Article: Why We Need Audiophiles
Reply #625 – 2009-04-28 23:33:09
The widest dynamic range commercial recording I've ever found (and I've been searching for about 10 years) has about 80 dB dynamic range. I remember you noting this a few months ago in a different thread here. Would you mind sharing which recording (CD?) this was? Now to blind listening again. I had time to read the Stereophile article about the HE2005 debate .Yet the ABX test, consisting of 10 different 60-second musical excerpts that took close to 2.5 hours to get through, failed to reveal statistically significant differences. No wonder in my opinion. Summed up reason: fatigue => high probability of grading errors => inconclusive results. You need to concentrate the entire time, especially at the beginning of an item because you have no clue as to which stimulus is the original and which the UUT. At Fraunhofer, we try to limit our blind test items to 20 seconds maximum each. I personally am even more strict when conducting blind tests. I try to limit one test session to 7-8 items, 12-15 seconds per item. This way even the slowest listener finishes the session within 30 minutes. Why do I do this? Because it seems to me the test results became more reliable when I started doing this. Moreover, listeners were much more willing to participate because they knew they didn't have to spend "more than an hour listening to boring and annoying music and speech". A motivated listener is better than an unmotivated one. Which is also why we usually offer candy to the participants If I would have been asked to participate in quoted 2.5-hour test, I would have rejected the "offer". The mentioned, supposedly experienced sound engineer setting up that ABX protocol could have done much better than that. Chris