The idea that "you need to be trained to do ABX" reminds me of this:
https://goldenearsaudio.com/
I suspect the "Golden Ears" name turns off some people as the phrase is often used derisively, but I first saw this over 20 years ago when it was sold on CD, and discussed (or maybe just advertised, the creator was a frequent poster) on rec.audio.pro. As an EE with interest in audio, I was tempted to get it, but then I didn't really want a career in audio due to a probable pay cut. But who knows, I might have ended up making MORE money.
Googling golden ears training finds a few discussions and mentions of competing products. I'm disappointed not to see more discussion of this sort of thing, ere or even on ASR. "How do you know you can hear what you hear?" The answer is too often "I just know."
https://goldenearsaudio.com/
I suspect the "Golden Ears" name turns off some people as the phrase is often used derisively, but I first saw this over 20 years ago when it was sold on CD, and discussed (or maybe just advertised, the creator was a frequent poster) on rec.audio.pro. As an EE with interest in audio, I was tempted to get it, but then I didn't really want a career in audio due to a probable pay cut. But who knows, I might have ended up making MORE money.
Googling golden ears training finds a few discussions and mentions of competing products. I'm disappointed not to see more discussion of this sort of thing, ere or even on ASR. "How do you know you can hear what you hear?" The answer is too often "I just know."
Professional perceptual testing and goofy amateur stuff needs to be differentiated. The role of training as used in professional perceptual testing is pretty well understood based on scientific research. There is a large body of information on perceptual testing science if anyone is interested enough in learning more about how to do it right.
For one example, please see: "Sensory Discrimination Tests and Measurements - Sensometrics in Sensory Evaluation," by Jian Bi.
For one example, please see: "Sensory Discrimination Tests and Measurements - Sensometrics in Sensory Evaluation," by Jian Bi.
Last edited:
Tom, I don't disagree with you.
adason, No. I just felt that listening so intently for a session to have a valid result was fatiguing and the concentration to identify differences and matches made it unpleasant.
I really appreciate the test and it surprised me that the results were in contradiction to my preconceived beliefs.
Snake oil aside, I hope that I always am willing to keep my mind open to challenges to my beliefs.
Some times, I just want to set back and enjoy the music.
I went to an ABBA tribute band concert in Johnson City last night with a good friend (Not my choice of bands, but she wanted to go so I figured what the hell). I have no ABBA albums and was never interested in them. That said, it was a lot of fun,and I recognized almost every song.
adason, No. I just felt that listening so intently for a session to have a valid result was fatiguing and the concentration to identify differences and matches made it unpleasant.
I really appreciate the test and it surprised me that the results were in contradiction to my preconceived beliefs.
Snake oil aside, I hope that I always am willing to keep my mind open to challenges to my beliefs.
Some times, I just want to set back and enjoy the music.
I went to an ABBA tribute band concert in Johnson City last night with a good friend (Not my choice of bands, but she wanted to go so I figured what the hell). I have no ABBA albums and was never interested in them. That said, it was a lot of fun,and I recognized almost every song.
Last edited:
As a general observation: on the one hand, it seems that everybody is able to identify stuff like distortion, sound space, dither, what have you in a casual way with no training and no special preparation or organisation. The results of susch casual activities are often given as gospel.
But when someone asks to do it systematically, like with an ABX where we can get an idea of the actual audibility, all of a sudden people get stressed and fatigued by being 'forced' to listen carefully or 'their brain gets scrambled'. Didn't they listen carefully at the casual listening where they identified the differences??
It all sounds very much like a cop-out to me. Those casual results are not at all confidence-inspiring, also because many of the reports are all over the place.
But it's nice conversation if you have nothing serious to do.
Jan
But when someone asks to do it systematically, like with an ABX where we can get an idea of the actual audibility, all of a sudden people get stressed and fatigued by being 'forced' to listen carefully or 'their brain gets scrambled'. Didn't they listen carefully at the casual listening where they identified the differences??
It all sounds very much like a cop-out to me. Those casual results are not at all confidence-inspiring, also because many of the reports are all over the place.
But it's nice conversation if you have nothing serious to do.
Jan
Goofy amateur stuff? I suppose a corollary could be that professional perceptual testing and training thereof is dressed-up sales and marketing (Shudder! It gives me microplastics just thinking about it!) After all, what would be the purpose of hand-holding an amateur so their results are somehow guided outside of what they would able to achieve independently? The commercial sense would be to sell something. The scientific sense? That would be in the mind of the scientist and I'm not sure I want to go there, either.
For those claiming to hear differences in casual (i.e. sighted, not level matched) listening the stress and fatigue caused by ABx comes mostly from failing to hear any differences. The "night and day" difference heard in casual listening is nowhere to be heard in level matched ABx. At first this causes stress and frustration and later it turns to full denial: ABx is faulty.
There are plenty of self-proclaimed listening experts claiming to hear differences far beyond the thresholds of audibility. Very few (if any) have been able to validate their claims in controlled listening tests. This does not require ABx. However levels need to be matched to 0.1dB and visual cues removed (blind test) to eliminate confirmation bias. Paradoxically it seems that to hear differences many need to listen with their eyes.
There are plenty of self-proclaimed listening experts claiming to hear differences far beyond the thresholds of audibility. Very few (if any) have been able to validate their claims in controlled listening tests. This does not require ABx. However levels need to be matched to 0.1dB and visual cues removed (blind test) to eliminate confirmation bias. Paradoxically it seems that to hear differences many need to listen with their eyes.
Denying some people can heard below average thresholds of audibility is like denying all athletes are exist. But of course it should be measure properly. Human have different ability in all things and there are average value in all of things.
Of course there are people that can hear below average thresholds.
And some can't hear down to the average threshold. That's in the definition of 'average'.
Dutch males are on average 1 meter and 84 cm tall.
Some are taller, some are shorter. Surely we don't need to explain the term 'average' here??
But people consistently rate their ability too high.
If you ask people if they have above or below average IQ, about 80% answers they are above.
Which of course is wrong - it should be 50%, by definition.
Half of the human population doesn't make an IQ of 100. Undeniable.
Jan
And some can't hear down to the average threshold. That's in the definition of 'average'.
Dutch males are on average 1 meter and 84 cm tall.
Some are taller, some are shorter. Surely we don't need to explain the term 'average' here??
But people consistently rate their ability too high.
If you ask people if they have above or below average IQ, about 80% answers they are above.
Which of course is wrong - it should be 50%, by definition.
Half of the human population doesn't make an IQ of 100. Undeniable.
Jan
See the article states a ITU-R BS.1116 compliant listening room is used. If one read the ITU-R BS.1116 (Methods for the subjective assessment of small impairments in audio systems) it even says somthing clear on trained listener and what it is. So interesting reading. https://www.itu.int/dms_pubrec/itu-r/rec/bs/R-REC-BS.1116-3-201502-I!!PDF-E.pdf
For the HD discussion it is interesting to se that a compliant speaker in the compliant room has this HD limits:
'7.2.2.3 Non-linear distortion A constant voltage input signal producing an average sound pressure level (SPL) of 90 dB is supplied to the loudspeaker. Related to that SPL, no harmonic distortion component, in the fundamental frequency range 40 Hz to 16 kHz, shall exceed the following values: –30 dB (3%) for f < 250 Hz – 40 dB (1%) for f > 250 Hz'
So not very hard to get an amp that is 20 dB lower in HD
Because they did not test it properly, so you can believe or not, it does not matter.
By the way, I took IQ test in high school, the result was above average, but the score was strange. I get very high score in language and logic, but other are average.
That's the point. If you haven't really tested your ability, you rate it consistently too high.
This is the reason double blind tests are so stressful. You are asked to verify that indeed you are above average.
The easiest way out for your ego is to state 'ABX (or whatever) sucks' with no reasons. It makes your ego happy again and you Feel Good without having accomplished anything useful ;-)
Jan
This is the reason double blind tests are so stressful. You are asked to verify that indeed you are above average.
The easiest way out for your ego is to state 'ABX (or whatever) sucks' with no reasons. It makes your ego happy again and you Feel Good without having accomplished anything useful ;-)
Jan
I never claimed that my listening skill is above average. But I believe that someone have better listening skill than me 😉
Interesting to this discussion about average and normalized groups. The ITU test is not about average and normalizing, interestingly enough.
First there are a preselection of participant on the listening tests. Then the researchers are aloved to post unselect participants with bad scoring on predefined tests. So not normalized group in any way. They just want the statistics from the best or maybe good enough listeners. 😎
First there are a preselection of participant on the listening tests. Then the researchers are aloved to post unselect participants with bad scoring on predefined tests. So not normalized group in any way. They just want the statistics from the best or maybe good enough listeners. 😎
Last edited:
I'm on the limit, here, for my IQ, but my understanding is that they put some clips in the test that is not to easy to detect for a worthy participant, but not to hard either. If you dont nail those, youre out, regardless of your score on the other real tests.
Sad people have to make decisions based on listening to bog standard crap systems. If the test system was a lot better it wouldn’t be stressful. Hel-loo!