Pro-Codec User Manual

5.6 A-B Auditioning and A-B-X Testing (A-B Tab) 5 THE PRO-CODEC PLUG-IN
the same button will alternate between that
button and the previously selected button. This
means you can repeatedly alternate between A & X with your eyes closed (or B & X, or A
& B) because you don’t need to move the mouse pointer. Eyes-closed auditioning might
improve your listening focus!
As with A-B mode, there is access to the input signal (I/P) by clicking in the B column
adjacent to a selected A codec. Press RESET to reset the internal counters and begin a
new test. The display window beneath the RESET button displays the number of
completed test cycles.
X is randomly set to either A or B. Listen critically, switching between A, B and X. When
you think you have identified X as either A or B, leave the selection on the identified A or
B, and press CHOOSE. This will increment the number of test cycles but not display
whether your choice was correct. The act of pressing CHOOSE also randomly
re-allocates either A or B to X. For a thorough and statistically significant test, repeat this
audition and CHOOSE cycle up to 20 or 25 times.
When you have completed a satisfactory number of cycles, press REVEAL, which will
show the percentage number of correct guesses:
A 95% confidence level is regarded as statistically significant - anything less than this
means you have not reliably been able to distinguish between A and B!
Note:
A 95% confidence level is not the same as a 95% correct score! Think of a single
cycle; you might by chance guess correctly and obtain a score of 100%, but with a
confidence level of only 50%.
As a quick guide, for a 95% confidence level you need a minimum of 8 correct
guesses from 10 cycles (80%), 11 from 15 (73%), 14 from 20 (70%), or 17 from 25
(68%).
www.sonnox.com 28 Go to contents