Thread Rating:
  • 0 Vote(s) - 0 Average
  • 1
  • 2
  • 3
  • 4
  • 5
Can DOS 2 Phantoms reproduce acoustic programme above 24kHz?
#31
(07-Feb-2021, 16:32)streamy Wrote: Let’s wait and see what @struts can record once he receives the UMIK-2.

[Image: c9ccf60b03524b275469be9040bf2061.jpg]

At last! After a few days of customs delay (it was shipped from Hong Kong) UMIK-2 is finally innadahouse!


Sent from my iPhone using Tapatalk
Reply
#32
(07-Feb-2021, 16:32)streamy Wrote: Let’s wait and see what @struts can record once he receives the UMIK-2.

So, I played the 24/192 versions of @streamy's test files via my Phantom Golds (which are on the 2.12.5 firmware) using Audirvana (UPnP) over Wifi and measured the output with the UMIK-2.  Unfortunately, as we suspected, there is no acoustic output above 24kHz.  The 1kHz, 20kHz and 22kHz tones all show very clearly (see below) but there is no trace of the 25kHz tone, either in the file where it is mixed with the 1kHz tone or on its own.  Unfortunately I don't have any hi res-capable device with an optical output that I can use to repeat this test in optical direct mode.  I will repeat this test with the 24/96 versions of these files when I receive them but don't expect any different outcome.  @streamy has offered to repeat my tests on his system with DOS 1 so we can finally get to the bottom of this.

   
Reply
#33
(03-Mar-2021, 17:45)struts Wrote:
(07-Feb-2021, 16:32)streamy Wrote: Let’s wait and see what @struts can record once he receives the UMIK-2.

So, I played the 24/192 versions of @streamy's test files via my Phantom Golds (which are on the 2.12.5 firmware) using Audirvana (UPnP) over Wifi and measured the output with the UMIK-2.  Unfortunately, as we suspected, there is no acoustic output above 24kHz.  The 1kHz, 20kHz and 22kHz tones all show very clearly but there is no trace of the 25kHz tone, either in the file where it is mixed with the 1kHz tone or on its own.  Unfortunately I don't have any hi res-capable device with an optical output that I can use to repeat this test in optical direct mode.  I will repeat this test with the 24/96 versions of these files when I receive them but don't expect any different outcome.  @streamy has offered to repeat my tests on his system with DOS 1 so we can finally get to the bottom of this.

Got this response from Devialet yesterday, fyi. In the DOS 2.12.3 release, Devialet has upgraded the audio pipeline from 16 to 24 bits, which provides the pipeline sound quality superior to all previously published DOS versions. We chose to limit the input sampling rate to 96 kHz to ensure stability and performance. Indeed, in some situations (depending on network conditions and the number of speakers), there may be stability issues with 176.4/192 kHz. Even though we do not limit the input file sampling rate on UPnP for compatibility purposes, we do recommend that our users limit playback to 96 kHz / 24 bits, if it is supported by their UPnP server. In any case, our operating system processes Roon and UPnP inputs in exactly the same way and may resample the original stream at the end of the signal path to ensure maximum stability.
Reply
#34
Very interesting.
This means that on DOS 1 any 24/192 sources where down sampled to 16/192.
And on DOS 2 everything above 24/96 will also be down sampled
Please excuse my poor reputation
Reply
#35
@daibaron I actually read the same from @howcho 's Devialet quote. So instead of 16bit in DOS 1 we now have 24bit in DOS 2, but at 24/48. If this is true I can understand why many report SQ improvement in DOS 2 vs DOS 1.
*
Devialetless!
Roon, ROCK/Audiolense XO/Music on NAS/EtherRegen/RoPieee/USPCB/ISORegen/USPCB/Sound Devices USBPre2/Tannoy GOLD 8
250 Pro CI, MicroRendu(1.4), Mutec MC-3+USB
Reply
#36
(04-Mar-2021, 11:12)ogs Wrote: @daibaron  I actually read the same from @howcho 's Devialet quote. So instead of 16bit in DOS 1 we now have 24bit in DOS 2, but at 24/48. If this is true I can understand why many report SQ improvement in DOS 2 vs DOS 1.

Yes, my interpretation aswell. Looking at MconnectHD that I use now for my Tidal Masters I can max get 24/48 and I guess I need a ROON core to get 24/96 which is the new max? Is it worth 1500 euros for a nucleus and then 500 euros for ROON subscription? I miss some stuff from really good recordings such as  some ECM stuff and Fleetwood Mac remasters, which I presume was 16/96 or even 16/192 through the DOS1 and NODE 2i optical? Pros and cons..... but it sounds better in general.
Reply
#37
(04-Mar-2021, 21:08)howcho Wrote:
(04-Mar-2021, 11:12)ogs Wrote: @daibaron  I actually read the same from @howcho 's Devialet quote. So instead of 16bit in DOS 1 we now have 24bit in DOS 2, but at 24/48. If this is true I can understand why many report SQ improvement in DOS 2 vs DOS 1.

Yes, my interpretation aswell. Looking at MconnectHD that I use now for my Tidal Masters I can max get 24/48 and I guess I need a ROON core to get 24/96 which is the new max? Is it worth 1500 euros for a nucleus and then 500 euros for ROON subscription? I miss some stuff from really good recordings such as  some ECM stuff and Fleetwood Mac remasters, which I presume was 16/96 or even 16/192 through the DOS1 and NODE 2i optical? Pros and cons..... but it sounds better in general.

You won't get 24/96 in Roon, it's downsampled to 24/48 by the speakers. 
   
Living room: Kii Three/BXT with Control.
Den: Tannoy Precision 8 iDP with TS112 iDP subwoofer.
In the cupboard, waiting for a sibling: 1st gen. Phantom Silver running DOS1
My Phantom Voyage
Reply
#38
(03-Mar-2021, 19:43)howcho Wrote:
(03-Mar-2021, 17:45)struts Wrote:
(07-Feb-2021, 16:32)streamy Wrote: Let’s wait and see what @struts can record once he receives the UMIK-2.

So, I played the 24/192 versions of @streamy's test files via my Phantom Golds (which are on the 2.12.5 firmware) using Audirvana (UPnP) over Wifi and measured the output with the UMIK-2.  Unfortunately, as we suspected, there is no acoustic output above 24kHz.  The 1kHz, 20kHz and 22kHz tones all show very clearly but there is no trace of the 25kHz tone, either in the file where it is mixed with the 1kHz tone or on its own.  Unfortunately I don't have any hi res-capable device with an optical output that I can use to repeat this test in optical direct mode.  I will repeat this test with the 24/96 versions of these files when I receive them but don't expect any different outcome.  @streamy has offered to repeat my tests on his system with DOS 1 so we can finally get to the bottom of this.

Got this response from Devialet yesterday, fyi. In the DOS 2.12.3 release, Devialet has upgraded the audio pipeline from 16 to 24 bits, which provides the pipeline sound quality superior to all previously published DOS versions. We chose to limit the input sampling rate to 96 kHz to ensure stability and performance. Indeed, in some situations (depending on network conditions and the number of speakers), there may be stability issues with 176.4/192 kHz. Even though we do not limit the input file sampling rate on UPnP for compatibility purposes, we do recommend that our users limit playback to 96 kHz / 24 bits, if it is supported by their UPnP server. In any case, our operating system processes Roon and UPnP inputs in exactly the same way and may resample the original stream at the end of the signal path to ensure maximum stability.

I would not overestimate what Devialet says currently. We have many communications, white papers, declarations from Devialet and its representatives where they talked about hires in all forms. Remember almost all in hires is 24bit. I’m quite skeptical regarding what they tell to questions of any of us. Their statements change way faster than our equipment. All can be possible, but why limit to 16bit when all their Linux SW components, a lot of it under different open source licensing models is made for up to 24/192k? Don’t think Devialet changed that to 16bit to as of 2021 to open up to 24bit? For me PR fuss!
Reply
#39
(04-Mar-2021, 22:30)ragwo Wrote:
(04-Mar-2021, 21:08)howcho Wrote:
(04-Mar-2021, 11:12)ogs Wrote: @daibaron  I actually read the same from @howcho 's Devialet quote. So instead of 16bit in DOS 1 we now have 24bit in DOS 2, but at 24/48. If this is true I can understand why many report SQ improvement in DOS 2 vs DOS 1.

Yes, my interpretation aswell. Looking at MconnectHD that I use now for my Tidal Masters I can max get 24/48 and I guess I need a ROON core to get 24/96 which is the new max? Is it worth 1500 euros for a nucleus and then 500 euros for ROON subscription? I miss some stuff from really good recordings such as  some ECM stuff and Fleetwood Mac remasters, which I presume was 16/96 or even 16/192 through the DOS1 and NODE 2i optical? Pros and cons..... but it sounds better in general.

You won't get 24/96 in Roon, it's downsampled to 24/48 by the speakers. 

Thanks @ragwo Then I will not go down that route
Reply
#40
@howcho You point out something very true.
What is the purpose of having high end solutions when the process is altered and outputted in a downgraded way.
Roon is now limited and not bit perfect anymore which is also off limit of the company philosophy.
Please excuse my poor reputation
Reply


Forum Jump:


Users browsing this thread: 1 Guest(s)