly & tried to pass it off as a "test"

Who are you trying to fool - yourself, maybe?

Quote
Quote from: Originally Posted by amirm link=msg=0 date=

That's right. Anything but running a simple test and reporting on its results.

The comparison was of a 24/192 file titled "15 Haydn_ String Quartet In D, Op_ 76, No_ 5 - Finale" to a 256 kbps MP3 based on it prepared using Lame. This file happens to be exact same one that was recently the star of an AES paper that Amir has been heaping praise onto.

The playback DAC+ headphone amp was an Asus U7 which is capable of handling 192 KHz sample rate files with good performance.

The playback transducers were the highly regarded Audio Technica ATH-M50 headphones.

The ABX log is as follows:

foo_abx 2.0 beta 4 report
foobar2000 v1.3.5
2015-01-06 21:04:53

File A: 15 Haydn_ String Quartet In D, Op_ 76, No_ 5 - Finale - Presto + cues 256kbps.mp3
SHA1: f24d8c506ae5d38fd7d3a8e7700ee8595cd5e025
File B: 15 Haydn_ String Quartet In D, Op_ 76, No_ 5 - Finale - Presto + cues.wav
SHA1: 961320fa0baa1983130304bed02df943a32cfe25

Output:
DS : Primary Sound Driver

21:04:53 : Test started.
21:05:18 : 00/01
21:05:39 : 01/02
21:06:39 : 02/03
21:06:45 : 03/04
21:06:47 : 04/05
21:06:50 : 04/06
21:06:54 : 04/07
21:06:56 : 05/08
21:06:58 : 06/09
21:06:59 : 07/10
21:07:01 : 07/11
21:07:04 : 07/12
21:07:05 : 08/13
21:07:08 : 08/14
21:07:10 : 08/15
21:07:31 : 08/16
21:07:31 : Test finished.

----------
Total: 8/16
Probability that you were guessing: 59.8%

-- signature --
b54eb2a632d09ae60dbb1c13774d4152ee32f110


In short my ears are as I have said not up to this sort of thing, but to please Amir...


It's not your ears that weren't "up to this sort of thing" - it was your motivation & morality!


Thanks for reproducing Amir's taunts, but you still didn't reproduce my posts resisting doing the test, and my explanation of what happened which was in other posts. That's intellectual deception, and its what you seem to do oh so naturally without even noticing, John.

Anybody who has hands-on experience with such things knows that distinguishing a well-made 256k MP3 from the wave file it is made from is far from being a slam dunk. Lots of people who posture about such things (including by implication one John Keny) get embarassed by DBTs.

Depending on the music, lots of people never hear any differences in tests like these no matter how hard they try. John, while you have tried to obfuscate it in this thread, I'm 68 and my hearing is not as good as it was 20-40 years ago. Since you eschew DBTs, you have no way of knowing what you can and can't hear due to the overwhelming false positives in your naive casual audiophile sighted evaluations.

How do you listen to an ABX test?

Reply #1354 –
Quote from: Arnold B. Krueger on
Thanks for reproducing Amir's taunts, but you still didn't reproduce my posts resisting doing the test, and my explanation of what happened which was in other posts. That's intellectual deception, and its what you seem to do oh so naturally without even noticing, John.
Oh, you want me to reproduce a whole thread when people can read all of your posts before & after your posted results

So, your excuse for cheating is that Amir taunted you - ah, that's sweet & child-like in it's naivety. Does this justify your cheated test?

Quote
Anybody who has hands-on experience with such things knows that distinguishing a well-made 256k MP3 from the wave file it is made from is far from being a slam dunk. Lots of people who posture about such things (including by implication one John Keny) get embarassed by DBTs.
So, let me restate your logic here - if a difference is small & difficult to discern then it's OK to press random selections?

Quote
Depending on the music, lots of people never hear any differences in tests like these no matter how hard they try. John, while you have tried to obfuscate it in this thread, I'm 68 and my hearing is not as good as it was 20-40 years ago. Since you eschew DBTs, you have no way of knowing what you can and can't hear due to the overwhelming false positives in your naive casual audiophile sighted evaluations.

Yes, I agree that well-made 256K MP3 is very good but that doesn't mean that therefore people doing a test should not bother to listen. It doesn't mean that a test done like this should be considered a valid test.It doesn't mean that it should be posted or displayed as an actual test.

All these things you are suggesting people should do by your posted example & your subsequent attempts at defending such a cheated test.

Your excuse about having done the test earlier & losing the log file, I don't believe

Your test needed a trusted proctor & I'm coming to the conclusion that a proctor is the only way to run a valid/reliable ABX test

How do you listen to an ABX test?

Reply #1355 –
Quote from: Arnold B. Krueger on
John, while you have tried to obfuscate it in this thread, I'm 68 and my hearing is not as good as it was 20-40 years ago. Since you eschew DBTs, you have no way of knowing what you can and can't hear due to the overwhelming false positives in your naive casual audiophile sighted evaluations.

So your excuses mount up:
- you can't say anything about the multitone tests you claimed to have been running since the 1990s because you are away from home?
- you have to pretend to do ABX tests because Amir taunts you & you are 68 with shot hearing?

Here's a thing, that you should have learned in your 68 years of being on this earth, Arny - "a liar needs a good memory"

How do you listen to an ABX test?

Reply #1356 –
Quote from: jkeny on
- you can't say anything about the multitone tests you claimed to have been running since the 1990s because you are away from home?


I can say what I will but it will be lacking supporting evidence. I can see where a liar and blow hard whose only source of evidence is what he makes up out of his own imagination won't appreciate this.

How do you listen to an ABX test?

Reply #1357 –
I see we've reverted to the original mudslinging about Arnold giving up. John, it's clear you have nothing more to say on the matter.  We have already rebutted the issue of listener selection.  Painting Arny as a liar (though only through intellectually dishonest means by quoting post #35 of a disussion and pretending as if post #21 #19 didn't occur) and saying he's dodging doesn't change the landscape. It's just more means of off-topic, ad-hominem deflection on your part.

The multitone angle is yet another smokescreen.  You're just going for what you think is low-hanging fruit; though it is completely tangential to this course you set us on by positing unsubstantiated "preponderance" claims.

How do you listen to an ABX test?

Reply #1358 –
Quote from: jkeny on
The hypocrisy & self-deception is all yours. Your hypocrisy that refuse to admit that most blind tests are of the type I gave as an example. Your attempts at deception (including self-deception) that ABX tests are exempt from needing some indicator of the quality of the test. Your refusal to admit that most blind tests are of the type I gave as an example.

That's complete garbage and quite frankly another ones of your disgusting lies.

It is complete nonsense to say that "most blind tests" are of the type of an ABX log you keep posting. Most blind tests are done by people who want to prove their claims to be true. I have told you this many times. Please have your brain checked.
Besides, you still haven't presented a shred of evidence.

It is a lie to say that I generally exempt ABX from needing some form of validation. I have explained different types (free-for-all public, personal, ...) of ABX tests and their implications to you, I have explained how you can easily check the conditions, the participants, the test protocol .. if you're doing a more formal test.
You said that is not enough for you, but that you accept biased sighted listening claims as long as they agree with your opinion! Give me a break.

Of course I do not admit to your lies being true. I'm not a liar.
"I hear it when I see it."

How do you listen to an ABX test?

Reply #1359 –
Quote from: jkeny on
Arny's cheated test is a perfect example of why quality measures are needed for ABX tests.

No John, you have repeatedly said you reject blind test therapy for your audiophile disorder condition. That adult tests for infantile beliefs are "stupidity" due to having controls. No "quality" measures for ABX is going to change the infantile beliefs of the Dunning-Kruger audiophiles. You've admitted this yourself!
Adults are well aware of what blind tests (like ABX) are and why they are needed. As well as "quality" measures associated. Dunning-Kruger has stripped you of the ability to be even cognizant of what ABX is, hence you referring in desperation and confusion to the PFM tomfoolery as an ABX "test". You are simply not in a position to criticize that which you can't comprehend...and reject.

Quote from: jkeny on
The only other solution is to ALWAYS use a trusted proctor for all tests

No John, proctoring of test done by known shysters and industry "biz" men who have a dire need for blind tests to be avoided...or cheated, is necessary. AES submittal type stuff is always proctored.

Quote from: jkeny on
which is what you people invoke when faced with positive results you don't agree with .

John, the "positive results" that your Santa hearing friends have caused you to "verify", are not "tests" of sound. They are IQ tests that you are failing. Unfortunately, things like expectation bias, power of suggestion, perception fallibility, etc, etc,, are all adult topics that puerile minds cannot comprehend. Hence the circular logic of "verification".
So while adults will believe you and your friends did indeed hear Santa, in your minds, we will not consider those "results" to be proof positive of his existence. Not going to get agreement there at all. That is an impasse you will be unable to surmount. Despite all the toy chucking and foot stomping.
Loudspeaker manufacturer

How do you listen to an ABX test?

Reply #1360 –
Quote from: jkeny on
The only other solution is to ALWAYS use a trusted proctor for all tests - which is what you people invoke when faced with positive results you don't agree with .

No.
If the person posting the positive log is not demonstrably dishonest (unlike amir and you) and does not have a strong vested interest in producing a positive result regardless of audibility AND if the audible difference is plausible (which is NOT the case if we talk about differences at 21+ kHz but the person's hearing doesn't even extend past 16k, and the used earphones also roll-off before that, and possibly even the sound card in the laptop..) then we can reasonably say: sure, there's very likely an audible difference here.

If these people however even go so far as proclaiming their logs as "undeniable truth", with some of the underlying files actually being faulty, then they have disqualified themselves from any rational and honest discussion.

I have explained this before, but your cognitive dissonance seems to prevent you from thinking.


Your next question should also be answered with that, but since you mention QA I have to point out your hypocrisy again:
Accepting claims based on biased sighted listening as long as they agree with my opinion is not a "QA mechanism", yet your business operates on it. It is a mechanism of confirming your own biases.
You have nothing but claims you agree with based on your opinion.


edit: corrected "unlike", thanks greynol
"I hear it when I see it."

How do you listen to an ABX test?

Reply #1361 –
Quote from: Arnold B. Krueger on
<snip>

It has been conclusively proven that casual sighted listening evaluations are 100% susceptible to false positives and false negatives, and yet you persist in say all of the above?

Please tell me how to quantify the false positive and false negatives in a sighted listening evaluation.  I want peer reviewed scholarly articles, and math on the same grade as we have for DBTs.


As it is always just a matter of probabilities it is virtually impossible to quantify "false positives and false negatives in a sighted listening evaluation" , but that holds true for controlled other listening tests as well.
As i´ve already stated in an answer to "kohlrabi" in a NHST you are in reality not testing the null hypothesis but instead evaluate the data _under_ the assumption that the null hypothesis is true.

"sighted" and "blind" tests are just both needed, any condemnation isn´t appropriate and does not help either.

In other sensory fields people are routinely asking an expert; if he thinks the difference is really, really small (means difficult to detect) then controlled test were done.
jkenys concern are not unresonable (although i would accept additional runs with posivite controls on appropriate sensitivity niveaus, be it in ABX or A/B or any other protocol) and obviousbly the inventors of the ABC/HR shared these concerns. The have developed imho a quite ingenious protocol which enables monitoring on the fly or additional analysis capabilites post hoc .
The authors of the BS 1116 do emphasize for good reason on the special need for additional control runs (and carefull reconsideration of the test) if the percentage of "no difference" gets to high.
Otoh the ABC/HR relies on the existance of a true reference and while that one is easy to identify in the case of codecs it isn´t wrt to other gear in most cases.

Does it really help during the evaluation of tests done, if somebody assures that proper training or incorporation of positive controls (or negative controls in case of a posivite result) could have (should have) been done, if nobody really knows that it _was_ done?

Not to mention that there surely exists a difference depending on the participants of any test. If a listener is routinely listening to files while using a computer it will be much easier for him to do some tests, but it obviously is much more difficult for someone who uses "classical" hifi gear.
I am a bit surprised that the posters hear negate the possibility that such listener could have difficulties with a certain test protocol including sort of rapid switching. Never seen participants getting lost in switching furiously from sample to sample?

How do you listen to an ABX test?

Reply #1362 –
...but the "classical" listener will have no difficulty with MUSHRA-style tests?

...so we should simply accept claims they make from sighted evaluations, or just the ones we can corroborate with our own sighted evaluations, or ... ?

But let's back up; how is giving the listener a priori knowledge of which DUT is being auditioned in a difference test "needed"?!?

What about in cases where the listener is the one claiming obvious differences when he has a priori knowledge, which is the usual situation in this community?

How do you listen to an ABX test?

Reply #1363 –
Quote from: greynol on
I see we've reverted to the original mudslinging about Arnold giving up. John, it's clear you have nothing more to say on the matter.  We have already rebutted the issue of listener selection.  Painting Arny as a liar (though only through intellectually dishonest means by quoting post #35 of a disussion and pretending as if post #21 didn't occur) and saying he's dodging doesn't change the landscape. It's just more means of off-topic, ad-hominem deflection on your part.
What are you talking about post #35 & #21? Are you are referring to AVS because I didn't mention anything about post #35 which is about "fair use & RIAA" post #21 is also about " fair use policy". I really don't know what you are getting at with your cryptic post numbering system.

Arny, has been seen to post a set of ABX results which show that for the majority of the trials he wasn't listening. You call it what you want - I call it deceit. But the point is what controls that there are in the test to stop the likes of Arny's behaviour? You talk about screening - so what are you going to screen? Arny says he can hear 0.5dB differences - is that what you screen? This isn't the same as being able to differentiate 256K mp3 files from lossless. So what is your screening which would have eliminated him from the test?

Surely, you would have expected his own integrity to eliminate himself from the test but it didn't.

So how are you going to prevent others who behave like Arny & submit false/invalid results? 

Quote
The multitone angle is yet another smokescreen.  You're just going for what you think is low-hanging fruit; though it is completely tangential to this course you set us on by positing unsubstantiated "preponderance" claims.

It's just another one of Arny's holes into which he dug himself & can't extricate himself. It's a systemic problem with him

How do you listen to an ABX test?

Reply #1364 –
Quote from: xnor on
It is a lie to say that I generally exempt ABX from needing some form of validation. I have explained different types (free-for-all public, personal, ...) of ABX tests and their implications to you, I have explained how you can easily check the conditions, the participants, the test protocol .. if you're doing a more formal test.

So, tell me how your validation will show up the likes of Arny's behaviour?

How do you listen to an ABX test?

Reply #1365 –
#19 and I will change that now.
Quote from: Arnyk@AVS link=msg=0 date=
But this misses the point. I have repeatedly reported my age (68) and hearing issues (not as sharp as my hearing was 10 years ago and not by a little). This raises the question of what relevance do my test results have to AVS readers? I did all this work around the year 2000 when my hearing was far better, and AFAIK not a lot has changed since then, except perhaps the unnecessarily aggressive and personal rhetoric. The tone is about the same but the source changed.


As I said, your going after Arny isn't getting you anywhere; it just makes you look petty.

If you don't like the outcome if a test because of a problem with the data, make adjustments and re-run the test.  All we know from your specific example is that Arny did not demonstrate the ability to distinguish an mp3 from its source; no more, no less.  If the question is whether it can be done by anyone at all, find an enthusiastically willing participant to take the test.  Maybe someone who claims to be able to hear the difference in the first couple of notes?

How do you listen to an ABX test?

Reply #1366 –
Quote from: greynol on
As I said, your going after Arny isn't getting you anywhere; it just makes you look petty.

His ABX results are a perfect example of the type of invalid results that are not trapped as invalid.
It is not unusual to start off an ABX test listening in the first couple of trials& when no difference is heard stop listening thereafter for the remaining trials. Thus completing the "test" run & producing a set of results - results of what value.

This is condoned here as was posted by Wombat along the lines of "so what, life's too short" - fine but don't do the test if you don't really want to participate.

As Amir said, pool Arny's test in with Amir's positive result & what do we get - a dilution of a valid effort (Amir's) by Arny's invalid results

How do you listen to an ABX test?

Reply #1367 –
Sigh. Pelmazo already took you to the woodshed on this.

If you don't like the outcome if a test because of a problem with the data, make adjustments and re-run the test. All we know from your specific example is that Arny did not demonstrate the ability to distinguish an mp3 from its source; no more, no less. If the question is whether it can be done by anyone at all, find an enthusiastically willing participant to take the test. Maybe someone who claims to be able to hear the difference in the first couple of notes?

Xnor already addressed your apparent desire to "pool" results several times. You evaded each and every instance.

How do you listen to an ABX test?

Reply #1368 –
Quote from: jkeny on
His ABX results are a perfect example of the type of invalid results that are not trapped as invalid.


Keny you are still lying to the forum by not presenting all of my comments about the test.

I've done a number of serious ABX tests since then with positive outcomes. You even recognized them in this thread and again made fun of my age-related hearing problems.

How dishonest and petty do you want to appear, John? I can't imagine any sane person doing what you are doing to yourself here.

How do you listen to an ABX test?

Reply #1369 –
Quote from: greynol on
#19 and I will change that now.
People can read it themselves - that's why I gave a link to the thread. If you want to explicitly quote it, go ahead

Quote
As I said, your going after Arny isn't getting you anywhere; it just makes you look petty.
A bit of a one-sided view, don't you think? Arny has been calling me a liar for most of this thread, says my business is a failed entity & numerous other slanderous posts.

Quote
If you don't like the outcome if a test because of a problem with the data, make adjustments and re-run the test.
You still don't get it, do you - it isn't valid data - it's somebody submitting data pretending it to be valid. It's not about whether anybody has or will post positive ABX results of the same files. 
Quote
All we know from your specific example is that Arny did not demonstrate the ability to distinguish an mp3 from its source; no more, no less.
Anybody can see from his ABX log that he randomly guessed trials 4 to 15. If this is what you call valid data then it reveals a lot
Quote
If the question is whether it can be done by anyone at all, find an enthusiastically willing participant to take the test.  Maybe someone who claims to be able to hear the difference in the first couple of notes?
Nothing got to do with it - you miss the point completely, yet again, - the random guessing was presented as valid test data

How do you listen to an ABX test?

Reply #1370 –
Quote from: Arnold B. Krueger on
Quote from: jkeny on
His ABX results are a perfect example of the type of invalid results that are not trapped as invalid.


Keny you are still lying to the forum by not presenting all of my comments about the test.
Does anything excuse your posting of these results as a valid attempt at an ABX test?

Quote
I've done a number of serious ABX tests since then with positive outcomes. You even recognized them in this thread and again made fun of my age-related hearing problems.

How dishonest and petty do you want to appear, John? I can't imagine any sane person doing what you are doing to yourself here.

Again, I will reiterate - you have stated that your hearing is shot (in some areas) - this should disqualify you from ABX tests as a matter of course - instead you posted random guesses as "valid" results when all it indicates is that you didn't do the test honestly.

Irrespective of whether you can or can't differentiate 256K mp3, the test results you posted are from an invalid test & they serve to show how easy it is to produce negative results in an ABX test. Positive results are far more difficult to achieve because of all the controls in place to trap false positives. 

Nobody has yet answered what ABX controls would catch the likes of Arny's results as being invalid

How do you listen to an ABX test?

Reply #1371 –
Quote from: jkeny on
Quote
If the question is whether it can be done by anyone at all, find an enthusiastically willing participant to take the test.  Maybe someone who claims to be able to hear the difference in the first couple of notes?
Nothing got to do with it - you miss the point completely, yet again, - the random guessing was presented as valid test data

I haven't missed the point at all, but you're trying to make it look that way.

Let's now enter the part of my post that you conveniently omitted:
Quote from: greynol on
Xnor  already addressed your apparent desire to "pool" results several times.  You evaded each and every instance.

To re-ask one of his questions, what is the purpose of the test you're trying to conduct?

Why should we be interested in pooling Arny's results with other results?  What hypothesis is it that you wish to test?

How do you listen to an ABX test?

Reply #1372 –
Quote from: jkeny on
Nobody has yet answered what ABX controls would catch the likes of Arny's results as being invalid

Nonsense.

GIGO.  You wish to re-frame the facts of the situation in order to answer a question that was never being asked and then pretend to score a huge win.  It's laughable.

How do you listen to an ABX test?

Reply #1373 –
Quote from: jkeny on
Quote
All we know from your specific example is that Arny did not demonstrate the ability to distinguish an mp3 from its source; no more, no less.
Anybody can see from his ABX log that he randomly guessed trials 4 to 15. If this is what you call valid data then it reveals a lot


The data is in the log to determine his lack of effort, is it not? How else would you be discussing this if it wasn't?

How do you listen to an ABX test?

Reply #1374 –
Quote from: greynol on
Quote from: jkeny on
Quote
If the question is whether it can be done by anyone at all, find an enthusiastically willing participant to take the test.  Maybe someone who claims to be able to hear the difference in the first couple of notes?
Nothing got to do with it - you miss the point completely, yet again, - the random guessing was presented as valid test data

I haven't missed the point at all, but you're trying to make it look that way.
Ok so your comment "If you don't like the outcome if a test because of a problem with the data, make adjustments and re-run the test." is incorrect - the data should be discarded as invalid, like in all well-run tests. It's not a case of not liking the outcome of the test, it's that the data has no validity.

Quote
Let's now enter the part of my post that you conveniently omitted:
Quote from: greynol on
Xnor  already addressed your apparent desire to "pool" results several times.  You evaded each and every instance.

To re-ask one of his questions, what is the purpose of the test you're trying to conduct?

Why should we be interested in pooling Arny's results with other results?  What hypothesis is it that you wish to test?
What are you talking about - how do we know that 256K mp3 impairments are small? Because of the accumulated (pooling) results of many ABX tests. Would JJ have accepted Arny's results as valid - no!