I knew about the great
www.pcavtech.com for some time but I finally tested PCABX with interesting results.
To my ears I could discern the following.
PC-ABX v1.75 (16 tests each), WinXP
http://www.pcabx.com/training/index.htm
(tested with Creative Audigy <- Creek OBH-11 <- Sennheiser HD600)
The percentage-values are for "Probability You Were Guessing". Only values < 1% are acceptable. I did ran some tests multiple times until I reached a value sub 1%.
+-----------------+
| Tone Quality |
+----------------+
Very Easy- 0.05%
Easy- 0.05%
Harder- 0.05%
Difficult- 0.5%
Might be Impossible- guessed
+-------------+
| Loudness |
+-------------+
Very Easy- 0.05%
Easy- 0.05%
Harder- 0.05%
Difficult- guessed
Might be Impossible- n/a
+---------------------+
| Noise, Distortion |
+---------------------+
Very Easy- 0.05%
Easy- 0.5%
Harder- not good enough (5%)
Difficult- n/a
Might be Impossible- n/a
Here I really noticed how much noise / distortion the Audigy has by itself.
The reference example didn't sound flawless (like it should) at all.
This will be a good test I'll re-run when I get the Audiophile 2496.
I wonder how much the A2496 by itself will improve upon this test...
+------------+
| Distortion |
+------------+
Very Easy- 0.05%
Easy- guessed
Harder- n/a
Difficult- n/a
Might be Impossible- n/a
Same as above. Very very bad. Anyone with a A2496 ran these tests with good results? Would be interesting to know how different this testing goes with an external DAC attached to the A2496 and with non attached.
+--------------+
| Articulation |
+--------------+
Very Easy- 0.05%
Easy- 0.05%
Harder- 0.05%
Difficult- 0.05%
Might be Impossible- not good enough (25%)
I also compared different compression mechanisms:
castanets_060_lame386_128-h.wav vs. castanets-060.wav- 0.05%
castanets_060_lame383_256.wav vs. castanets-060.wav- guessed (100%)
Very nice tool. Highly recommend.
I don't understand however how I could possibly use this tool to e.g. compare an analoge sample of an Audigy with an analoge sample of an A2496 while using an Audigy. I guess you just can't.