There’s no sense in having a “controlled” test for multichannel vs. stereo. The thing is immediately obvious. It’s more obvious then choosing between the rear-wheel-drive and the 4-wheel-drive versions of a car.
I would strongly disagree with this.
If there is one thing science is good at uncovering, it's that we can be wrong in inferring from what seems "obvious" to us.
All you have to do is realize that all of the more daft claims in high end audio - from super pricey ethernet cables to AC cables to audiophile fuses etc - come with testimonials of people who truly believe what they hear is "obvious." Listening tests that control for bias tend to wipe away this "obviousness."
The retort about stereo vs multichannel would likely be "But in those other cases we know they are wrong, theory and practice tells us there is no change to the signal so we can presume they are likely imagining sonic changes. Whereas in stereo vs multichannel, we know the sound is in fact different - theory and practice suggests it."
True. But even when real world differences are audible, bias effects still occur. An obvious example is all the studies done on speakers and how sighted bias effects can skew the perception of the sound. Many have found "obvious" differences can be much reduced...sometimes essentially disappear...in some cases, when controlling for bias variables.
It's quite possible that something we prefer under sighted conditions could change when "only the sound" is being evaluated, even possibly surround sound vs stereo.
Basically, science has shown us to NEVER underestimate the power of biases in regard to our perception and inferences.
But, as I've argued many times here, that's a condition underlying every inference and conclusion we make, all day long, so pragmatism has to play a part in what we demand for *reasonable* (and provisional) inferences. This is why I've mentioned that, especially in regards to the reasonablness of our everyday claims and reports of experience, a general heuristic of "extraordinarhy claims require extraordinary evidence" comes in to play.
So applied to my own surround vs stereo comparisons: Unlike, say, hearing differences between ethernet cables, on technical/theoretical grounds it's likely, expected, I'd hear a difference between a track played on my stereo vs my surround system. So my reporting of hearing a difference is entirely plausible. It's also plausible I could prefer one over the other. So...not an extraordinary claim that "contradicts" any science on the matter. It's like my saying I prefer movies on my projection screen vs my TV set. Can I produce a scientific level of confidence? No. But there's nothing unreasonable in accepting the claim *provisionally* with the lower non-scientific level of confidence we accept most non-extraordinary claims. And if one wants to bring particular studies on the subject to bear, we are still talking about probabilities and even then we have to look at how closely the scenario subject A rendered his opinion matches the scenario of the study trials. That's why scientists are always SO CAUTIOUS about what they can infer from study trials!
So: It's one thing for a person to say
"Well, I'm looking for a scientific level of confidence about such claims, where the confounding variables have been controlled for." That's a perfectly reasonable goal of course. But it's another to say "
Since the claim does not meet the standards of science (sighted), I can dismiss it as wrong, or in error, or as mere fantasy, or presume it is unreasonable to accept such claims even provisionally."
You can do that to dismiss extraordinary or implausible claims (e.g. lots of the audiophile snake oil), but not where plausible audible differences are there and preferences
can indeed differ.