Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[Sashaaa]

So this is a bit of an odd issue I have been having. When playing back DSD files for some reason there is extra treble compared to when played back in other players. I always convert DSD to PCM in playback so I haven't tried doing native DSD playback. I don't know the exact numbers, but it sounds almost like playing a emphasized CD without emphasizing it except not as extreme. I can send two audio samples, one with a DSD file being played back from foobar2000 , and one with a DSD file being played back in MC, but all my DSD files are obviously copyrighted so I didn't want to send unless needed.

Here are some of the things I've tried:
Disabling all DSP
Converting to 88.2khz instead of 176.4khz
Enabling/Disabling Sox for the resampling
Restarting computer/media center
Upgrading to 29.0.66 (I don't remember which build I was on before but I installed it roughly a month ago)

Any ideas what's causing this? Could be a bug, I've only started using Media Center a couple months ago so I haven't used older versions. Let me know if there's anything I could try.

Relevant ystem specs:
M1 Max 16in Macbook Pro running macOS 12.4
Media Center 29.0.66 for Mac
RME Fireface UFX+ Audio Interface (for speakers)/RME Adi-2 Dac (for headphones)

[AGAWA]

MC has flat response and other players roll-off treble? Could be possible explanation.

[Sashaaa]

MC has flat response and other players roll-off treble? Could be possible explanation.

I don't really think that would be the case, as converting the track to pcm flac via dbpoweramp yields the same results as playing them back in other players I've tried, and I don't really think that dbpoweramp is rolling off treble starting at 13-15khz. Also I am comparing the dbpoweramp converted file to the dsd one within media center.

[AGAWA]

play SACD ISO without extracting - and see if it behaves the same way.

[blgentry]

If you audibly hear a treble rolloff with the other player, it's likely far below 13kHz.  Most adults have significant hearing loss at 15k and above.  Many can barely discern 13kHz.

If you've measured it with a tool, then I stand corrected.  But if you are relying up on your ears, you might be surprised.

I wish I had a good idea of what's going on here.  Clearly you hear something unique going on with MC compared to at least 2 other players you've mentioned.  Seems like it has to be something other than the filter frequency (which is above 20kHz).  I don't have any good theories to offer, but I wish I did.

Take care,
Brian.

[Sashaaa]

If you audibly hear a treble rolloff with the other player, it's likely far below 13kHz.  Most adults have significant hearing loss at 15k and above.  Many can barely discern 13kHz.

If you've measured it with a tool, then I stand corrected.  But if you are relying up on your ears, you might be surprised.

I wish I had a good idea of what's going on here.  Clearly you hear something unique going on with MC compared to at least 2 other players you've mentioned.  Seems like it has to be something other than the filter frequency (which is above 20kHz).  I don't have any good theories to offer, but I wish I did.

Take care,
Brian.

It could be below 13khz, I'm actually pretty young and can definitely hear up to 17khz, with it starting to roll off at around 18khz. Regardless though, the most important information is below that anyways. It's definitely audible, sometimes I do find that my ears can fool me but I don't think they are in this case.

Regardless though, I decided to run a double blind test just to be sure (with replaygain) and you can see the results below (I got it correct 16/16 times):

Code: [Select]

foo_abx 2.0.6d report
foobar2000 v1.6.9
2022-07-16 16:48:33

File A: dbp Subterranean DSD.wav
SHA1: 8bd87d8937b69e8c4ac2c373ed30033da4a14f8e
Gain adjustment: -0.26 dB
File B: MC Subterranean DSD.wav
SHA1: 833a05bab61de7f3ae57cd223e1b4150211dbaa7
Gain adjustment: -0.00 dB

Output:
WASAPI (event) : Speakers (High Definition Audio Device), 24-bit
Crossfading: NO

16:48:33 : Test started.
16:48:56 : 01/01
16:49:11 : 02/02
16:49:38 : 03/03
16:49:54 : 04/04
16:50:11 : 05/05
16:50:33 : 06/06
16:50:49 : 07/07
16:51:05 : 08/08
16:51:26 : 09/09
16:52:06 : 10/10
16:52:33 : 11/11
16:52:51 : 12/12
16:53:13 : 13/13
16:53:39 : 14/14
16:54:04 : 15/15
16:54:24 : 16/16
16:54:24 : Test finished.

 ----------
Total: 16/16
p-value: 0 (0%)

 -- signature --
6cff08b456dee473b93f1506fe68b7a18b7a2ca2

While I don't really have a way to measure the average frequency response throughout a file, I do have some tools to view spectrograms which may give a better idea of what I'm hearing. I took 2 screenshots of each file in izotope RX 9 which you can view here https://imgur.com/a/o25dn5h . Each file was cut to the same length, and roughly matched in level (replaygain only shows a difference of -.2db's), so while this isn't exactly a perfectly scientific comparison, I do believe it shows the difference that I'm hearing accurately enough. If needed I could also just upload the 30 second audio samples I'm using so that you can hear for yourself but I didn't know if that was allowed or not (the song is subterranean homesick blues by bob dylan).

play SACD ISO without extracting - and see if it behaves the same way.

Just tried this with the same track, and the issue persists. Was worth a shot though!

Thanks for the responses!
-Sasha

[Sashaaa]

So just to do some further troubleshooting I decided to create a macos virtual machine on my mac and installed media center 29.0.58 (forgot to go for the latest build). I'm mainly doing this to see if the issue is isolated to my machine or if it's something that is present on other macs . The issue is still definitely present in the vm and it has the same effect as playing back files in media center outside of the VM. I do have an older non apple silicon mac that I could try, but I would imagine it behaves the same way.