• Welcome to ASR. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

DAC listening test files - update

This is Josh’s foobar log report, and it is valid. One question @Josh83 , MacOS or iOS was set to 16-bit resolution? It seems so from the report.

Code: 
foo_abx 2.2.3 report

foobar2000 v2.25.10

0126-08-18 20:16:49

File A: dactest_comp.flac

SHA1: 5636356ac8b5937a78ec403b31ee999fa44531eb

File B: dactest_ref.flac

SHA1: 12185333584a38af8a8f11dd2b4e40a198e3f0c8

Output:

Core Audio : Default, 16-bit

Crossfading: NO

20:16:49 : Test started.

20:19:18 : Test restarted.

20:19:18 : 01/01

20:19:24 : Test restarted.

20:19:24 : 01/02

20:20:03 : Test restarted.

20:20:03 : 02/03

20:20:11 : Test restarted.

20:20:11 : 03/04

20:20:35 : Test restarted.

20:20:35 : 04/05

20:20:43 : Test restarted.

20:20:43 : 05/06

20:21:02 : Test restarted.

20:21:02 : 06/07

20:21:11 : Test restarted.

20:21:11 : 07/08

20:21:22 : Test restarted.

20:21:22 : 08/09

20:22:01 : Test restarted.

20:22:01 : 09/10

20:22:01 : Test finished.

----------

Total: 9/10

p-value: 0.0107 (1.07%)

-- signature --

eb13c985e5d1efc0fa3f7027c105c63a8ec6ef25
 
pma said:
Yeah, you are right. Josh has taken the 2_DAC_test, in 44.1kHz. 44.1kHz/24bit. It is in the post #1 as part IV. I am sorry for confusion, I am on vacations with my iPad only, no test database. Please forget my plot, it was from another test.

But for sure there was no clipping. Peak value was -8dBFS.
Click to expand...
Aren't the -8 dBFS relative to the ADC full scale?

I mean, I don't see any evidence of the DacMagic+ clipping in this recording. It was just an idea to check.
 
pma said:
Part IV. Topping D10s vs. DacMagic+
A/B preamp switch was used
Click to expand...
Done and superbly failed (Windows 11 PC -> SMSLPL200MKI (as DDC) -> SMSL DO400 and Audiotechnica ATH-M50x)

foo_abx 2.2.1 report
foobar2000 v2.1.5
2026-08-19 18:30:49

File A: dactest_ref.flac
SHA1: 12185333584a38af8a8f11dd2b4e40a198e3f0c8
File B: dactest_comp.flac
SHA1: 5636356ac8b5937a78ec403b31ee999fa44531eb

Output:
Default : Speakers (SMSL USB DAC) [exclusive], 32-bit
Crossfading: NO

18:30:49 : Test started.
18:34:39 : Test restarted.
18:34:39 : 00/01
18:34:55 : Test restarted.
18:34:55 : 00/02
18:35:58 : Test restarted.
18:35:58 : 01/03
18:36:13 : Test restarted.
18:36:13 : 02/04
18:36:32 : Test restarted.
18:36:32 : 02/05
18:38:18 : Test restarted.
18:38:18 : 03/06
18:38:27 : Test restarted.
18:38:27 : 03/07
18:38:36 : Test restarted.
18:38:36 : 03/08
18:38:46 : Test restarted.
18:38:46 : 03/09
18:38:57 : Test restarted.
18:38:57 : 03/10
18:39:08 : Test restarted.
18:39:08 : 03/11
18:39:14 : Test restarted.
18:39:14 : 03/12
18:39:27 : Test restarted.
18:39:27 : 04/13
18:39:41 : Test restarted.
18:39:41 : 04/14
18:39:56 : Test restarted.
18:39:56 : 05/15
18:40:10 : Test restarted.
18:40:10 : 05/16
18:40:10 : Test finished.

----------
Total: 5/16
[COLOR=rgb(184, 49, 47)]p-value: 0.9616 (96.16%)[/COLOR]

-- signature --
89c70bd980da160dc633d9cf6bc17316aba9487a
 
RandomEar said:
I mean, I don't see any evidence of the DacMagic+ clipping in this recording. It was just an idea to check.
Click to expand...
I understand. Clipping would immediately result in sections of very high distortion in the delta file.
 
pma said:
This is Josh’s foobar log report, and it is valid. One question @Josh83 , MacOS or iOS was set to 16-bit resolution? It seems so from the report.

Code: 
foo_abx 2.2.3 report

foobar2000 v2.25.10

0126-08-18 20:16:49

File A: dactest_comp.flac

SHA1: 5636356ac8b5937a78ec403b31ee999fa44531eb

File B: dactest_ref.flac

SHA1: 12185333584a38af8a8f11dd2b4e40a198e3f0c8

Output:

Core Audio : Default, 16-bit

Crossfading: NO

20:16:49 : Test started.

20:19:18 : Test restarted.

20:19:18 : 01/01

20:19:24 : Test restarted.

20:19:24 : 01/02

20:20:03 : Test restarted.

20:20:03 : 02/03

20:20:11 : Test restarted.

20:20:11 : 03/04

20:20:35 : Test restarted.

20:20:35 : 04/05

20:20:43 : Test restarted.

20:20:43 : 05/06

20:21:02 : Test restarted.

20:21:02 : 06/07

20:21:11 : Test restarted.

20:21:11 : 07/08

20:21:22 : Test restarted.

20:21:22 : 08/09

20:22:01 : Test restarted.

20:22:01 : 09/10

20:22:01 : Test finished.

----------

Total: 9/10

p-value: 0.0107 (1.07%)

-- signature --

eb13c985e5d1efc0fa3f7027c105c63a8ec6ef25
Click to expand...
And several replies in less than 10sec indicate that is very easy to identify the difference between the two files.
But I guess that when the difference is spotted, and Foobar replay section is well calibrated, then it is that easy.
 
Thomas_A said:
I do not understand why Josh use a 10-trial test when he nailed all 16 trials for 5 DAC combinations.
Click to expand...
ABX DBTs are very, very, hard to do.
The world is full of people sitting on their butts typing on their keyboards telling everyone else what they should do.
There are very few people like Josh who actually get up and do it.
 
@NTTY Difficult to say anything. All we get are speculations, both here and at AS. There may be hundreds opinions, like those about quantum physics nonsense. We have some test data, including positive results, and this is all that counts, to me. Not the endless debates and opinions. Nothing based on sighted tests when nobody knows about test conditions and level matching. They are in the land of fairy tales.
 
welwynnick said:
There are very few people like Josh who actually get up and do it.
Click to expand...
Yes, but he is not the only one. He is known based on his AS review, of course I appreciate the efforts with abx tests. Debaters are here and there, but we need results, not opinions. And we have been collecting quite a lot of results in a few weeks.
 
Last edited:
NTTY said:
And several replies in less than 10sec indicate that is very easy to identify the difference between the two files.
But I guess that when the difference is spotted, and Foobar replay section is well calibrated, then it is that easy.
Click to expand...
And there is a usual question, if it was, hypothetically, based on a short sample in the length of 1/10 s order, that was found during a learning cycle, how would this be important during normal listening.
 
Proctored tests needed, really. This is all far too 'politicized' by now otherwise. No trust left.
 
Last edited:
welwynnick said:
ABX DBTs are very, very, hard to do.
The world is full of people sitting on their butts typing on their keyboards telling everyone else what they should do.
There are very few people like Josh who actually get up and do it.
Click to expand...
I’ve done my share of blind tests in the past and also do them occasionally as seen e.g. on ASR. My point above is still valid though.
 
Thomas_A said:
I’ve done my share of blind tests in the past and also do them occasionally as seen e.g. on ASR.
Click to expand...
Me too - too many in fact. I spent too long in the pursuit of HiFi; but it's just not that important. I should have spent more time with my family instead.
I wouldn't encourage anyone to walk in my footsteps.
 
Last edited:
pma said:
By now we were talking about tinniest differences between DACs "sound". But, will you hear a difference if we test such poor component like this one is?

Test - Topping D10s/Cosmos ADC loop vs. Tube preamp (6J1) in the loop

link to the test files:

6J1_test.zip​


drive.google.com
drive.google.com

Please feel free to try the test and tell me if you hear the difference, and do not forget to post foobar abx test report.
I have also included a delta file (Deltawave difference) for you to hear how horrible is the "sound" of tube distortion in isolation. But, can you hear it when playing music?
Click to expand...
Just for fun I ran foobar's bit comparator on two two files first

Code: 
Differences found in compared tracks.
Non-zero offset detected.
Extra leading/trailing sections contained non-null samples.

Comparing:
"E:\Users\SASULLY\Downloads\6J1_test\file_x.wav"
"E:\Users\SASULLY\Downloads\6J1_test\file_y.wav"
Compared 4520000 samples.
Differences found: 9039943 values, 0:00.000000 - 1:34.166646, peak: 0.597617 (-4.47 dBFS) at 0:59.112750, 2ch
Channel difference peaks: 0.403949 (-7.87 dBFS) 0.597617 (-4.47 dBFS)
File #1 peaks: 0.880109 (-1.11 dBFS) 0.773436 (-2.23 dBFS)
File #2 peaks: 0.899219 (-0.92 dBFS) 0.805715 (-1.88 dBFS)
Detected offset as 1 samples.

Comparing again with corrected offset...
Compared 4519999 samples, with offset of 1.
Discarded 1 trailing samples containing non-null values from file #1.
Discarded 1 leading samples containing non-null values from file #2.
Differences found within the compared range: 9039890 values, 0:00.000000 - 1:34.166625, peak: 0.250247 (-12.03 dBFS) at 0:00.954271, 2ch
Channel difference peaks: 0.156435 (-16.11 dBFS) 0.250247 (-12.03 dBFS)
File #1 peaks: 0.880109 (-1.11 dBFS) 0.773436 (-2.23 dBFS)
File #2 peaks: 0.899219 (-0.92 dBFS) 0.805715 (-1.88 dBFS)



Total duration processed: 1:34.167
Time elapsed: 0:00.136
690.90x realtime


Then I tried the files with fooABX in training mode for a few minutes, setting it to restart rather than continuous play, just to hear if I stood a chance.

I did this at first without 'zeroing in' on some micro second of the track. Just listenin' closely as the A or B or X track played, using my Amir-approved Truthear GATE in-ear headphones at normal playack volume, via foobar2k v2.25.9 on a Win 11 desktop, fooabx v2.2.3 . No replaygain, no DSP. Foobar's output device is AB13X USB audio (exclusive), 16 bit (not dithered). (I have a USB-C adapter attached to the output pin of the TruthEars)

Then I tried a short segment, where there is a lot of sibilance in her voice (@~ 48 sec)
Then I tried just the loudest part of the track (@ 1:02 and after, i.e., the bridge)
Then I tried a part with the lowest bass notes (@ 13-14sec)

Didn't matter. It was clear that my 65 yr-old ears did not stand a chance. I would literally be guessing.
So I did not bother to take a 16-trial test. Call me biased or insufficiently teched-up.

I loved the music. Nice voice.
 
Last edited:
pma said:
One question @Josh83 , MacOS or iOS was set to 16-bit resolution? It seems so from the report.
Click to expand...
I was interested in that and also the incorrect date - 0126. I tried the same version of foobar2000 and foo_abx and see the same thing so I guess it's a MacOS specific reporting bug?

Code: 
foo_abx 2.2.3 report
foobar2000 v2.25.10
0126-08-19 22:42:42

File A: dactest_ref.flac
SHA1: 12185333584a38af8a8f11dd2b4e40a198e3f0c8
File B: dactest_comp.flac
SHA1: 5636356ac8b5937a78ec403b31ee999fa44531eb

Output:
Core Audio : Default, 16-bit
Crossfading: NO

22:42:42 : Test started.
22:42:45 : Test restarted.
22:42:45 : 01/01
22:42:45 : Test restarted.
22:42:45 : 02/02
22:42:46 : Test restarted.
22:42:46 : 03/03
22:42:47 : Test restarted.
22:42:47 : 04/04
22:42:48 : Test restarted.
22:42:48 : 05/05
22:42:48 : Test restarted.
22:42:48 : 06/06
22:42:49 : Test restarted.
22:42:49 : 07/07
22:42:50 : Test restarted.
22:42:50 : 08/08
22:42:50 : Test restarted.
22:42:50 : 09/09
22:42:52 : Test restarted.
22:42:52 : 09/10
22:42:52 : Test finished.

 ----------
Total: 9/10
p-value: 0.0107 (1.07%)

 -- signature --
af12524848d0586d8ec3fe38f45213bf86b4b15e
 
  • Like
Reactions: pma
huw said:
;) I was interested in that and also the incorrect date - 0126. I tried the same version of foobar2000 and foo_abx and see the same thing so I guess it's a MacOS specific reporting bug?

Code: 
foo_abx 2.2.3 report
foobar2000 v2.25.10
0126-08-19 22:42:42

File A: dactest_ref.flac
SHA1: 12185333584a38af8a8f11dd2b4e40a198e3f0c8
File B: dactest_comp.flac
SHA1: 5636356ac8b5937a78ec403b31ee999fa44531eb

Output:
Core Audio : Default, 16-bit
Crossfading: NO

22:42:42 : Test started.
22:42:45 : Test restarted.
22:42:45 : 01/01
22:42:45 : Test restarted.
22:42:45 : 02/02
22:42:46 : Test restarted.
22:42:46 : 03/03
22:42:47 : Test restarted.
22:42:47 : 04/04
22:42:48 : Test restarted.
22:42:48 : 05/05
22:42:48 : Test restarted.
22:42:48 : 06/06
22:42:49 : Test restarted.
22:42:49 : 07/07
22:42:50 : Test restarted.
22:42:50 : 08/08
22:42:50 : Test restarted.
22:42:50 : 09/09
22:42:52 : Test restarted.
22:42:52 : 09/10
22:42:52 : Test finished.

 ----------
Total: 9/10
p-value: 0.0107 (1.07%)

 -- signature --
af12524848d0586d8ec3fe38f45213bf86b4b15e
Click to expand...
Wow, new world record, 10sec total duration and 1sec between 2 trials. I’m not sure I can click that fast between the ABX buttons. Challenge accepted :)
For curiosity, how many times did you try to get that one positive result?
 
NTTY said:
For curiosity, how many times did you try to get that one positive result?
Click to expand...
Yes, as said several times before, it would be fair if the posters would mention all their results. As @NTTY did in his DM to me.

Just an example: 3/10, 6/10, 8/10, 9/10. Then it is fair and tells something. Like finding the ultra short section of the music sample that was used for the successful 9/10 result.

I am still out, when I will be back, I will try to summarize results of all the tests from this thread.
 
Last edited:
It would be better if everone made 16 or 20 trials. I made 100 random tests with 10 trials and got 9/10 two times and 0/10 one time. So better use the standard 16 trials or 20 trials.

Like the group delay test here
www.audiosciencereview.com

Audibility of group delay at low frequencies

In the spectrogram tab You can change the appearance settings as well -- just change it to whatever method or resolution is needed to see things clearer. FDW doesn't affect the spectrogram at all, but it can clear up some of the reflection-induced rotations in seen in the phase traces. From...
www.audiosciencereview.com www.audiosciencereview.com
 
Thomas_A said:
It would be better if everone made 16 or 20 trials. I made 100 random tests with 10 trials and got 9/10 two times and 0/10 one time. So better use the standard 16 trials or 20 trials.
Click to expand...
This is absolutely correct. Quite tiring, so it can be separated into two consecutive tests, with some pause, and 9/10 + 7/10 would equal to 16/20, for example. Mentioned before, again.
 
You must log in or register to reply here.

Similar threads

pma
Fever FE-6J1-2 tube preamplifier kit review + listening test files
2 3 4
Replies
63
Views
4K
deanznz
D
E
MA1 setup help- Crashes HELP!
Replies
14
Views
846
sibi1865
S
W
How good is your hearing?
Replies
1
Views
375
Wavelength
W
Italian Job
Quad artera pop/thump
3 4 5
Replies
94
Views
4K
Italian Job
Italian Job
N
New Jico SAS stylus = crazy mistracking / sibilance
Replies
6
Views
1K
sqrl
sqrl
Back
Top Bottom