Psychotic Acoustics!

[Alloroc]

Following Chris' suggestion, here's the the thread to continue on the discussion with regard to Psychoacoustics and human Psychophysiology.

My own interest in this topic stems from the time I worked for Philips Electronics - consumer electronics division. I worked for them for 6 years before I set up my own business three years ago. I was based in Eindhoven, Netherlands but worked from home a lot of the time. I visited most sites including factories and research facilities. I spent nine months working on a project in the Philips Research Europe centre in Redhill, Cambridge, England. My project, was to work on the Philips Nexperia programmable logic/system system on chip range of semiconductors. My job was to design and build a new datacentre to support the project.

So what does all this have to do with Psychoacoustics? Well not a lot really. The lead engineer was a massive fan of analogue audio and had a most amazing vinyl collection. He believed that to the human ear, analogue audio, from software to source to amplification, was the most pleasing to listen to. I didn't argue. He introduced me to the science of Psychoacoustics and how various algorithoms are employed to 'tune' a digital source/device/format to produce a more pleasing sound.

So, please add your comments, thoughts and opinions to this thread and I request, with respect to all, please keep an open mind. Science is a really, really expansive subject!

[Kobus]

You have most of us running for our encyclopedias and dictionaries........ to no avail, zero help there.

I have been following this with great interest on the other thread and would love to hear/learn more.

Kobus

[stantheman2]

I find the whole are of the psychoacoustics of sound reproduction fascinating. I am an engineer, so my prediliction is towards applying scientific principles to these issues. However, my own experience is that science does not yet have all the answers.

For example, I replaced a "good" quality Monster Cable between my DVD player and my receiver with a solid silver digital cable from Catcables. From an engineering perspective, this should make absolutely zero difference - as long as a conductor can pass a digital data stream, its material and construction should have no effect. Yet, there was a very dramatic improvement that was immediately obvious to everyone in my family, none of whom has any interest in technical things.

Another example is CDs vs DVD Audio/SACD. Mathematically, one can prove that the 44.1 kHz sampling rate is sufficient to completely reconstruct any music stream up to 20 kHz, the accepted upper range of audibilty. Yet, all else being equal, the higher sampling rate of DVD Audio and SACD sounds better, Heck, even oversampling of CDs improves the sound. From a mathematical, data stream analysis perspective, this is cannot be true. Yet, it is easily audible.

Yet a third example is double-blind testing. Such testing on cables, amps, etc., consistently demonstrates that listeners cannot recognize differences between components or cables, to any statistically significant degree. Yet I hear very audible differences, and so does virtually everyone. Otherwise we would all be satsified with $50 boom boxes. The only rational conclusion I can come to is that double-blind testing somehow does not capture, or have the abilty to measure, those differences.

[Chris D]

By nature, aren't all of us geeks that post here "Psycho Acoustics"?

I find psychoacoustics fascinating. Just the fact that a change in a sound's volume, frequency, timing, distance, source, instrument, type, etc, can DRAMATICALLY alter the way the human brain perceives it, even more than the numerical change itself. We really are complex creatures, us humans.

You know, there's got to be a good book out there, or website, or doctoral thesis, or something, describing and explaining various phenomena. I'd be interested in reading something like that, if anyone knows of one.

So is the first type of psychoacoustics that "if a tree falls in a forest, and nobody is around to hear it, does it make a sound"????

[RobP]

Another aspect of this topic that fascinates me is how the music can actually take a physical effect on our bodies and spirit. I am amazed how wonderful I feel after listening to a great live symphonic concert, I walk away feeling euphoric and refreshed, anyone here ever experience this?

[Chris D]

Yes, this is true. There is a psychological effect that music has on a body and its well-being. However, music/audio can also be used for PHYSICAL benefits as well. A doctor that has associated with Clark Synthesis in the past has done a study and uses Clark tactile transducers for physical therapy. Here's the use for "physiotones":

http://www.clarksynthesis.com/specialty/Physiotones.htm

And here's a Clark Synthesis link on "physioacoustics":

http://www.clarksynthesis.com/specialty/Physioacoustics.htm

I'd still like to find something on psychoacoustics.

[SteveCallas]

Yet a third example is double-blind testing. Such testing on cables, amps, etc., consistently demonstrates that listeners cannot recognize differences between components or cables, to any statistically significant degree. Yet I hear very audible differences, and so does virtually everyone. Otherwise we would all be satsified with $50 boom boxes. The only rational conclusion I can come to is that double-blind testing somehow does not capture, or have the abilty to measure, those differences.

That's the ONLY rational conclusion you can come to? The more plausible conclusion is that when one knows which piece of equipment is being used, their conscious or subconscious bias makes them believe they hear differences that don't really exist.

Another thing to consider is that blind testing requires level matching, whereas just switching a piece of equipment and doing casual listening doesn't. A difference of a db or two can easily be interpretted as clearer, brighter, better bass, etc. There is no "flaw" to blind testing that hides audible differences, the "flaw" is in sighted testing.

[stantheman2]

A double blind test is a scientific experiment designed to test a hypothesis. For audio equoment, the hypothesis could be "component X and y sound different". Most double blind testing in such cases finds that the tested individuals, under the test conditions, cannot distinguish between the two components, to a statistically significant degree. The italics here are important. The double blind test does not prove or disprove the hypothesis; it simply indicates that the individual test, under those specific test conditions and with those specific test subjects, yields a certain result. Another test, under another set of conditions, with different tets subjects, could yield a different result. In some of the double-blind testing reported in magazines, some indiviuals could distinguish between X and Y component to a statistically significant degree. Were they wrong?

My point is that there is a preponderance of evidence that components and interconnects do influence the sound. It is true that much of the other evidence is more anecdotal, and not as "rigorously scientific". However, either all that other evidence is wrong, or else the design and specific conditions of the double blind test do not measure a difference that does in fact exist. Either answer coud be correct. In either case, at the end of the day, you are putting your "faith" in one position, based on your personal interpretation of the evidence, and your own personal decision as to how to weigh the evidence. You can choose to givbe double-blind testing 100% of the weight. but that does not make my positon any less valid because I believe double-blind testing is not the only viable evidence.

[Alloroc]

You know, there's got to be a good book out there, or website, or doctoral thesis, or something, describing and explaining various phenomena. I'd be interested in reading something like that, if anyone knows of one.

Lot to catch up on here. Been a busy weekend.

Anyhow, there's lots of reading out there....

Here's a good place to start....

http://www.auditory.org/postings/2003/463.html

and

http://www.elsevier.com/wps/find/bookdescription.cws_home/675639/description#description

Also, I read, or tried to read 'Psychoacoustics : Facts and Models' a few years back. It was a very technical read but it gave me a reasonable grasp on the subject.

http://www.physicstoday.org/pt/vol-54/iss-6/p64.html

Here's another interesting link. This lady editied a collection of essays that examines the psychology of music, aptly titled 'The Psychology of Music'! Explore her page - there's lots of great material.....

http://psy.ucsd.edu/~ddeutsch/

and

http://books.elsevier.com/us//apcatalog/us/subindex.asp?maintarget=&isbn=0122135652&country=United%2BStates&srccode=&ref=&subcode=&head=&pdf=&basiccode=&txtSearch=&SearchField=&operator=&order=&community=apcatalog

I have a bunch of links on an old laptop on this topic. I'll root them out when I get a chance......

[SteveCallas]

stantheman, I appreciate the debate.

In some of the double-blind testing reported in magazines, some indiviuals could distinguish between X and Y component to a statistically significant degree. Were they wrong?

Could you point me to some of these?

I'm not going to try and debate with you over whether or not blind testing will prove whether two pieces of gear sound different, because from your stance, even 10000 tests with the same results don't prove anything, and that is fine. I've done some of my own blind testing with amps, and will soon be doing some with DACs, and I have reached my own conclusions, just as you have probably reached yours. All I'm saying is that the logical conclusion - when blind tests show there to be no distinguishable differences - is that visual bias plays a role, not that double blind testing is somehow flawed.

[NMyTree]

I always thought and felt that it's in the harmonics of the music notes, the melodies. And then when you string certain harmonics or notes together, it has a profound emotional, physical and hormonal/chemical effect on the brain and body. I think there's several ranges of harmonic notes and note sequences that trigger specific endorphins, hormones and so forth in the brain/body.

I also believe that music triggers specific memory markers or flags, which stimulate emotional memories. Hence the reason why we hear a specific old song and all of the sudden we are transported back in time to a moment in our lives when we either were having a lot of fun, in love, in emotional distress or what ever the event; and we actually feel like we are re-living those events or momements. The emotions and memories coming rushing back like a flowing river and we catch a ride on those waves. It's amazing.

I'm no expert or scientist. Basically these comments are based strictly on my own personal experience as a music lover, since I was a a very young child; and the experience of other music lovers I know (friends, family and so forth).

[stantheman2]

Hey, SteveCallas, I also enjoy this type of stimulating discsussion. I can't answer your specific question about my sources - it was in one of the main magazines several years ago. I do/have subscribed to Sound & Vision, Home Theater, SGHT, Audio, Stereo Review, Sterophile, but I don't catalog the articles.

Also, please note I do not believe double-blind testing is inherently flawed in any way. I respect the process, and the result of those tests. Double-blind tests are quantitative and objective. However, the whole field of psychoacoustics is based on the relaization that both qualitative and qualitative, objective and subjective elements affect how we perceive sound, particularly music. The act of listening is a combination of the objective and the subjective.

The engineer in me says that differences in components SHOULD be easy to identify in double-blind tests. The engineer in me says that, if double-blind tests do not reveal such differences, then "meaningful" differences must not exist. Yet I DO hear differences that are meaningful to me, and make me prefer one component over another. Intense listeners like magazine reviewers or subscribers DO hear differences. Casual listeners DO hear differences. That dichotomy endlessly fascinates me. In all intellectual honesty, I do not feel I can reject or discount that large body of evidence because it is not confirmed by double-blind tests. So, for me, while double-blind testing is accurate and is valid, it somehow does not capture the complete picture.

[SteveCallas]

The engineer in me says that, if double-blind tests do not reveal such differences, then "meaningful" differences must not exist. Yet I DO hear differences that are meaningful to me, and make me prefer one component over another

I don't want to keep repeating the same thing over and over, so this will be my last attempt at this train of logic. If you can't hear differences with an unsighted, level matched test, but you can with a level matched sighted test, then it's easy to isolate the variable causing you to hear differences - knowing which piece of equipment is playing. It's just simple logic. If one doesn't want to see this truth for what it is, and look for off the beat explanations, so be it :T

[Victor]

... Double-blind tests are quantitative and objective. However, the whole field of psychoacoustics is based on the relaization that both qualitative and qualitative, objective and subjective elements affect how we perceive sound, particularly music. The act of listening is a combination of the objective and the subjective.

The engineer in me says that differences in components SHOULD be easy to identify in double-blind tests. The engineer in me says that, if double-blind tests do not reveal such differences, then "meaningful" differences must not exist. Yet I DO hear differences that are meaningful to me, and make me prefer one component over another. Intense listeners like magazine reviewers or subscribers DO hear differences. Casual listeners DO hear differences. That dichotomy endlessly fascinates me. In all intellectual honesty, I do not feel I can reject or discount that large body of evidence because it is not confirmed by double-blind tests. So, for me, while double-blind testing is accurate and is valid, it somehow does not capture the complete picture.

I know nothing of psychoacoustics, but I do understand the Double Blind Test methodology. It is unassailable from the scientific point of view. This is a view that demands absolute proof in a rigorous sense of that word every time you make a statement, - no exceptions.

With this in mind, the Blind Test approach to audio is a must. Think of it as a Control Group in an experiment. Without a Control Group your results are subject to unpredictable variations because of the infinite degrees of freedom that always exist in any experiment. The Control Group narrows the findings down to the subject matter while dismissing the ‘noise’ associated with the environment.

Bottom line is, - do you want to know the truth or you do not. Naturally your results are only as good as your test methodology. The logic as we understand it, dictates that in any environment, if there is a difference between the two groups, then there is a tangible reason to cause that difference. The Blind test makes this environment controlled and statistically will show that difference every time. You remove the Blind Test and you will show differences some time only and at other times you will not while nothing in the two groups has changed.

So, you see, - the Double Blind Test must be a part of any testing procedure beyond any conceivable doubt, otherwise we may as well flip coins in order to get answers and overall act as blind kittens.

[stantheman2]

One more try, and then I too will move on. First, the subject of this thread, at least as I understood it, is psycoacoustics. Here is a definition from Wikipedia:

Psychoacoustics is the study of subjective human perception of sounds. Effectively, it is the study of psychology of acoustical perception.

One purpose of a double-blind test is to exclude "subjectivity" to the extent possible. So, double-blind testing and pschacoustics are really separate subjects.

However, what really surprises me about the previous posts is the apparent viewpoint that, if differences are not discernible by some selected rigorous, "scientific" tests, then any perceived differences are somehow not real, or at least not valid. I must vehemently, but respectfully, diisagree.

One example from another area will illustrate my point. A number of year ago, Road & Track compared the Honda Accord with the Mercedes 230 sedan. They performed extensive, rigorous, industry-standard objective testing on both. By each and every objective measure, the two cars were virtually identical - overall dimensions, weight, weight distibution, wheelbase, front and rear track, interior room, interior noise levels, trunk space, acceleration, braking, cornering ability, slalom speed, etc. Yet the reviewers found the two cars to be significantly different! I dare say that anyone who drove both cars would find them different, and would prefer one over the other, even the previous posters. The same is true of audio components.

[Shawn Parr]

I think a bigger difference in the audio specific realm is who you get as subjects for your double blind tests.

Listening is a skill and an art. Anyone can hear music and a system, few people actually listen and can discern the subtle differences. You actually need to train yourself to listen properly. This can be very easy for some, and very difficult for others.

As such it is very possible, even with perfectly accomplished double blind testing, to get mixed results depending on if you choose individuals at random, or individuals that actually can listen and discern.

A great example of this was a Roland demo. Their multitrack recorders used to include a proprietary compression scheme that they claimed no one could tell from uncompressed. In reality if you knew what to listen for, while being very subtle, the difference was there. One individual that sat through their blind testing multiple times always detected it 100%. But since the subjects were always a huge mix of people, the results were skewed towards it being undetectable as no one else in the room was able to listen to the fine details. Although interestingly enough one of their samples everyone got right every single time. (They used something like 5 or 10 different songs with compression on and off).

[Alloroc]

Stan, I totally agree with your assesment.

Psychoacoustics is relevant and fundamental to both subjective and objective standpoints. In fact one could say it transcends them.
Manufacturers rely on psychoacoustics principles when developing digital audio products - be it hardware or software(inc. media formats). For example, Meridian. The firm concentrates on research activity in psychoacoustics, DSP, and servo control aimed particularly at good engineering for high-quality fidelity sound.

Sidenote - One interesting aspect from their research and development is addional audio channel - the height channel!

Anyhow, I'm neither a scientist or an engineer nor a musician for that matter. I use music to improve my life, by allowing myself to be taken to wherever my mind wanders to when I'm listening. I believe that I have, over the years, have developed the skill to truly listen to music for pleasure. Not as easy as one might think. However, listening to music to understand its structure, composition or accuracy to a source of sound is not what I'm about. Those are valuable skills but more in the realm of what musicians, audio reviewers or music students pay attention to (perhaps, 'blind test' testers too!?).

I listen to music to maximize various kinds of enjoyment of the sound, to give me meaning and creative stimulation and to let it affect me emotionally and physically. Such listening can have powerful benefits for the mind, the body and the spirit.

There is an old piece of audiophile advice worth repeating here: Listen to the music, not to the system. We spend much time tweaking our systems to hear improvements and listen to the noise instead of the signal. We often pay attention to the tones instead of the instruments, to the sounds instead of the meaning contained in the musical communications. There is no one sound to a piece of music. It can be heard in an infinite number of ways, which hold little semblance to the physical properties of the source or received signal. Music, like most of the art world, is a projective test to our imagination, emotions and attention.

[Chris D]

So, double-blind testing and pschacoustics are really separate subjects.

Absolutely right. We already have a thread where we're discussing the usefulness of "double-blind" testing. This thread is supposed to be separate from that, where we can discuss PSYCHOACOUSTICS. i.e. that between a 20 Hz tone and 80 Hz tone played at the same SPL volume, the human bran will interpret the 80 Hz tone as being louder.