Sorry to go back a few posts, and I know you are getting hammered from many sides. But, I do not see a proven "fact" in your statement above. I am not saying the Clark paper is factual either, and Clark is not here to defend it. But, it is food for thought.
What you keep harping on is merely your belief or hypothesis about longer term testing, not a fact. There has been no controlled experiment to prove your belief is true or factual, as far as I am aware. Even proving (or repetitive hand waving) that says the Clark paper is wrong, still does not prove as a fact that long term listening "can often reveal unpleasant aspects of the sound" or anything else.
My own belief, while I do not try to glorify it as fact, is that longer term testing is a good thing. I try to do that myself as much as possible in auditioning new components. I further believe that we, as audiophile consumers, need longer term testing specifically as a substitute because we do not generally have access to well controlled, double blind short term listening test capabilities for most all components, especially speakers. We are also not conditioned, trained or experienced at doing short tern, double blind. That, and it is a lot of trouble to try to do.
I do believe that better double blind unsighted auditions, if we had access to them which we don't, would help us discover component differences much more quickly. I still believe that longer term listening can be of some value, at least in confirming our short term impressions with a wider mix of recordings and under less stressful conditions. This greater assurance or comfort level is important in buying decisions.
I have done a fair number of long term listening tests to a variety of components, some of which I bought. Anecdotally, I cannot recall a single case where my sonic opinion changed materially between day 1 or 2 and day 30, if indeed the audition carried on that long. I think I did feel that I was generally more confident in my opinion after longer listening. And, I have not turned over my system frequently due to sonic dissatisfaction discovered after purchase. I generally keep my stuff for years.
I might add that component comparisons for my own system are something I have been doing for many decades, and I have tried to learn how to listen and to make good choices for myself. I have gotten better at it. We had a discussion elsewhere about learning a "tell" - the specific sonic characteristics - that make up the difference between components (not something extraneous in 2 different versions of a recording as we were discussing). Often it takes a little while to discover and then pay attention to that in comparative listening. But, I think with experience, one gets better and quicker at identifying it.
I don't disagree with most of what you said & I've said the same myself - long term listening is the best, easily available method we have of evaluating components that we wish to put in our playback systems. Yes, it can be flawed but the long term nature of this listening generally irons out most, if not all of these flaws. I'm sure you have changed your impression of a component usually in the first few sessions of listening - your initial impressions change i.e you hear a difference but don't know if it is more revealing, more accurate (or less revealing, less accurate) until you have listened further?
My criticism of Clarke's (& Nousaine's) paper still stands & hasn't been refuted - I don't call it hand waving - just pointing out how some people leave their critical analysis at the door when they see AES in a paper. And also pointing out obvious experimenter bias in these tests (is this the Clarke that co-developed ABX testing with Kreuger?)
I consider when enough people report the same observation that this constitutes a valid observation to me & therefore a fact (if I confirm it myself). I see this in people's reports of their long term listening revealing aspects of a component that they hadn't first noticed. I consider this reliable evidence that long term listening is a more robust & accurate way of evaluating components. Yes, this may then suggest an A/B listen is done for confirmation. But the point is this new aspect that is being listened for in this A/B listen would never have been uncovered without the long term listening so if we were relying on just A/Bing we would be missing this difference
Last edited: