Here's my parting shot before I bow out of this thread.
If you're a player, and you want a Brazilian rosewood guitar, then go ahead and get one. Anything that inspires your playing is worth spending a little more money on (if you have it to spare), but don't
take part in any blind testing.
If you're a player and want to separate the myth from the truth, do some blind testing. Although it is very difficult to carry out thorough scientific tests, a few simple A vs B tests are worthwhile. Wear a blindfold, and compare 2 similar guitars but with different back and side tonewoods. Do this both as a player and a listener. Repeat this test whenever you have the opportunity. I'll be surprised if you do significantly better than random chance (50% if you're just testing pairs of guitars).
If you're a luthier, you have a choice. The vast majority of us have probably never done any serious blind testing, so it is not possible to know
whether the differences we think we are hearing are real or not. Blind testing is an effective way of not only eliminating possible distractions and prejudices, but can also improve auditory ability (shown by a report published by the Acoustical Society of America, amongst others). You may be disillusioned, and even demoralized, as Kenny Hill was...
Kenny Hill wrote:
"...I've done blindfold tests on two occasions, and I don't think I will do it again. It's too unnerving. My own blindfolded listenings have been among very sophisticated players, dealers and myself with my own guitars included among the guitars of high quality and low. Blinfolded, all of us were completely lost. We couldn't tell ports or not, spruce or cedar, our own guitars, even cheap or expensive. That is demoralizing. I never want to do that again..."
...but you will have more information about what differences (or changes) are important or significant in terms of sound production.
I've done some blind testing myself, and been involved with a lot more that was carried out as part of the Leonardo Guitar Research Project (see more here
, and elsewhere in the luthier's forum). The Leonardo project results involved both experienced and student luthiers, and also professional and student players. There is a lot of information in the results, and as I have said before, the test conditions were not perfect, but the overriding conclusion was that most of the time, most players and listeners could not tell the difference between similar guitars built with tropical and non-tropical tonewoods.
In another recent test, one of my guitars was played alongside other maker's guitars. Although I wasn't quite as lost as the listeners reported by Kenny Hill, I did fail to pick out my own guitar, even though it had maple back and sides (I mistook it for a Kevin Aram Indian rosewood guitar), and also failed to pick out the one cedar top guitar in the group. Some listeners did a little better, but not significantly more so than you would expect by chance.
There is no doubt in my mind that if you change the tonewood used for back and sides of a guitar to one with different physical properties, then the sound of the guitar will change. The question is, is that change significant, and I think a good definition of "significant" is whether a player or listener can distinguish any change in a blind test.
I also have no doubt that some listeners will have greater ability to hear these subtle differences than others, and those in this thread who believe they can reliably hear a difference may be at the upper end of the "auditory ability" scale, but I would still suggest that the only way to be certain of this is to do some blind testing. It's scary (particularly for luthiers), but I do not regret having done it, and (unlike Kenny Hill), would be very happy to do more.