• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

Our perception of audio

Fitzcaraldo215

Major Contributor
Joined
Mar 4, 2016
Messages
1,440
Likes
632
I don't quite know what you're getting at there. When you stress understanding, my immediate thought is that anyone who truly understands statistics would probably distrust the use of statistics in anything but pure mathematical puzzles. As soon as you try to apply them to complex, interesting real life you are making assumptions that are really just guesswork. These give you the financial crashes, and self-driving car crashes, that the statistics don't predict.

https://en.wikipedia.org/wiki/Statistical_assumption

To me, the highlighted text is just a fancy way of saying that you should think hard about your hand waving before attempting to justify the statistical approach you have taken. What follows may be extremely rigorous and mathematical but ultimately founded upon hand waving.
Uh, oh. I had better stop taking all the drugs I am taking, since "proof" of their probable efficacy is only based on statistics from sampling groups of test subjects, not entire populations.

I have known a few people who truly understand statistics, some distinguished consultants in their fields. It did not strike me that they distrusted the proper use of statistics at all. Actually, the reverse was true. But, like anything, statistics can be misunderstood, misused or incorrectly applied, but let's not tar the entire field because of that.

Also, statistics can have at least two useful applications. One is purely descriptive of tendencies of a data sample. The other is predictive. I grant you, the latter does not avoid the possibility of a crash, as in trying to predict the behavior of complex financial markets in time future based on behavior in time passed. Stuff happens in human activity based on external factors not considered in the statistical model. But, I still think it beats just flipping a coin to try to predict the future.
 

svart-hvitt

Major Contributor
Joined
Aug 31, 2017
Messages
2,375
Likes
1,253
Uh, oh. I had better stop taking all the drugs I am taking, since "proof" of their probable efficacy is only based on statistics from sampling groups of test subjects, not entire populations.

I have known a few people who truly understand statistics, some distinguished consultants in their fields. It did not strike me that they distrusted the proper use of statistics at all. Actually, the reverse was true. But, like anything, statistics can be misunderstood, misused or incorrectly applied, but let's not tar the entire field because of that.

Also, statistics can have at least two useful applications. One is purely descriptive of tendencies of a data sample. The other is predictive. I grant you, the latter does not avoid the possibility of a crash, as in trying to predict the behavior of complex financial markets in time future based on behavior in time passed. Stuff happens in human activity based on external factors not considered in the statistical model. But, I still think it beats just flipping a coin to try to predict the future.

Regarding your mentioning of «people who truly understand statistics».

When this came out a couple of years ago I thought it was about time:

http://www.amstat.org/asa/files/pdfs/P-ValueStatement.pdf

Link to the whole paper:

https://www.tandfonline.com/doi/pdf/10.1080/00031305.2016.1154108

It isn’t often you see a professional organization warn the public against its own profession. It takes some integrity to do that!
 

Blumlein 88

Grand Contributor
Forum Donor
Joined
Feb 23, 2016
Messages
20,524
Likes
37,057
Regarding your mentioning of «people who truly understand statistics».

When this came out a couple of years ago I thought it was about time:

http://www.amstat.org/asa/files/pdfs/P-ValueStatement.pdf

Link to the whole paper:

https://www.tandfonline.com/doi/pdf/10.1080/00031305.2016.1154108

It isn’t often you see a professional organization warn the public against its own profession. It takes some integrity to do that!

The second paper reminds me of something an economist said. The extra classes and years to get a phd in economics are necessary to get an economist to believe and pay attention to details he was taught in Econ 101.

The statements in the paper seem to fit with what a very first class in statistics would teach students. You are seeing if your data match a model. By itself the p value proves nothing nor disproves anything. Nor does it say how large or important an effect is should it match your data.

I do think it is worthwhile to tighten up on the p=.05 value simply from the result of practical experience. As was recently recommended that research go to a p=.005 guideline. The p value was always arbitrary. Subsequent experience in quality control of manufacturing and experimentation in physical sciences indicated you made more progress and cut way down on chasing ghosts if you went with 2.5 or 3 sigma as your threshold. Those circumstances are less complex and it is simpler to know your model fits with what you are doing. In the messier areas of psychology, sociology, medicine I would think the need for tighter thresholds is even greater. In all cases you have to remember the p value is a tool and not a validator.
 

Brad

Active Member
Joined
Nov 8, 2016
Messages
114
Likes
35
The p-test has caused manu problems in medical and psychology research - like 60% of studies not being reproducible. The main problem (apart from p-hacking) is that from a small random sample of a population you can get any p value. The expectation value gives much more insight into your population.

The other problem with interpreting statistics for real events, is that the events follow power-law distributions and not Normal distributions. Most of our intuition is for Normal distributions. An example is earthquakes. If it was a Normal distribution we would almost never see magnitude 9 earthquakes. With the correct power law, we find that they occur with disturbing frequency
 

Nowhk

Member
Joined
Dec 22, 2017
Messages
99
Likes
15
Even in recording people are coloring sound.
I know. They know. In fact they are working on making the "final" product. But if also a listener, in the end of the chain, adds its own color (by his gears), isn't wrong to say that he's shaping its own "reality" about that song as well? They are mixing-on for his own reality. So the whole idea about transparency should be wrong, from scratch. No? Please note that mine aren't provocations, I'm just trying to understand some stuff :)

Not in this context. What I am describing there is NOT that your system became better. But that your brain hears more detail that may have always been there so nothing is improved with respect to the new gear.
I see, and I agree. Many gears can be able to "show" details that some others can't (due to my attention, expectations and experience).
And the idea (I think) is to reach a chain of gears to a level that changing somethings won't do any differences (as @Cosmik said, change a cable or a DAC won't take effects). So lets talk about the second part of your reply (which is what I was talking about).

Now if you are talking about speakers there, then sure, there is coloration there by default.
So again, once you reach a point were you would catch some details, they will change as well across different listenings, time to time, state by state (again, as @Cosmik said, drink a coffee in the morning will affect the whole).

But this confirm my puzzlement: why looking for somethings "transparent" when an objective reality isn't reachable at all? (neither between artist and listener OR different listening for a chosen listener, like me).

Why struggle to get a transparent setup/environment and try to get a deterministic listening when stuff in the life (music in this case) aren't deterministic at all, but fluid all of the time depending of uncontrollable things? (such as setup coloring, as related example).
Why would you consider a Bose color not acceptable and a 10k setup color (which THERE IS, even if tiny) acceptable? In both case you won't get the artist reality behind the song, you will distort the piece as well and just add your own preference "layer". No? :p
 

Dirk Wright

Member
Joined
Apr 28, 2018
Messages
23
Likes
13
---While this is responsible for some of the faulty observations, just as big of a factor is the elasticity of our perception of audio. When evaluating new additions to our systems, we become far more attentive. We are dying to know if the new addition made a difference. We pay far more attention to what is played and as a result, hear detail, nuances, etc. that we did not when we were just enjoying music. What happens then is what you say: we attach those improvements to the new device and bias makes sure that when we go back to "before" configuration, we don't hear those improvements.

There is a paper on how our hearing adapts to new sonic signatures over time. Thus, when we make a change in our equipment, we notice the change (if any), but after a while the new sonic signature becomes the new normal. This means that long term listening tests are completely wrong and useless. I uploaded it to the FB group I created called "Audio Science and Engineering". It's an AES paper as I recall.

This is, of course, independent of expectation bias, etc.
 

Frank Dernie

Master Contributor
Forum Donor
Joined
Mar 24, 2016
Messages
6,445
Likes
15,781
Location
Oxfordshire
There is a paper on how our hearing adapts to new sonic signatures over time. Thus, when we make a change in our equipment, we notice the change (if any), but after a while the new sonic signature becomes the new normal. This means that long term listening tests are completely wrong and useless. I uploaded it to the FB group I created called "Audio Science and Engineering". It's an AES paper as I recall.

This is, of course, independent of expectation bias, etc.
I have long time attributed this characteristic to people thinking equipment "burns-in".
 

JJB70

Major Contributor
Forum Donor
Joined
Aug 17, 2018
Messages
2,905
Likes
6,148
Location
Singapore
I've always wondered how much of the alleged "burn in" effect of audio equipment is actual equipment running in and how much is adjusting to a new sound. I can accept that mechanical moving parts and diaphragms will have a period where they loosen up and bed in (but I'm not sure it should need long) but I have always suspected that when people say a piece of equipment needs time to burn in it actually means a period of time for a user to become accustomed to it. When I got my T5P headphones I really thought they were a bit bright and with a recessed mid range, after a couple of weeks they sounded much more balanced and natural, but I suspect that while there probably was some change in the sound a lot of it was me getting used to their sound. One of the nice things about headphones is that unlike speakers it is very easy to switch headphones depending on mood or what you're listening to, which means that unlike my speakers which I think have a neutral sound and very natural I do still notice differences between the sound of my different headphones.
 

Sal1950

Grand Contributor
The Chicago Crusher
Forum Donor
Joined
Mar 1, 2016
Messages
14,073
Likes
16,609
Location
Central Fl

solderdude

Grand Contributor
Joined
Jul 21, 2018
Messages
15,891
Likes
35,912
Location
The Neitherlands
When I got my T5P headphones I really thought they were a bit bright and with a recessed mid range, after a couple of weeks they sounded much more balanced and natural, but I suspect that while there probably was some change in the sound a lot of it was me getting used to their sound.

This ^ I call it 'brain-in' instead of burn-in.
 

solderdude

Grand Contributor
Joined
Jul 21, 2018
Messages
15,891
Likes
35,912
Location
The Neitherlands
Here was the last post I made on WBF.

(In case it disappears: "How do Power Cords perform all this good stuff when they (at least, the hot and neutral leads) spend most of their time isolated from the equipment circuitry via rectifier diodes in the non-conducting state?")

Aahhhh shocks ... we don't have a mains cable thread here it seems ... shocking .... at least when you touch both bare wires.. or just the live one :rolleyes:

I seem to be missing all the good stuff it appears to bring.
When running on batteries only... will it still sound better when I connect the batteries through such a cable ?
So many questions ...
 

NorthSky

Major Contributor
Joined
Feb 28, 2016
Messages
4,998
Likes
937
Location
Canada West Coast/Vancouver Island/Victoria area

solderdude

Grand Contributor
Joined
Jul 21, 2018
Messages
15,891
Likes
35,912
Location
The Neitherlands

Mr. Caelin Gabriel obviously has extensive knowledge about this subject. I believe, however, he is also good at mixing facts with fiction.
Lets analyze some of his quotes.


Mr.Caelin Gabriel: "ALL of the wire and connectors can and do affect the performance of the component's power supply."

Me: wires are in series with the components power supply (induction + resistance). And also sort of in parallel (capacitance) so yes, wire could affect the performance of the connected power supply in a very small amount.
A higher resistance will reduce peak powers drawn which is important as shown below here. A higher inductance will reduce the 'energy' of HF garbage.
One has to realize that last meter is in series with tens to hundreds of meters of 'crappy cable' with tens to hundreds times the inductance and resistance and within that length lots of other devices also 'distort' that voltage.
A very complex system of 'noise generators' and (short peak) currents drawn that are 'flattening' the peaks of the AC mains voltage.
Yeah ... the last meter will make the difference ... sure it will.

Unless one is using a regenerative UPS of course, one on each device. Where each one of those devices emits their own HF garbage and usually has lots of filters in there... that are bad ... and usually have trouble providing high peak currents. Ohh the waspnest this is.

All true... but I have yet to see measurements that clearly show a real world situation where the DC output of a power supply is really 'better' (cleaner or has less ripple referenced to the circuit ground/common) when 2 different (decent quality) mains cables are used.
Any audio device works on DC so when there is no significant variance in the DC.
Why not show pictures with 'proof' (that aren't doctored in some way, which is so easy to do.. the doctoring part) ?


Mr.Caelin Gabriel: "There are two aspects to power transmission: the electromagnetic wave and the current flow. The current itself cannot be contaminated but the electromagnetic wave can be modulated with other frequencies. We usually call these other frequencies noise or Electromagnetic Interference (EMI). Within the various parts of a power circuit there may be EMI in certain parts that is not present in others. Electromagnetic energy can be transformed or redirected to lessen their effects.

me: True. What's it have to do with the DC output from a powersupply ? In the sense that 1 meter of cable, that does not appear to have any different inductance, resistance or capacitance, can change the DC output so dramatically that it can actually change the waveform of an audiocircuit fed with that DC.


Mr.Caelin Gabriel: "Some power cords use capacitors, inductors, or ferrites in an attempt to control the electromagnetic fields around the audio component. The success of such an approach is completely dependent upon the specific design and the reactance of the power supply of the component to which the power cable is attached."

me: True.
This means that some 'filtered' cords improve AC 'quality' (but not necesarily the resulting DC) while the same cord can make some AC components (i.e. RF garbage) worse.
So will almost all AC filters in any device. As a lot of devices using SMPS (Switched mode Power Supplies) have an AC filter to comply to minimum requirements.
Why would one buy such a cable... Shunyata cables don't appear to have a filter in it nor do 99.99% of all power cables so why mention these cables?


Mr.Caelin Gabriel: "The power cord is not the last 6 feet, it is the first 6 feet from the perspective of the component. As stated in #1 the local current and electromagnetic effects directly affect the sonic performance of the component."

me: For emitted signals the power cord is the first 6 feet. For the received power (what a device works on) it is the last 6 feet in series with the hundreds of feet before it.
Mr.Caelin Gabriel stated in #1 response: "ALL of the wire and connectors can and do affect the performance of the component's power supply." he certainly did not say: "the sonic performance of the component" but said "power supply" which are 2 very different things.
One is a fact "power supply" the other an opinion "sonic performance".


Mr.Caelin Gabriel: "A primary source of audible sonic degradation is caused by the power supplies in our audio/video components. Most components use FWBR (full wave bridge rectifier) power supplies that generate an incredible amount of transient noise when the rectifiers switch off. The design of a power cable can significantly affect the reactance of these signals within the power supply. The power cable is effectively part of the primary winding of the power transformer. "

me: Well any rectifier (FWBR or SWR) will indeed generate very small amounts of RF peaks. How 'bad' this RF type of signals is depends on whether the power supply has sufficient 'snubbering' (I love that word) and if there is a filter used or not.
Those switching artifacts are only present on the input side of the rectifier and not on the DC side. A poor design can indeed 'conduct' some of these (very small amplitude) RF signals in so called 'common mode' in the DC circuit.
Nothing any cable can improve performance on.

Common mode signals may be a problem when connected to other devices not by themselves as all DC power lines and 'common' are all modulated in the same way effectively having no influence between the DC rails at all so are not 'modulating' the actual music signal.


Mr.Caelin Gabriel: "Most power supply filters are ineffective at blocking very high frequency noise components and much of it is passed through to the DC rails. The sonic effects of this include: high background noise levels, blurred or slurred transients and a general lack of clarity and purity of the sound or visual image."

me: yes, most power supply filters are ineffective at GHz noise components.
The 'sonic effects' part is totally made up. No proof has ever been presented that transients in music signals are affected nor that clarity (in the 3-5kHz range) is affected nor that 'purity' is affected.


Mr.Caelin Gabriel: "As long as power supply design is based upon FWBRs or switching supplies, the power cord will always be significant.”

me: Indeed without a power cord the device will not be powered and do nothing. In that sense it is true. Very significant.


Mr.Caelin Gabriel: "A power cable on the other hand is not transmitting a signal. It is conducting A.C. power and its sonic superiority will be determined by its ability to deliver current (steady-state and instantaneous) and its ability to deal with the EMI effects of the components to which it is attached."

me: Well an AC power cord is feeding the mains voltage to the device. Through it there are short (and high peak current) very short 'pulses' with lots of harmonics that are only present at the peaks of a sinewave. The rest of the time nothing happens in there. There is no 'audio' drawn through the cables at all.
Just short peaks.

Below the harmonics (current, not voltage) that are 'injected' into the mains using a FWBR.

fft-single-phase-single-winding1.png


Below a scopeshot I made of the currents running in the diode (so on the secondary side, the current in the mains side of course is much lower depending on the winding ratio, it is part of this article

3300uf-schottky-33ohm-big-trafo.png


Mr.Caelin Gabriel: "Silver, copper, brass, gold and others all "sound" different."

me: no they just have a different amount of resistance when diameter is the same, and a very strong effect on ones expectation bias by association.
I have yet to see any cable material acting as a bandwidth filter in and well outside of the audible band or add audible amounts of distortion.
I have seen poor connectors do that though, cable materials .. no and that is with audio cables. So.. an audiophile assumption.
How would any of these materials change the pulses present in a power cable and have an audible effect when no audio is passing through these materials ?
Would like to see some technical evidence. I haven't been able to find any.

Mr.Caelin Gabriel: ""Many components use a power inlet IEC that has an integrated "L" or "pi" filter. The quality of these devices varies dramatically."

me: Yes, very true. What does it have to do with audio quality ? It just is a loose remark.


Mr.Caelin Gabriel: "Many of these IEC packages were created for office and computer products and are required to pass certification tests for EMI emissions."

me: Yes, very true. What does it have to do with audio quality ?

Mr.Caelin Gabriel: "Shielding can be a two-edged sword. On one hand, it can reduce radiated fields from impacting other components. On the other hand, the shielding may induce re-radiated fields onto the cable or component that it is being used in. Sometimes the cure may be worse than the illness."

me: Yes, very true. What does it have to do with audio quality unless the power cords run parallel with (poorly or not shielded) audio cables.

Mr.Caelin Gabriel: "There is no silver bullet and there is no rote formula that works in all cases."

me: Hmm ... I thought Shunyata cables did ?
 
Last edited:

Wombat

Master Contributor
Joined
Nov 5, 2017
Messages
6,722
Likes
6,459
Location
Australia
It seems to me that Mr Gabriel hasn't got much further than the 'water flowing in a hose' level of understanding electricity.
 

NorthSky

Major Contributor
Joined
Feb 28, 2016
Messages
4,998
Likes
937
Location
Canada West Coast/Vancouver Island/Victoria area
Thank you Mr. Solder for the time you took with your analysis.
And if you don't mind I would like to share with Mr. Gabriel.
Can I quote your post in its entirety and link it to here?

Cheers,
Bob A. C.
 

solderdude

Grand Contributor
Joined
Jul 21, 2018
Messages
15,891
Likes
35,912
Location
The Neitherlands
You could do that, but my answers he has heard hundreds of times already.
He probably is not going to reply or will blow more smoke your/our way and simply remark that we are 'wrong'.

It looks like Mr. Gabriel is well aware of proper EMC practices and knows what signals pass through a cable and can be found on the mains power grid.
He also seems to be well aware of mains filters and how they are constructed IRL and its actual properties.

I expect him to have access to tons of measurement gear or can do tests in EMC labs. I have done lots of tests in EMC labs myself as well.

Mains signals are not anything like the square-waves he likes to test with as these have high harmonics that do not pass through mains transformers anyway as these are bandwidth limited usually below 10kHz. RF passes through transformers by capacitive coupling between windings.

The guy has cables to sell and seems to do well at it.
Money should roll, especially from those who can afford it and don't mind spending a small fortune on audio jewelry like this with questionable motives.
It just won't be my money in this case and hope others find comfort in skipping on expensive nonsense either.
 

SIY

Grand Contributor
Technical Expert
Joined
Apr 6, 2018
Messages
10,386
Likes
24,752
Location
Alfred, NY
"A talented con man, or a slick politician, does not waste his time trying to convince knowledgeable skeptics. His job is to keep the true believers believing. He is not going to convince the others anyway. " -Thomas Sowell
 

JJB70

Major Contributor
Forum Donor
Joined
Aug 17, 2018
Messages
2,905
Likes
6,148
Location
Singapore
If expensive power chords really do make a difference then why don't those proclaiming their benefits simply present measurements and results of double blind testing to add some substance to the claims? Ditto with interconnects.

I find it genuinely sad when companies that are very competent and capable of innovative design get in on the act. I understand why they do it as if people want to throw $$$$$$$$$'s on this stuff why would you not sell it but for some reason I just find it sad when companies that do know better jump on the bus. A few months ago I saw a shop tried to sell me a Meridian headphone amp/DAC (which I didn't buy). The unit itself was very expensive (especially when there is no shortage of excellent performing DACs and headphone amps from companies like Fiio, Topping, SMSL, JDS Labs etc at very affordable prices) but then I was assured that to take advantage of its potential I would have to buy the separate power supply which was almost as expensive again. The dealer realised there wasn't much point going on when I asked him why if Meridian were so great they were incapable of designing the onboard power conditioning to function well enough, as other manufacturers seem able to do?

When I hear people start talking about the sonic properties of different metals used in power chords, my immediate response is to laugh, my second response is to be irritated and my third response is to say, OK then, show us some evidence.

Similarly with cables. Cables do need to be an appropriate gauge for the application and I have no issue with paying a little more for a cable which is well made, with good quality connectors etc because I do think a good, tight connection with untarnished connectors is important and if you're going to pay for something you may as well buy quality. However the price uplift to get such cables is trivial and can't be compared to audiophile cables.
 
Top Bottom