hey just a side note. i am not trying to offend anyone or disrespect anyone. my posts are just for educational value for those who are interested.
if you don't personally see the value of a direct comparison or double blinded testing, that is totally fine. if you prefer to do a 'long live-in listening' to form subjective impressions about headphones, that is also very valid approach. My responses were simply to correct inaccuracies in the other poster's comments. I meant no offense.
Below is the in-depth scientific explanation of how memory works:
There is a mistake made when the previous poster related "short term of auditory memory" only specifically to a side-by-side comparison. Your auditory memory applies to ALL listening situations. Just because you are more familiar with a sound from "living in," doesn't mean that your "working memory" will last any longer or you skipped using your "working memory."
You simple have converted some of that "working memory" into "long-term memory" through repetition. "Short-term/working" memory actually encodes information acoustically in your frontal lobe (more specifically: the prefrontal cortex which is linked with complex cognitive behavior: judgement, concentration, and problem-solving) and parietal lobe (which includes the "primary auditory cortex" responsible for auditory sensory information & language). The conversation of "working memory" into "long-term memory" is done by hippocampus which converts the acoustic information into semantic information (signs/symbols/concepts) that is stored throughout neural connections within the brain. Your long-term memory will store your auditory experiences as subjective conceptual interpretation rather than the exact audio information.
To truly learning something like the skill of multiplication, you will use "procedural memory" based on "implicit memory" (aka doing many multiplication problems or reciting the table, you learn the associations). That is separate type of memory from the recognition or recall of different auditory information. For auditory information, they are usually stored as episodic memories aka "personal memories where you remember the sensations and emotions of that event." Just like when you hear a familiar song, you remember that time at the college bar where you last hear it and how you felt happy because the cutie waitress was winking at you.
So taking the "stored memory" of your background experience and comparing to another "stored memory" or "working memory" is tricky and can very subject to many bias and inaccuracies. Human long-term memory can be quite prone to error for specific sensory information as that information is stored conceptually or with emotional-associations. Short term memory is actually more reliable for judging audio information as your brain stores that acoustically as given.
Also, there is nothing about a direct comparison that prevents you from first gaining 'experience' with the headphones and building up your "long-term memory" prior to doing a direct blinded comparison. The only issue is that the conversation to your long-term memory can alter the actual auditory information as the information is not stored acoustically, so your "experience/long-term memory" can be actually be inaccurate & subject to conceptual bias.
So that's that on the auditory memory aspect of everything.
Now for Double Blind Testing:
The article linked presented a straw man argument. The double blinded test the author set-up was incorrectly set-up and already flawed. The application of a double-blind test would not be for which sound you like better. The application of this test would be for determining if a listener can reliably pick out the correct sampling rate out of 2 test tracks identical in all ways except their sampling rates. The trial is run multiple times. The sampling rate is unknown to the listener, hence the first "blinding." Every other factor should be controlled (same headphones, same dac, same amp, same player, same volume... etc). The tester recording the results should not know if the listener was correct or incorrect. Hence, the double blind. That is all there is to a double blind test. Isolate the factor you want to see make an impact on the sound, and blindly test if you can reliably detect the difference in sound per changes to the factor. Remove researcher bias by blinding him as well.
Now for my actual personal opinions:
I do personally agree that there are sonic elements that you can gain more appreciation for after long periods of time. For headphone comparisons, not only does that give you time to get used to an unfamiliar sound signature and prevent brain bias from picking what is familiar to you, but also you can notice subtle things with different specific tracks after listening to them many times. However, does that mean you have proven whether there is a difference or not? No, you have simply formed subjective opinions. Was the subtle difference there the whole time and your brain just noticed it now? Was it something that your brain just never converted into a long-term memory previously? Or was it due to some change in your physical/digital audio system? You cannot know the answer to that question without empirical testing.
To prove a difference exists, the golden standard used in the scientific community is a double-blinded test with proper statistical analysis. Whether you wish to engage in more controlled testing is a personal choice, but to refute the validity of a properly constructed double blind testing as a means to gather information to a certain confidence interval is to argue against the empirical scientific process itself, which we have used as a species for all notable scientific breakthroughs since the 17th century.
Hope this post clears up any misunderstandings & hope everyone has a great day :) Any further questions can be PMed to me as to no longer derail the thread. Cheers!
Edited by money4me247 - 3/22/15 at 6:57pm