That is exactly what I'm working on now (not auditory perception), and in my work, the answer is not to attempt to make each error as close to perceptual thresholds as possible, rather to determine which combination of errors is most tolerated by the brain as "transparent" enough for the goal.
One could argue Earl Gedlee got a good chunk of the way there some years ago but the results were unpalatable to many such that no attempts to seriously duplicate or debunk were made?
How does one design for "transparency" if the various components of transparency involve trade-offs? How does one optimize the trade-off to make the result "most-transparent"?
That is exactly what I'm working on now (not auditory perception), and in my work, the answer is not to attempt to make each error as close to perceptual thresholds as possible, rather to determine which combination of errors is most tolerated by the brain as "transparent" enough for the goal.
Excellent point & something that could be considered in the context of audio reproduction - we all know that as we improve one part of our playback system flaws in other parts of the system become evident - flaws that were masked before. One could almost claim that digital audio was a victim of it's own success - it's low noise perhaps exposes some elements that were masked before. How perceptually aggravating this is can only be judged by our auditory perception.
I'm can understand the approach that if we strive for each device in the chain to be as flawless as possible then we should have a chain of devices that is also flawless. But two questions arise - what do we consider flawless enough - what test signals, what is flawless enough ? And do we not have to consider the interaction between devices in the system (again with the correct test signals) - it's not necessarily behaving as flawless as when it is measured alone.
Not taking that approach & trying to balance the combination of errors to be as perceptually transparent as possible is a tough job - I wish you the best in this endeavour.
Isn't it more, just fit for purpose? When I was a BT apprentice, we were told that the design of the old dial telephone circuit involved first of all getting it to work and then a process of component removal and shared circuit elements to get it as simple as possible. It's actually a beautiful example of a minimalist designThis sounds like usual business advice to "match" the components for the best sound. Like "if you buy our XX power amplifier, you should also buy our YY preamplifier to get the best sound". Is that what you are saying? 😀
,
No, but I see your point, and I suppose my work could be used toward such a goal. But what I do is pretty far from sales. It is used to inform the design of both XX and YY. This works for the company that is funding my research, since all their products are all-in-one. No one buys multiple products to work together.This sounds like usual business advice to "match" the components for the best sound. Like "if you buy our XX power amplifier, you should also buy our YY preamplifier to get the best sound". Is that what you are saying? 😀
One could argue Earl Gedlee got a good chunk of the way there some years ago but the results were unpalatable to many such that no attempts to seriously duplicate or debunk were made?
That's really interesting to me! I don't know his work. Although I can certainly google that myself, do you have a favorite link that describes this?
One could argue Earl Gedlee
Earl Geddes, please.
GedLee stands for Geddes + Lee. 😀
View attachment Result - Intelligibility Test - Headphones.txtAttached is a standard speech intelligibility test file.
After a few folks confirm they have done this I will show the word list.
\You can use headphones or near field loudspeakers and try out your hearing abilities or your actual sound system and include room effects.
One listen/write, one listen back with one correction ?.
Dan.
Last edited:
Sorry Pavel, I was going to type 'Gedlee Metric' then changed to 'Earl Geddes' and fingers and brain got confused
GedLee LLC is a good starting link for information on Earl's work.
GedLee LLC is a good starting link for information on Earl's work.
Not taking that approach & trying to balance the combination of errors to be as perceptually transparent as possible is a tough job - I wish you the best in this endeavour.
Thanks! It is tough. I spent my first 20 years using a reductionist approach, and then dabbled in gestalt methods. Neither extreme gives the whole picture. Bayes is my friend. I know I'll die before the "answer" is found, but I just want to get one smidge closer.
How does one design for "transparency" if the various components of transparency involve trade-offs? How does one optimize the trade-off to make the result "most-transparent"?
That is exactly what I'm working on now (not auditory perception), and in my work, the answer is not to attempt to make each error as close to perceptual thresholds as possible, rather to determine which combination of errors is most tolerated by the brain as "transparent" enough for the goal.
Isn't there a relation to subliminal stimuli as well?
Ferrè, E. R., Sahani, M., & Haggard, P. (2016). Subliminal stimulation and somatosensory signal detection. Acta Psychologica, 170, 103–111
Isn't there a relation to subliminal stimuli as well?
Ha! Subliminal and sub-cognitive perception is a lightning rod for ferocious criticism on audio forums.
But anyone whose research uses psychometric function fits (including mine) sees this routinely. It is a definitional issue. Any responses that are statistically significantly above random but below the arbitrarily defined "threshold" fit those terms. The subjects invariably state they are guessing, but their data shows something is going on. I use the best fit of their data to model, and so the subliminal/subcognitive data is included in the models.
Ha! Subliminal and sub-cognitive perception is a lightning rod for ferocious criticism on audio forums.
No way; you must be kidding..... 🙂
responses that are statistically significantly above random but below the arbitrarily defined "threshold" fit those terms. The subjects invariably state they are guessing, but their data shows something is going on. I use the best fit of their data to model, and so the subliminal/subcognitive data is included in the models.
I mentioned it recently when explicating the construction of a basic psychometric function and the meaning of the term "threshold" and subsequently a member proposes that it would be better to drop everything below the 50% mark.
Somehow i got away with this kind of "heresy", although it must have been one of my "toxic" posts where the "BS is wrapped in a thick crust of theoretical knowledge" .
Actually i was moved being promoted from the "know nothing categoy" to the next higher level.........
That's a really nice ref with which I was unfamiliar. Thanks!Ferrè, E. R., Sahani, M., & Haggard, P. (2016). Subliminal stimulation and somatosensory signal detection. Acta Psychologica, 170, 103–111
I suspect it's human nature to dislike facts that an opponent can use against you. It's important for me to avoid that, but I suspect I don't always succeed.No way; you must be kidding..... 🙂
I mention it because I try to remember to cut people some slack when they act so... naturally. 😉
That's a really nice ref with which I was unfamiliar. Thanks!
Your welcome!
I suspect it's human nature to dislike facts that an opponent can use against you. It's important for me to avoid that, but I suspect I don't always succeed.
I mention it because I try to remember to cut people some slack when they act so... naturally. 😉
Presumably the feeling of "against me" due to kind of religious like identification with an idea provokes the suprising reactions.
It should be just a competition of facts and arguments (and maybe theories overall) .
So, similar approaches.....
Best avoid those that offend your ear 😉Even all recordings sounding "nice" to oneself is impossible.
Best avoid those that offend your ear 😉
Correct 😉 Better to have less impressive things and none that offend the ears than many impressive things and ONE that offend the ears. BUT... as we move up the ladder, nothing really seriously 'offend' our ears, so it is still a hard decision. May be having two amps, each for different purpose, is a possible option?
Sorry Pavel, I was going to type 'Gedlee Metric' then changed to 'Earl Geddes' and fingers and brain got confused
GedLee LLC is a good starting link for information on Earl's work.
Could be worse:
Geddy Lee | Rush.com
😛
Scott by your posts it seems that you listen only to punk/noise/industrial stuff ?.
I guess you missed the ones on Hindi devotional music, old timey blues, and you even commented on the post concerning David Lewiston's Asian field recordings.
Sorry for the obscure reference.
Number Six is shown into the metallic cylinder. He passes transparent tubes holding Numbers Two and Forty-eight along with a third, empty tube, each labelled as "Orbit". Climbing a stairway, he finds a robed man in a mask watching surveillance videos of Number Six.
Of course, I agree with you, Jakob.The goal for the reproduction of sound isn´t that well defined (or agreed by all parties) as PMA seems to state.
Too long to explain why some can consider the pure technically position, when it is not accompanied by a ton of nuances, like an extreme simplification.
And the excess of confidence in the deep knowledge of the science, so poor in the same time, both presumptuous and naive.
It is obvious that we feel the same on this subject.;-)
No surprise that people that have a simplified approach are so definitive in their statements, feel themselves in danger when others question their certainties and react the way they do.
If I learned a little electronics, it was to try to better master the tools for recording and reproducing music. I do not always have the impression that some EEs, who decided to apply their knowledge in the field of audio, have taken the effort to go (and learn) the other way.
Because our lives are short, we cannot know and do everything. Why challenge to others, who are fishing in other waters, the weight of the fish they bring back?
Special dedicace to Scott, with all my admiration, that has a deep knowledge of lot of aspects of electronic and the physic involved, probably more than everybody else in this forum. And never spread it.
I would like to take this opportunity to thank him.
I asked him questions, in private. He took the time to answer the Padawan I am with a few nice high density words that immediately made me change the observation post in my way of analyzing and "feel" the circuits.
Gotcha. I must remember to put the cat out tonight.I mention it because I try to remember to cut people some slack when they act so... naturally. 😉
- Status
- Not open for further replies.
- Home
- Member Areas
- The Lounge
- John Curl's Blowtorch preamplifier part III