Trust your ears? Trust your measurements? Look to psycho-acoustics for help? Is the psychology of being human the most important thing?
by, 06-25-2012 at 09:43 PM (1657 Views)
I am going to quote Barrows here. It is a complete post on his careful methodology of listening. I quote because I think it is an excellent description of how to audition for differences in components. I agree with it completely. He was responding to Julf's question about how he listened carefully:
Well, thanks Julf... As far as listening goes, I suspect that each of us may find things we can do differently, that is, what works for me, may not be the same for you. But a few things: blinded does not work for me, it just introduces stress which makes it harder to remain objective. Stress is really one of the main problems, along with expectations. If one is tired, dealing with work/relationship stress, or not having a good day, forget about listening for evaluation. Listen alone, I find things like group "shootouts" really distracting, and the presence of others and their opinions in the listening area is distracting at best.
I generally use some trusted tracks at first, but this just gets a baseline on overt differences. Longer term listening with a wide variety of music ultimately tells me more about how something sounds. Not to get all "groovy", but letting the music come to you really helps me, rather than trying to define the exact boundaries of the soundstage or get super analytical about what the differences are. One should accept that their system will sound different at different times, due to room temperature and humidity levels, not to mention AC quality variations, and temperature stability of components, and consider that these factors may be influencing the results. Of course, keep a reference system as constant as possible, and only change one variable at once.
I find it easier to evaluate the generalities first (does it sound "better", ie, do I get drawn into the musical interplay and the emotional power of the music more easily) before trying to decide what specific playback qualities might have changed.
More and more as I listen for pleasure, I find I move away from the analytical "wow, that soundstage is really precise" and more towards getting the gestalt of the music. I know things are really good when I find my mind engaged in feeling how two musicians are reacting to the subtle nature of the emphasis of their playing.
1. do not be stressed
2. be alone
3. empty away expectations
4. do not be hungry
5. do not be drunk
6. be well rested
7. darken room
8. close eyes (but only if it does not make you nervous)
9. listen to a wide variety of music over a long term when possible
When I worked at PS Audio I evaluated things both in PS' listening room, and at home. But I waited 'til after business hours to use the listening room when no one was around to avoid the distractions of "work". Using two systems I was familiar with also helped me to have some additional perspective-I kind of miss that now.
Next another post by Barrows in answer to Julf asking how he protects himself from bias.
"You have a good point about the stress, and I totally agree with it. But if you can't resort to blinded tests, how do you protect yourself from perceptual biases?"
At some point I have to accept that some bias may creep in, but my experience shows that is (usually at least) not the case. Considering that the additional stress from being blinded generally makes the results moot, that is not an option. I find that short term tests are unreliable, and most subject to the kinds of bias that folks here talk about, but multiple longer term listening, in my experience, eliminates bias as a problem for me. Many times I will go into a test, where my expectation is that something will be one way, but my listening results are counter to my expectations-this happens often enough that I gain confidence in my ability to hear difference without biases affecting me.
Like anything, this kind of listening is a skill which is developed purposefully over time-I would not expect random subjects picked off the street to be able to do as well.
When I first started working at PS Audio, I took it upon myself to do critical listening for at least an hour every single day, as part of my job, often testing different components, cables, power conditioners. After about a year of this, I had developed skills which I did not posess previously, despite being an audiophile before working there.
Now, I prefer to listen for pleasure, and keep my analytical hat off most of the time, but every once in awhile I am called upon to test something new, and I find my skills remain pretty sharp (but not as good as they were when I was doing critical listening every day).
And next a post where I replied to Barrows with some of my thoughts on the matter:
I think your listening methodology is sound. It also will enhance your ability to perceive real differences as far as human senses can go. I would also mention if someone had asked me how to go about listening carefully for comparative purposes my instructions would have been identical to yours. Not just very similar, but identical. That is how I have done such listening for some number of years.
While you might be able to avoid bias better than most, and I do think repeated experience helps in that regard, you still are subject to it. Human minds and perceptions have certain common ways of processing things we are all subject to those. Somewhere around the edges of the perceivable or beyond bias would affect you or any of us.
You have written:
At some point I have to accept that some bias may creep in, but my experience shows that is (usually at least) not the case.
Quite a bit of careful work has shown this statement is common, and nevertheless people's senses usually get biased much more than they think it does. Again this is where the fascination of the issue is for me. I have accepted the idea no matter how careful and unbiased I think I have been, no matter how certain I am of my perceptions they can be wrong. They aren't always wrong, and there have been a number of situations where I was getting a niggling nagging idea something was up and eventually figured out why. Found a reason for it and determined it was true. That rightly reinforces the idea you trust your ears. But with interconnects there simply appears to be no reason that anything audible is happening and that should reinforce the idea you also doubt your ears. It bears investigating and it might be someone will come up with something. My best idea now is there is nothing there beyond my ears getting fooled.
So the murky, messy, in-between area is fascinating and requires some different practical or philosophical way to deal with it. If you assume all hearing abilities are completely understood it makes it simple, but in fact we know that isn't the case. On the other hand much is understood and shouldn't be dismissed either. We can say the same for technical considerations. Much is understood, but not everything. Then there are the psychological aspects of it. While we understand quite a bit more than nothing, we probably are on shakier ground here than the other two areas.
So you have interactions in at least three broad areas of knowledge. Even 3 sets of interaction get much too messy to keep straight in people's minds without some extra effort and care in applying that knowledge. That doesn't leave this area hopeless. Much of science deals quite effectively with much more than 3 variables and incomplete information.
A friend always warns me about reasoning by analogy. He is right and the more he points it out the clearer it is he is right. Still I will give in to the temptation here.
Ever see someone try to adjust a color television when they don't have any real knowledge of what the controls do? Or do you have a memory of doing that yourself? You know what stuff looks like, and there are only a few controls. So how could it be so hard to at least get close? Well you can end up with quite the colorful psychedelic result. I have seen people work and work and finally be satisfied with the adjustments. Then argue when you point out how far from right they are. After staring long enough they really don't see the green skin tint or the blue-greenish skies. If you finally get some reference (once used a white piece of paper held up to fluffy white clouds on screen that had a cyan tint), then it will flick something and they see it. Usually at that point they get disgusted and shake their heads at how hard this has been.
But with just a bit of extra understanding you can make it work pretty well. The brightness is a black level control, wait for dark scenes and adjust. Contrast is a white level control wait for the white dress or fluffy clouds to adjust. The color balance is tougher as magenta and cyan are harder to get in TV scenes. Skin tone can stand in. One can do a pretty decent job that way. All you have is a bit of extra understanding at how it works, and know which of the controls to adjust in which order. Only takes a handful of minutes to do what someone else might spend hours on and not get right and be convinced they have. Yet, even more technical methods give even better results. Would be foolish to claim only using your eye and viewing over time can one get truly good results. Yes, an experienced person can improve on the results with the eye, but not at the level of instrumentation. And yet like hearing the eye is the final arbiter. Without the human visual perception the rest has no point.
The etiology of how these differing opinions about cable come about with the overlapping, incomplete knowledge from three broad areas of knowledge is where progress can be made. If only we can get agreement such is the case, and people quit picking just one of the three areas while claiming it to have near religious-like infallibility. Then progress could be made on reliable methods to work around all this.
I invite discussion of this. Not about whether we hear cables or not or similar divisive subs-vs-obs commentary. But about how the fact we have three different areas of knowledge all of which are incomplete, all three of which are very useful, and which overlap. Discussion about effective ways to use the three together for a better more complete result and not as warring camps against each other fueled by religious fervor.