Can You Hear the Difference?

"The study compared data from more than 12,000 different trials from 18 studies where participants were asked to discriminate between samples of music in different formats. The new study found that, indeed, listeners can tell the difference between low- and high-resolution audio formats. Moreover the effect is dramatically increased with trained test subjects who could distinguish between formats approximately 60% of the time."

Sixty percent doesn't seem that dramatic since 50% would be just guessing, unless the sample size of the trained subjects is very large.

I'll reverse the off-topic on that one. You have only to leaf through the reviews of a hi-fi magazine to realise the extent of the audiopseud industry. If you note the existence ofpeople who pay silly money for hi-fi mains leads you are not likely to find among them a meaningful musical discernment.

The neighbor of an old associate of mine works at a place that does testing for custom hearing aids and related work that got to telling stories about the number of audiophiles that have came through proclaiming to have superior hearing to everyone and wanted to prove it with a documented hearing test.

Interesting thing was by far the majority scored as average or below on the basic sensitivity and range testing and only one she could recall ever tested as being even slightly above average for their age ranges, 30 - 50 years, but still well below what most kids of age 10 or less, like her own, could score. (obviously a fault of their equipment and not the listeners actual ears of course.)

Personally, the last hearing test I had I did score above average for my age but still I came up short of 20 - 20K and topped out at just over 17K which is not bad for a, at the time 40+, guy who openly admits to years of abusing his hearing for both work and at play.

And I am obviously not an audiophile either being I used lamp cords for my main speaker wire and ethernet line to connect the rear surround speakers for years.

If I'd spent $1,000.00 apiece on a pair of speakers, I'd claim to have perfect hearing, too.

I can tell the difference between which set of speakers in my pickup are still factory and which are aftermarket four-way Pioneer units. (The front doors need new speakers.)

As for audio playback formats, I can pick out the back to back playback differences between a song up to the standard 96K sampling rate using good headphones and maybe half that in the pickup which has a Pioneer DEH-4400 head unit.

After that, I think it's just a waste of memory space and if anything some songs seem to lose something in sampling formats higher than those but I am no expert and it could just be conversion algorithm artifacts I am sensing.

Not that it matters being I listen to metal, hard rock and occasional techno/dubstep most of the time anyways.

My truck still has all factory sound system. I hear NPR just fine.

And vacuum tube amplifiers with tubes imported from Russia...

No!

Put someone(trained subject) in an acoustic booth and play two versions back to back, OK.

Joe sixpack, no way!

Separate the tests by two hours and it's a total crapshoot.

No!

Back in the early '60s, I think I probably could have. I still remember the first time I heard high quality stereo played on high quality headphones from a reel-to-reel tape recorder running at 15"/sec.

At my current age, I probably could not.

And, judging by what they choose to listen to, I suspect that most young 20-somethings couldn't either, today

Try this: Sound Generator Software

My hearing shuts down above 8 kHz, so I'm just guessing what any highs sound like.

Let's take MP3 @ 128kb/s as an example. That is a popular compression level used for delivery of the program material to the end user and is pretty much acceptable on a one-pass basis. So long as not too many encodings and decodings have happened beforehand you can 'get away' with it. (BTW, MP2 at 256kb/s is far less obtrusive to my ears.)

Going back a couple of decades, you might have argued that a good quality cassette played on a well set up machine sounded indistinguishable from the master tape to most users with domestic replay equipment. However, few would argue that the studio tape decks with Dolby A (as against B or C) were not needed for production purposes. Those people used to working with studio equipment and listening on good monitor speakers would instantly be able to detect the difference on even one pass.

Now we have the opportunity to bring that same 'original' studio quality into the home and a good number of people have the equipment and experience to appreciate it. Let's raise the bar and all be able to enjoy it instead of choosing the lowest common denominator.

That said, many people, especially youngsters listening to low-dynamic-range, band-limited rock music, actually are oblivious to the finer points and will even run their amplifiers into clip on a routine basis. Audio education needed here!

It is certainly true that some people have ehanced senses compared to the 'normal' population. (And in a normal population distribution, 16 percent of the people will be above the 1 standard deviation of 'average'.)

I used to work in an engineering 'cube farm' with about 100 8'x8' 'cubes' in a large room. About once a month I'd hear that maddening (to me) warbly sound when a landline phone is left off the hook. Sometimes I'd have to walk all the way over to the other side of the room to find the phone and hang it up, passing dozens of other engineers who said they didn't even hear it.

But I've got a friend whose hearing is even sharper. He's an EE and audiophile and used to build his own stereo system speakers because he could easily hear differences in sound quality, and was able to build high-end devices at a fraction of the retail cost.

Sounds bunk to me. There are differences, but just not noticeable to the average human range of hearing and the studies results were well within the range of chance guessing (50%).

Moreover the effect is dramatically increased with trained test subjects who could distinguish between formats approximately 60% of the time

Meaning the untrained average person was pretty much 50/50 guess or worse. Trained only 60%? The conclusion to the results sounds like people trying to justify their jobs and products, and that's how the article reads also.

Compressed data formats (definitely noticeable) and specialty music pieces played on hi end systems which would benefit (slightly), well that's a different matter.

What are "trained test subjects"? You either hear it or you don't.

It various. There's the average person, say people that have had some experience listening to the differences in music played on different formats and live so have some methods of comparison right up to professional audiophile "experts" who have taken courses on the subject and claim to be able to accurately tell the differences between similar speaker frequency responses.

difference between low and high definition audio

This article is very short on specifics. It states that there is no definition of High Definition audio. So, does it have increased frequency range, increased dynamic range, or what?

Increased frequency range (beyond the normal range of hearing) should have no effect on what the ear can hear. (I'm sure there are a lot of audiophiles that will argue with this!). Increased dynamic range might make a difference, depending on the dynamic range of the audio selection being tested. This is not specified either!

The bottom line is that it's hard to argue without any specifics!

Apparently a needle tracking a vinyl grove and moving the diameter of a hydrogen atom ( 1.2 × 10^-10 m ) can induce a detectable sound. For approximation purposes, call this 10^-10 m.

The total movement width of a needle is ~ 2 x 10^-4 m (0.2mm). Call this 10^-4 m.

This means, to a big approximation, 10^6 possible positions of the needle.

A 16-bit CD has 64k possible positions ( 6.4 x 10^4 ).

The difference in resolution is enormous.

A CD is very bright, clear and boring - it clearly lacks detail.

High resolution CDs are a different matter. Let's take a 24-bit CD, which has a resolution of 2^24. This is 16 x 10^6 which is higher than my crude approximation of a vinyl record resolution, and sounds it. ( see http://www.linnrecords.com/ - they have demos of the different resolutions. )

I'm old now (74), with a max hearing frequency of ~ 8kHz but I can still clearly hear the quality difference.

I have found that initially CD clarity is exciting but after 10 minutes it becomes boring. This is more evident with classical music.

Tests of 30 seconds of music I think are convenient for the testers but inadequate to allow a person to form a clear picture.

Apparently a needle tracking a vinyl grove and moving the diameter of a hydrogen atom ( 1.2 × 10^-10 m ) can induce a detectable sound.

Sorry, this sounds as if we are getting deep into audiopseud territory. Can you provide some evidence of this assertion?

After conversion to units I vaguely understand, 1.2 Å, I'd be inclined to agree you on this point. That's an indiscernible movement.

I think you need an atomic force microscope probe before you can hear those hydrogen atoms stuck in the grooves.

https://en.wikipedia.org/wiki/Atomic-force_microscopy

I call BS and then some.

Given then mass and rigidity of the needle VS the mass of a single hydrogen atom and its rather squishy plastic medium it's attached to plus the rigidity of the mechanism that converts the needle movement to electrical signals plus the lower limits of the electrical sensitivity of the amplification components there is no way in hell a record player can detect a single hydrogen atom physically and convert it into a sound wave from a speaker.

hydrogen audio vinyl record myths.

Now as for reality what it does take to detect a single hydrogen atom is certainly not physical contact in the sense of a record player needle but some fairly high-end precision physics done with some pretty specialized equipment like an electron microscope.

Detecting a single hydrogen atom takes an electron microscope.

To be fair, it's not about detecting hydrogen atoms themselves, it's about detecting a movement of the order of size of the diameter of a hydrogen atom. The moulding inaccuracies will be greater than that anyway.

That's my whole point. Theory Vs reality where reality shows that that theory is off by a factor of a 1000 or more from what the limits of reality play out at.

Theory Vs reality where reality shows that that theory is off by a factor of a 1000 or more from what the limits of the systems in reality play out at.

An audiophile uses a hydrogen atom as a physical reference and claims it's an audible standard where calculations can be made and referenced to whereas reality uses overall system dynamic range and compounded sensitivity, or lack thereof, as whole from start to finish as a minimum standard.

Simply put, If it can't be measured at the speakers in terms of actual sound vibrations produced it isn't a real valid influence.

It's worse than that. Double the distance from the source and the SPL* drops 6dB.

*Sound pressure level, is measured with a sound intensity probe at the source for radiated power or at the ear for audio engineers as an effect on our eardrums and hearing.

It's black magic, because reflected sound adds, or subtracts sound depending on the phase relations.

Further to the above, I have located a bit more than hearsay...

Working with order-of-magnitude type approximations:-

The dynamic range of a stylus is greater than 100dB ( see www.analogplanet.com/images/vinyl-dynamic-range.pdf ) This is 10^10 ratio between max and min levels of play (worst case)

If you take the max width of the modulation (i.e. needle movement) peak-to-peak to be ~0.1mm (10^-4m) (actually too big) then the smallest movement that the needle can audibly detect is 10^-4 / 10^10 = 10^-14 which is much smaller than the diameter of a hydrogen atom (~10^-10 m)

So my original assertion that a movement of the needle approximately the diameter of a hydrogen atom is detectable as a sound is, in fact, an underestimate of the sensitivity of the vinyl record needle.

(In "Vinyl dynamic range" he states: "The subtlety of detail in the grooves of
an LP record is astounding. Quiet, yet clearly audible, details (such as the reverberation characteristic of Carnegie Hall) are represented by stylus motions of less than an ultraviolet wavelength (1/100,000,000 of a meter)-a dimension approaching the size of a complex organic molecule." Note the "clearly audible details" which is very much larger than "smallest detectable sound" )

(And note that any test of smallest detectable sound is carried out in a good anechoic chamber with ears accustomed to true silence where you can hear your heartbeat and the rush of blood in you head. e.g. [http://www.acoustics.salford.ac.uk/facilities/?content=anechoic] with a sound level below -8dB. Lowest sound a human can detect is 0dB to -4dB)

I think the operative term in this is "trained". And if you've studied statistics you'll know that 60% is right around the mean for multiple choice guessing. In any case, I have to believe that speakers and the technology involved are getting better and better. For example, I can vouch for the argument that raged for decades about the difference between pipe and electronic organs. Even those that were "samplers" were deemed not as good due to the lack of sub sonic harmonics that exist within the organ chambers and the building they were built into. This had been a given up until recently. https://www.youtube.com/watch?v=tFV9bC4BsK0 (I have not heard this instrument in person). It is reported that the"experts" have deemed this monster to be the virtual equal to an installed pipe organ.

Thanks for posting such a good blog .

Plastic injection molding