Setup: MacBook, USB A-USB B printer cable, LH Labs Linear Power Supply 4, LightSpeed 1G USB cable to LH Labs Pulse X Infinity, high gain, Femto Time Mode digital filter, XLR balanced headphone out, HIFIMAN HE1000 [beta]. I was using the original 24/96 file and the original 16/44.1 downsampled file, so the DAC did have a delay for sample rate switching. I wanted to do this in case someone made the argument that an upsampled 16/44.1 file to 24/96 somehow makes the ABX test invalid. Foobar's output was via ASIO, minimum latency, small 256 sample buffer size.
foo_abx 2.0 report
foobar2000 v1.3.8
2015-07-13 11:31:47
File A: A.flac
SHA1: 8ce7cdf117f2f7a0a4c934dcd04724c593842218
File B: B-1644_Original.flac
SHA1: a82d0e8be147c62b4532a1b9316795d13c7b16e8
Output:
ASIO : foo_dsd_asio
Crossfading: NO
11:31:47 : Test started.
11:33:54 : 00/01
11:34:50 : 00/02
11:35:39 : 00/03
11:36:03 : 01/04
11:36:41 : 02/05
11:36:58 : 02/06
11:37:10 : 02/07
11:37:38 : 02/08
11:38:02 : 02/09
11:38:47 : 03/10
11:39:34 : 04/11
11:39:47 : 05/12
11:39:47 : Test finished.
----------
Total: 5/12
Probability that you were guessing: 80.6%
-- signature --
b6ec774cd9e35aef1a5436dac9938588881fcbbf