Here we go again...Not knowing what you don't know vs Measurements. a pseudo-sarcastic rant.

[johnp98]

I have also been thinking about this topic a lot recently and here are some musings that I have come up with:

Double blind wine tasting? Coke vs Pepsi and functional MRIs? Very interesting topics to read about that highlight the complexities of human perceptions and all the factors that influence it.

I have read a lot of behavioral economics and psychology books (probably the best being any of the books by Dan Ariely or Daniel Kahneman) and think the audiophile community is a prime example of various psychological quirks such as (I am sure I will get some of the exact terminology wrong):

Bandwagon Effect: A person is more likely to go along with a belief if there are many others who hold that belief.
Related to audio – Review for this speaker are amazing thus it must be amazing.

Choice-supportive Bias: Once a decision is made, people tend to over-focus on its benefits and minimize its flaws.
Related to audio – I have spent my life loving tube amps and vinyl, thus it must be best.

Confirmation Bias: Paying more attention to information that reinforces previously held beliefs and ignoring evidence to the contrary.
Related to audio - I only agree with those who believe in listening with their ears (or those who only use measurements)

Ostrich Effect: Avoiding bad news about a decision by ignoring data that might be negative.
Related to audio - I paid an arm and a leg for this amp, and now you say it measures similar to a much cheaper amp!?!

Belief bias: An effect where someone's evaluation of the logical strength of an argument is biased by the believability of the conclusion
Related to audio - Wire can pick up RF interference, thus I probably should shield from it.

Experimenter's or expectation bias: The tendency for experimenters to believe, certify, and publish data that agree with their expectations for the outcome of an experiment, and to disbelieve, discard, or downgrade the corresponding weightings for data that appear to conflict with those expectations
Related to audio - Absorb the first reflections!

Dunning–Kruger effect: The tendency for unskilled individuals to overestimate their own ability and the tendency for experts to underestimate their own ability.
Related to audio - I can hear a difference between these two components as I have been an audiophile for years!

Ikea effect: The tendency for people to place a disproportionately high value on objects that they partially assembled themselves, such as furniture from IKEA, regardless of the quality of the end product
Related to audio - I worked hard setting all this stuff up and doing this new tweak, and boy what a difference!

Etc…

There are many psychological quirks that are very applicable to everyday life and specifically audio that I think we all could learn from.

It is easy to see how many people fall into these traps when it comes to audio. We often think that we personally do not fall into any of these biases, but we do. Thus I think we need to learn to live with them. We can use these biases and quirks to our advantage such as: If you want to impress your friends (or even prime yourself) then maybe larger and more pretty speakers might truly be worth it. If you tinker and change the position of the speakers, you might get the ikea effect and it might sound better. If the album your listening to is well reviewed and has a great soundstage, you will be primed to hear that better. Etc.

I think many people need more respect for the scientific side of audio. We need scientific reviews/analysis/comparisons. We need double blind tests. We need to challenge all beliefs. We can quantify and find associations that hold scientific validity.

I think many people also need more respect for human side of audio. Large speakers truly look impressive. Cosmetics are important. Finding your preference is important. Buying new things is fun.

The joy of being an audiophile is walking this line and supporting both sides!

 

[AJ Soundfield]

I've tried it both ways, flat and with the curve, and I prefer the curve.

Yet neither can be right for all recordings, which can vary quite a bit spectrally between thin and bass heavy. Precisely why Toole et al advocate for "tone controls" that can be used on a per recording basis.
http://seanolive.blogspot.com/2009/10/audios-circle-of-confusion.html

Btw, when I see bass that flat (aka happy mic/eyes effect), I fear there were frequencies boosted that should not have been.

cheers

 

[AJ Soundfield]

Just wondering... How does Toole or anyone scientifically account for this variation...

The tests are blind (not "blindfold"). Turns out (consumer) "ears" within standard audiometric variations (no severe loss etc) overwhelmingly prefer the same things, despite beliefs to the opposite.
Now when audiophile eyes/beliefs etc swamp the experience:
http://seanolive.blogspot.com/2009/04/
https://www.pnas.org/content/105/3/1050
https://www.aes.org/e-lib/browse.cfm?elib=19405

cheers

 

[welldun]

Yet neither can be right for all recordings, which can vary quite a bit spectrally between thin and bass heavy. Precisely why Toole et al advocate for "tone controls" that can be used on a per recording basis.
http://seanolive.blogspot.com/2009/10/audios-circle-of-confusion.html
View attachment 40126

Btw, when I see bass that flat (aka happy mic/eyes effect), I fear there were frequencies boosted that should not have been.

cheers

I added the raw measured response of the 4 subs to the graph. Now you can see before and after. See attached.

 

[highstream]

It’s not just ears. Discerning and appreciating sound is an interactive process between ears and brain. We’re not only all “wired” differently genetically, our psycho-neural developmental paths also differ significantly. Measurements, however informative, still skim the surface of that process and our understanding of it.

@huskydriver I hope it’s not out of place, or opening a can of worms in this thread to ask which PS Audio products you consider scientifically suspect. You can tell from my comments that my appreciation of McGowan is quite mixed, yet I know of only one product they sell that might be questionable from a user standpoint and perhaps a scientific one, that is, the Noise Harvester AC Cleaner. I don’t know enough to judge the science, but as a user I found no difference with it in different locations. That’s just a sample of one, though. Beyond that, like them or not, their products represent versions of the state of audio engineering and sound science, which as McGowan among many others in their business readily recognize and emphasize, is hardly mature, academics’ claims notwithstanding.

 

[Mike-48]

@welldun - I sympathize with you and go through similar frustrations. Indeed, they cause great doubt. And I am not a beginner -- I've been playing with EQ for about 17 years now, with two automated systems and many manual ones, though I have no formal education in it.

I agree with @Sonnie and think it's an important point, measurement at < 500 Hz is not too tricky. The mics are not directional, sound is not very directional, and almost everyone likes smooth bass. BUT of course, you have to decide which dips are correctable (the shallower, wider ones) and which are not (the deep narrow ones). Then you must decide how much to boost the bass below, say 200 Hz, a boost usually needed for good tonal balance at home listening levels.

In the higher frequencies, one should avoid correcting dips caused by crossover diffraction, I'm told. Have fun figuring out where they are and, if you have an automatic system, how to get it not to correct them.

I consider 500 Hz to 2000 Hz a kind of transition zone. Above 2000 Hz, IME, all bets are off. Mics get more directional, and so do speakers. The precise mic and speaker position may influence measurements more than you expect. I've measured with two mics, both used the way they were intended, and gotten different results. One pointed straight up, the other between the speakers. Neither was supplied with a calibration curve for the other direction.

Then, what should you EQ? Time-windowed on-axis sound? Or pink noise with an RTA, which is averaged over time and includes more room sound? Most auto-EQ systems don't tell you what they do. I'm not sure which is more theoretically correct. I don't know that there's a definitive answer to that. I will suggest using 1/6-octave smoothing in graphs meant for EQ work, as the ear doesn't hear every little blip.

So one method is to stop around 500 Hz (I go to 900). Another is continuing up, but in the higher ranges, using measurements to help spot potential issues, and with our ears, figuring out whether they sound better EQed. Sometimes response may look flat but sound better with a -2 dB dip around 2000 or 3000 Hz. I find that particularly true of classical music with a lot of violins, whether solo, in ensembles, or massed.

I can sum up my exploding head in two main points. First (bold paragraph above), it's pretty easy and non-controversial to EQ lower frequencies, and you gain a lot by doing so. Second, in higher frequencies, measured results, though precise and quantitative, are less clear about what we hear. The latter is influenced also by directionality, distortion, the relative quantities of direct and reflected sound, and I'm sure other factors. So take measurement with a grain of salt, and do the best you can. If you have an auto-EQ system, hope for the best, and don't be afraid to experiment. Some of them can sound pretty good, even if we don't know exactly what they are doing.

All of this is less important with home theater. It's pretty clear that when actively using our eyes, we are far less sensitive to small changes in tonal balance. One still wants to try, but getting it exactly right is a lot less critical.

So now it seems like we need both a good set of trained ears as well as an expert at interpreting graphs... the problem just doubled!

in the end, you must like what you hear!

+1 to both!