Hehe, yeah, we've been thinking about that. Let's take the worst case, if out of 100 people 75 are deaf we should still see a significant trend towards the accurate implementations from the remaining 25. If more than 75% of people are deaf, there's not much of a need bother with quality anyway, hence - spoken cynically - the outcome will be the same.glokraw wrote:How will you know who has two good ears? There are many musicians whose hearingUrs wrote: It's going to be something akin "can you spot the difference between a an expensive filter and a cheap one?" and "what's the defining threshold?". In other words, we're going to explore the boundaries of perception with a true double blind AB test.
has been thrashed, but still think their audio opinions are polished platinum.
Hearing range could be even more a concern when the testing is done on filtering.
Maybe do a small separate test using youthful orchestral music students
for comparison, to verify that results from the damaged-goods rock generation
are meaningful, or just marginal? A double-blind test among the double-deaf?![]()
Ouch![]()
Cheers
What's more critical is public discussion. If someone posts "hey, Filter 3 is the best, surely", many people might have their conclusion affected. Hence we need to do this in a way that the actual algorithms are different for each user - which could again be done by a random seed. We'll see...
