deaconblues
500+ Head-Fier
- Joined
- Jun 2, 2005
- Posts
- 500
- Likes
- 16
Your strawman isn't very convincing. That article is about the difficulties of developing psychoacoustic models due to the variability among individual listeners. What that article states is that our current knowledge does not allow us to predict an individual's sensitivity to various encoding artifacts.
However, an ABX test doesn't seek to predict anything at all. An ABX test simply measures an individual's ability to match an unknown sample with itself. An ABX test may demonstrate variability among individual listeners. For example, a test may show that some people can reliably tell the difference between 192kbps mp3 and wav while others cannot. Variability among listeners isn't a factor that somehow reduces the validity of ABX tests. ABX tests with multiple subjects actually MEASURE that variability. You are confusing the development of models with making real world measurements.
However, an ABX test doesn't seek to predict anything at all. An ABX test simply measures an individual's ability to match an unknown sample with itself. An ABX test may demonstrate variability among individual listeners. For example, a test may show that some people can reliably tell the difference between 192kbps mp3 and wav while others cannot. Variability among listeners isn't a factor that somehow reduces the validity of ABX tests. ABX tests with multiple subjects actually MEASURE that variability. You are confusing the development of models with making real world measurements.