|
diyAudio Member
Join Date: Jul 2002
Location: Toronto, Canada
|
Re: Millwood
This is what I found on the subject of testing. But I still don't know , how I can arrange it in time and space, so you will not have any doubts as to my results. Can you provide suggestion now without repeating yourself?
Quote:
Originally posted by millwood
Andy, I am happy to provide one, if you are OK with it. I am not sure how scientific it is, tho.
How about getting a few test subjects. Divide them into three groups. Those who believe in pad sound, whose who don't, and those who don't know about it.
Provide two identical gaincard amps, one with a silpad (amp S) and one with an AO pad (amp AO). do the usually trick of volume adjusting them, etc.
Conduct a series of double blind A vs. B listening tests, and ask the subjects two questions:
1) if A is different from B;
2) if A is preferred over B.
A / B can be either amp S, amp AO, or both. So the possible combinations are
amp S, amp S
amp S, amp AO
amp AO, amp S
amp AO, amp AO.
at alpha of 95%, run a hypothesise that amp S sounds the same as amp AO. (t-test I believe).
at the same confidence level, run another hypothesis that amp S is preferred over amp AO.
it would be interesting to know how well doug blackburn will do in such a test, 
|
Quote:
Originally posted by ALW
What time interval do you propose for your tests and evaluations?
The ear can only detect certain fundamantal changes during a relatively short assessment time (read: minutes, hours).
During this time one can make simple assesments related to amplitude and frequency balance. This is the problem with DBT - short time spans, matched amplitudes etc. reduce the ability to determine any difference at all. It's a null result, not a proof.
The assessment of the underlying improvement in the music (totally unrelated to both the above criteria) often only comes after several days listening, or at least a good few hours. This is a critical point - frequency response errors, changes in amplitude have NO EFFECT whatsoever on the fundamentals of the music, providing the devices are operating within their envelopes, and the changes are not so extreme as to cause other ill-effects.
This is what makes many people think that speakers are the worst part in the chain, as they have the grossest reponse / amplitude errors. They are most important to the sound least important to the music.
Quick A/B's do not work, they fail to take into account the way we hear, the processing of which occurs in the brain.
I think Jonathan Carr's method, proposed previously, of making externally identical amplifiers labelled A / B or C, with the salient details identified internally is a better method.
One has to ensure that the insides cannot be seen / viewed without obvious evidence you then simply pass these amps onto a number of individuals who report their experience / preference over a long time period. The data / listening notes are then analysed subsequently.
How many times have you changed something with the immediate result that you think it's better, only to find, several days later that it isn't? This is a common theme amongst those of us actually doing this stuff.
I've done a lot of design work with remote partners, where improvements have been made and both of us have reported identical results, sonically, and this has often been with people I would view as having daimetrically opposed views to my own, with different systems, musical tastes etc.
I strongly believe that better is better for most people and this has been a most surprising finding for me, the difference in opinions solely comes down to the actual location you are along the audio path to nirvana - I may not like your system, you may hate mine, but an improvement to a common area is likely to be the same to both of us.
Andy.
|
Quote:
Originally posted by SY
Quite the opposite. Peter is claiming to have possibly found an interesting phenomenon. Putting aside the notions of analogy, there's not much in the way of a clear mechanism in the electronics to explain this- unless Audio Freak's hypothesis is correct. Either way, there's potentially something important here, and those of us who would love to see the state of the art advanced, rather than just speculated about, are trying to figure out a way to pin this down.
Now, measurement is a fine thing IF you know what you're looking for. And IF you know you're not chasing a chimera. So the first thing that has to be done is to take the instrument that is giving the "funny" result (in this case, Peter's perceptual system) and run the experiment. It's possible that there's a real phenomenon here. It's also possible that Peter is falling into the normal human trap of bias and expectation. It's improper to assume one or the other in the absence of evidence. So the way to cut this problem in half is by LISTENING, not measuring.
And fortunately, through more than a century of sensory research, the methods exist to use the human faculty of listening to determine what's going on. Since Peter, like most of us, doesn't have much experience with sensory testing, my suggestion has mutated to something about which he (and apparently others) feel just as confident, but less emotional.
So, let's give the rhetoric a rest and start with some baby steps.
|
__________________
www.audiosector.com
“Do something really well. See how much time it takes. It might be a product, a work of art, who knows? Then give it away cheaply, just because you feel that it should not cost so much, even if it took a lot of time and expensive materials to make it.” - JC
|
|
|
|
.This thread originated somwhere else, but since it is a discussion between two persons mostly, about matters that may not be of interest to other mebers, I decided to open a new thread on this specific subject and don't pollute the other thread
