Sample: 15 seconds of track 4 from Radiohead's 'In Rainbows'.
Detail: static-like sounds accompanying bass instrument (between 0:03 and 0:05, also between 0:08 and 0:10) easily distinguishable.
Tested: Opus 1.1 @ 96, 128, 160 and 199kbps.
Interesting report.
Lothario, have You performed ABX (http://www.hydrogenaud.io/forums/index.php?showtopic=16295) test on this sample? Even if an issue is obvious it's still good to have a log of performed test.
Also this sample is very loud (as all modern CDs) and maybe the issue comes from there (clipping?). Dunno.
P.S. Ouch, now have seen "Sample easily ABX-ed at high bitrates". Well, anyway log would be nice
Hm, I can't hear what you mean on my system, these sound the same to me. Tried going down with the bitrate, but all I hear is the cymbals getting smeary.
Please post a log. Then we can investigate further.
I reproduced the finding. Here is my log.
foo_abx 2.0 report
foobar2000 v1.3.6
2015-01-12 03:26:13
File A: 1-04_WeirdFishes_sample.flac
SHA1: e7663211690ce2eb0d7e70cdc4b2a8582af5d1b4
File B: 1-04_WeirdFishes_sample_199.opus
SHA1: 7ca517bdbf34d7f68eea7d99a2149e4ebc7d9501
Output:
DS : プライマリ サウンド ドライバー
Crossfading: NO
03:26:13 : Test started.
03:26:55 : 01/01
03:27:18 : 02/02
03:27:41 : 03/03
03:27:56 : 04/04
03:28:31 : 05/05
03:28:52 : 06/06
03:29:13 : 07/07
03:29:25 : 08/08
03:29:50 : 09/09
03:30:01 : 10/10
03:30:16 : 11/11
03:30:46 : 12/12
03:31:03 : 13/13
03:31:35 : 14/14
03:31:57 : 15/15
03:31:57 : Test finished.
----------
Total: 15/15
Probability that you were guessing: 0.0%
-- signature --
e0c6eb8f7f237c81a539dd8da8fddb0490fd4994
I have to say that the difference is subtle; the buzzer like tone B sound from right, 0:04 is more buzzy and fuzzy in the Opus sample.
Apologies, I tested at a variety of rates (96, 128, 160 and 199) so didn't keep the original logs to post. Fortunately, I had the same equipment with me this evening (Sony MDR-7506 headphones plugged directly into my ThinkPad X220 work laptop) so was able to re-do the 199kbps test.
As before, the lossless file has a static-like sound in the right channel between 0:03 and 0:05 and, to a slightly lesser extent, between 0:08 and 0:10. This sound is diminished in the Opus file.
Oddly, I scored 10/10 first time but failed a single instance this time around. Perhaps I should have refrained from using the term 'easily' in the sub-heading. Anyway, here's the log of tonight's test:
foo_abx 1.3.4 report
foobar2000 v1.3.6
2015/01/11 19:38:17
File A: C:\Users\ThinkPad\Downloads\1-04_WeirdFishes_sample.flac
File B: C:\Users\ThinkPad\Downloads\1-04_WeirdFishes_sample_199.opus
19:38:17 : Test started.
19:38:56 : 01/01 50.0%
19:39:10 : 02/02 25.0%
19:40:00 : 03/03 12.5%
19:40:55 : 04/04 6.3%
19:42:34 : 05/05 3.1%
19:43:40 : 06/06 1.6%
19:44:35 : 07/07 0.8%
19:46:01 : 08/08 0.4%
19:48:00 : 09/09 0.2%
19:52:04 : 09/10 1.1%
19:52:40 : 10/11 0.6%
19:53:07 : 11/12 0.3%
19:53:46 : 12/13 0.2%
19:54:52 : 13/14 0.1%
19:55:32 : 14/15 0.0%
19:56:23 : 15/16 0.0%
19:56:25 : Test finished.
----------
Total: 15/16 (0.0%)
Here's my quick attempt, using speakers:
foo_abx 2.0 report
foobar2000 v1.3.6
2015-01-11 22:38:44
File A: 1-04_WeirdFishes_sample.flac
SHA1: e7663211690ce2eb0d7e70cdc4b2a8582af5d1b4
File B: Weird Fishes-Arpeggi.opus
SHA1: a4f9457598d2851f377c0c9aa5c8665cf394d438
Output:
DS : Primary Sound Driver
Crossfading: NO
22:38:44 : Test started.
22:39:10 : 01/01
22:39:24 : 02/02
22:39:59 : 03/03
22:40:28 : 03/04
22:40:54 : 03/05
22:41:32 : 04/06
22:41:47 : 05/07
22:42:06 : 06/08
22:42:16 : 07/09
22:42:41 : 08/10
22:43:56 : 09/11
22:44:04 : 10/12
22:44:15 : 11/13
22:44:20 : 12/14
22:44:25 : 13/15
22:44:49 : 14/16
22:44:49 : Test finished.
----------
Total: 14/16
Probability that you were guessing: 0.2%
-- signature --
767f88ae4ab963ccb2227a2ea3fd05381a5cddcb
I can't say exactly what it is, the original just sounds somewhat louder and clearer. Though when I compare them with ReplayGain, they show the same track gain and a higher peak in the Opus version.
BTW, I made my own Opus (199k) from the Flac, uploaded here.