Best guess is no.
Many of the engineers working on today's music are in their 20's. Do they listen to music made in the 70's? What is familiar to them?
If familiarity drives the preference, not a better subjective sound, I would expect the amount of distortion added would be reduced with each generation of studio engineers....maybe. Or not
Wouldn't someone who is familiar with the real sound of instruments be more comfortable with that sound without distortion added than someone who wasn't?
When changing the levels and bands of 2nd order harmonics in the track the sound difference to me is akin to the difference between bands in different venues.
This makes sense, as changing the environment does change the even order harmonics you will hear. It also changes the band affected. What adding 2nd harmonics doesn't do is add the time delay that the walls of a given venue do. This can be added separately though if you wished.
This is probably what those old preset effects did on receivers and amps of the 90s. You know the ones with 'hall' settings etc.
I suppose I'm saying that a high 2nd order distortion class A valve amp, or higher 2nd order distortion drivers may just be an analogue version of this as far as the ears are concerned.
I think the psychological difference between the 2 may just be the bias that the expensive and beautiful valve amps and exotic high efficiency speakers are functionally and visually different, and this is confirmed by the experience that they sound different. Whereas a cheap Kenwood receiver from 1992 is cheap and looks crap.
Who knows.
I will say though that human bias is ridiculously more powerful than people think. In one experiment, a group of top wine experts were given a red wine to taste. They were told the price (expensive) and asked to evaluate it. They described it with the usual descriptions for generic red wines. None of them had a clue it was actually a white wine with colouring added.
Coke tastes 17% sweeter in a red container, and sour from a yellow one.
The odds of management choosing the best candidate for a job at interview on average is 52℅. That's 2% better than a coin toss, yet your odds of getting a job increase by 10℅+ over other candidates if the people doing the interview are drinking a hot drink while they interview you.
(Psychometrics is a hobby)
The list goes on.
Fascinating !
Thanks for posting this is very useful when trying to set uo blind testing and DIY comparisons.
Cheers
A.
Thanks for posting this is very useful when trying to set uo blind testing and DIY comparisons.
Cheers
A.
Interesting , and scary video.
To my ears the sound with the added distortion did fit his descriptions. However , although the sound may have been "warmer" etc etc IT did not sound as "convincing" to me. It did not sound as "real". And also the added dist. sounded like a driver that is struggling in some way, like its has added noise from some form of lack of control, fuzzy, blurry.
Sorry for the unscientific explanation but thats what comes to mind when I hear the distortion manipulated sound.
I would much rather listen to "warmer" by manipulating frequency response
To my ears the sound with the added distortion did fit his descriptions. However , although the sound may have been "warmer" etc etc IT did not sound as "convincing" to me. It did not sound as "real". And also the added dist. sounded like a driver that is struggling in some way, like its has added noise from some form of lack of control, fuzzy, blurry.
Sorry for the unscientific explanation but thats what comes to mind when I hear the distortion manipulated sound.
I would much rather listen to "warmer" by manipulating frequency response
I can vouch for 15% but not 17%🙂
There is a logical fallacy that always comes up in these distortion discussions. The fact that our auditory sense can be tricked or influenced by bias (well documented) does NOT mean we aren't sometimes capable of incredibly accurate perception.
The problem is that the ear is very capable, but we don't normally
just hear, rather we interpret and listen. This is why we can tell what someone with a heavy accent is saying, because we are interpreting the sound.
Anyone who has ever built a real speaker knows we can hear very slight differences especially with experience and training. But you really need measurements to reliably know whats going on. Further you need knowledge to diagnose and fix problems.
I'd say that the person who wants to believe their 20$ amazon specials sound like world class speakers is just as deluded as the person with the 1000$ speaker wires on their tube amp.
Not saying that there isn't room for good research, but proving that people can be tricked or aren't very perceptive in general proves nothing.
Quite true, but the problem is that it is hard to know who is and who is not "capable" or when and for how long. For example, I distrusted our "expert" panel of evaluators, so I had a Guage Capability Study on them as a group (10) (I did not do the study and so it was blind to me. I was not a participant.)
As a group they were not capable of a statistically significant evaluation. But there were 2 out of ten who actually were. When asked, everyone thought it was them. We never released who was and who wasn't, but we did downplay the "expert" evaluations quite a bit after that. (I found out who "was" and I could see why - they were the two most "committed" and the most objective. Always looking for the problems and not taking any evaluation lightly.)
From my experience the more sure someone is that they are "capable", the less likely they are. It makes it very hard to accept casual listening evaluations as "good" evidence. Around here, I usually don't know the people making the claims personally so how can I take their opinions seriously. Now if you have objective data that you can show me, or blind testing data, I'll take that seriously, but no single and/or un-blind evaluations are going to sway my position.
That is a good video, but for me Daniel Kannemans book "Thinking Fast and thinking slow" ios the best reference. Kanneman won the Nobel Prize for doing the basic research on almost all of what was shown in that video.
Very interesting, I like casual listening approach, but I agrees it does seem the more confidence people have in themselves, the less likely usable results. This is probably more due to ego. But I do see if I can fine measurable data which may correspond with the casual listening, this has proved quite valuable to me.
No!
Self confidence may be delusional, but I do not buy in to the argument that this is the prevalent case. If you have the expertize, then the confidence exhibited flows from it. WHG
Self confidence may be delusional, but I do not buy in to the argument that this is the prevalent case. If you have the expertize, then the confidence exhibited flows from it. WHG
It may not prove anything, but it does suggest good research may be hard.
But it's not really that hard - we (well, some people) know how to do it - what makes good research 'hard' is that it's time-consuming, expensive, and fiddly. It's a lot easier to do half-*** studies, or give results of casual listening.
For fun, here's another "tricks" video, this one directly about audio:
YouTube
in examining what in can find in terms of reasonably detailed loudspeaker measurement it seems that the flatter the frequency response is the less distortion it contains. is this correct or just an improper correlation?
This answer depends on weather you view yourself as biased or not though.
The problem is, bias isn't overcome with experience in a field. What typically happens is bias such as 'expert bias' where the operant believes the expert anecdotal comments without evidence becomes replaced by conformation bias as you become the expert.
To be justifiably confident in an opinion requires continued attempts at falsification (usually by an independent body to avoid conformation bias) as opposed to just experience.
Being an expert carries increased probability of holding confidence in incorrect information, NOT decreased.
There is only one Warren Buffet. Not because what he does is unbelievably intelligent, but because intelligence INCREASES the chance of cognitive bias.
As he says:
'I'd rather believe the guy with an IQ of 120 who is 70% confident he is right than the guy with an IQ of 200 who is certain he's right......the second guy'll kill ya!'
The anecdotally best thing I've learned about cognitive bias is probably the simplest... People live their lives as a character in a story. Not by facts. The closest thing you can come to the truth is to decide the 'character' you play is a guy who doesn't like patterns or stories that end.
Part of the reason for this is that almost everything we do in our modern day is part of a series of learned behaviours that follow established patterns rather than Independent thought. This is what makes a new job so stressful, and why having your own table plant that moves job with you or bringing your pet to work is so effective at reducing stress. Its something that stays the same in a world of change.
You can clearly see the level of distress greatly reduce in a PET scan when a social group meets up and the topic changes from a new unexplored subject to a familiar scripted one. You can also tell if someone is going to accept evidence that contradicts their previous convictions without asking them because their brain lights up like a Christmas tree if they do.
Last edited:
The person with most to lose is less likely to admit he is wrong. This seems to be getting more common in society at large where people in positions of power and responsibility are freer now than ever before to deny their mistakes because no one can be sure what to believe
One problem that I have encountered numerous times regarding bias in the audio world is the DIY situation. A guy builds a set of speakers he is sure will sound great because he made them and/or designed them. On listening his confirmation bias (the Beranek effect) prevents him from hearing all the flaws, just the positives. He has now firmly established an anchor to what he considers good "sound quality". All future evaluations, particularly other peoples speakers, will be heard through this complex filter of biases preventing the person from developing an appreciation for anything new. It's another form of Toole's "circle of confusion".
I found that the only way to break this circle is to constantly go back to objective measurements, which tend to have far less bias in them, although they can also "confirm" false expectations. "If something doesn't measure better then it is not better" breaks the circle.
I got to the point where I could no longer subjectively hear differences that I could measure (much more careful subjective evaluations may be able to, but my personal casual ones were not.) Hence, I stopped developing the designs. (Then I stopped making them!)
I found that the only way to break this circle is to constantly go back to objective measurements, which tend to have far less bias in them, although they can also "confirm" false expectations. "If something doesn't measure better then it is not better" breaks the circle.
I got to the point where I could no longer subjectively hear differences that I could measure (much more careful subjective evaluations may be able to, but my personal casual ones were not.) Hence, I stopped developing the designs. (Then I stopped making them!)
It's wise to know when to stop.😉 I'm not a compulsive measurer, it seems some people are. Surely microphones at a point in space and with the technology available are a useful tool, I think many DIYers overestimate their power though, as with most technology they may only marginally understand.
2 true references.
Solid post Earl, I agree with most of what you said.
I would like to suggest that there is only 1 true reference and then 2 separate options when it comes to designing loudspeakers:
(1) Listening to live (no PA or electronics) vocals and instruments is the only true reference point.
(2) Objective test and measure techniques for design.
(3) Listening tests to assess quality of reproduction of loudspeakers.
Are (2) and (3) equally important?
Are they equally important at all times / stages of development / to all designers / customers...?
My biggest issue with the objective only option is ... How do we know what to measure and what to ignore....Which measurements are the best indicators of the best sound?
Is this set in stone for ever ie 100 years ago the early driver / loudspeaker designers started the ball rolling.... Material science, computer analysis, electronics...All unrecognisable today....
Yet all the 21st century design houses use the teachings / principals of the 50 to 100 year old pioneers....
Only in loudspeaker design do you get multi $Billion corporations led by / consulting with 60 to 80 year old guys who still think the Loudspeaker design cook book is the bible!
Solid post Earl, I agree with most of what you said.
I would like to suggest that there is only 1 true reference and then 2 separate options when it comes to designing loudspeakers:
(1) Listening to live (no PA or electronics) vocals and instruments is the only true reference point.
(2) Objective test and measure techniques for design.
(3) Listening tests to assess quality of reproduction of loudspeakers.
Are (2) and (3) equally important?
Are they equally important at all times / stages of development / to all designers / customers...?
My biggest issue with the objective only option is ... How do we know what to measure and what to ignore....Which measurements are the best indicators of the best sound?
Is this set in stone for ever ie 100 years ago the early driver / loudspeaker designers started the ball rolling.... Material science, computer analysis, electronics...All unrecognisable today....
Yet all the 21st century design houses use the teachings / principals of the 50 to 100 year old pioneers....
Only in loudspeaker design do you get multi $Billion corporations led by / consulting with 60 to 80 year old guys who still think the Loudspeaker design cook book is the bible!