10 minutes my *ss! ;) I definitely had a clipping problem. Fixed that by using a -10db analog attenuation baseline for all tests. Also increased the recoding time to 20 seconds and now I am getting fairly consistent results between runs. Tested each setup three times, so there are 6 results for each run (left and right).
-0db baseline: 99.7, 90.0, 87.2, 84.8, 92.2, 86.4 Average: 90.1 -10db digital: 75.5, 82.9, 84.1, 86.4, 74.8, 80.5 Average: 80.7 -20db digital: 69.9, 70.1, 68.4, 73.8, 66.7, 71.2 Average: 70.0 -30db digital: 55.0, 68.7, 53.1, 60.7, 50.6, 56.5 Average: 57.4 -10db analog: 83.0, 83.3, 80.2, 75.7, 86.0, 82.8 Average: 81.8 -20db analog: 74.4, 73.8, 65.4, 70.8, 77.7, 75.7 Average: 73.0 -30db analog: not tested (not enough jumpers!) So as you can see the loss of information in terms of correlation depth reported by ADM is equal to the amount of attenuation. The reason we also lose almost as much when using analog attenuation is because this test is flawed - we are not level matching before the ADC so we are limited by the recording resolution. I would invite someone else to try this with the level corrected before the ADC. This is why the benchmark would be ideal for this - it has 0, 10 and 20db stepped gain selectable from the front panel. -- seanadams ------------------------------------------------------------------------ seanadams's Profile: http://forums.slimdevices.com/member.php?userid=3 View this thread: http://forums.slimdevices.com/showthread.php?t=60041 _______________________________________________ audiophiles mailing list [email protected] http://lists.slimdevices.com/mailman/listinfo/audiophiles
