T+a dac 8 dsd

[Miska]

5 hours ago, guiltyboxswapper said:

Hmmm that almost appears anti-competitive.  The newer XMOS ASICs clearly have much more headroom than the Amanero designs...

 

XMOS has it's own problems... Then there are the ones used by Japanese manufacturers which is yet another implementation. So far I've been most impressed by RME's FPGA based UAC2 implementation.

[bibo01]

14 hours ago, blue2 said:

What do you mean exactly? Are you referring to the processing capability of the Amanero board, as Linux can be run on a variety of hardware?

TIA

I was referring to processing capability  of Amanero board.

[blue2]

20 hours ago, OE333 said:

 

Complexity in hardware very often brings other unwanted side effects (e.g. jitter). In many respects the Amanero is a very fine solution. With a Windows source it performs just perfect.

 

I keep my fingers crossed that this now can also be achieved for Linux....

Not quite perfect - there's the problem reported by @louisxiawei

[OE333]

1 hour ago, blue2 said:

Not quite perfect - there's the problem reported by @louisxiawei

 

This problem only appears with very few DACs and there is a solution for it.

If you encounter this problem with your DAC, please contact the T+A service dept. or your T+A dealer.

 

[twluke]

19 hours ago, Miska said:

For a little while, such existed. There were big- and little-endian versions of the firmware. Driver still detects and supports both. The newest firmwares are big-endian only.

 

Since the introduction of firmware (actually a series of 200xbe) supporting native DSD of BE format on linux by Amanero, I've been quite accustomed to see a warning message in dmesg like below:

[    1.292138] usb 1-10: new high-speed USB device number 3 using xhci_hcd
[    1.440637] usb 1-10: config 1 has an invalid interface number: 3 but max is 2
[    1.440638] usb 1-10: config 1 has an invalid interface number: 3 but max is 2
[    1.440639] usb 1-10: config 1 has no interface number 0
[    1.441153] usb 1-10: New USB device found, idVendor=16d0, idProduct=071a
[    1.441154] usb 1-10: New USB device strings: Mfr=1, Product=2, SerialNumber=3
[    1.441155] usb 1-10: Product: Combo384 Amanero
[    1.441156] usb 1-10: Manufacturer: Amanero Technologies
[    1.441156] usb 1-10: SerialNumber: 413-001
[    2.689600] usb 1-10: Warning! Unlikely big volume range (=32767), cval->res is probably wrong.
[    2.689602] usb 1-10: [10] FU [PCM Playback Volume] ch = 2, val = -32767/0/1
[    2.691127] usb 1-10: Warning! Unlikely big volume range (=32767), cval->res is probably wrong.
[    2.691129] usb 1-10: [10] FU [PCM Playback Volume] ch = 1, val = -32767/0/1

This is part of dmesg from a linux kernel (4.15.0-34-generic, Ubuntu 18.04) running on an i7 6700K PC, connected to a DAC 8 DSD with 2005be Amanero firmware, playing DSD512 via HQP/Roon. I'm not sure this warning can be relevant to the current noise issue but my impression is that it can be a sign of firmware with certain bugs.

 

Here is another example of firmware with no warning on dmesg:

[    1.288020] usb 1-10: new high-speed USB device number 3 using xhci_hcd
[    1.437251] usb 1-10: New USB device found, idVendor=16d0, idProduct=071a
[    1.437252] usb 1-10: New USB device strings: Mfr=1, Product=2, SerialNumber=3
[    1.437253] usb 1-10: Product: Combo384 Amanero
[    1.437254] usb 1-10: Manufacturer: Amanero Technologies
[    1.437254] usb 1-10: SerialNumber: 413-001

One might notice that there is no complaints about invalid interface or volume range. This comes when the DAC 8 DSD is flashed to 1099c firmware which supports native DSD of LE format up to DSD256. The sound quality is excellent and beautiful.

 

I'll be happy if Miska can explain the background of this warning message.

 

TIA

[Miska]

1 hour ago, twluke said:

Since the introduction of firmware (actually a series of 200xbe) supporting native DSD of BE format on linux by Amanero, I've been quite accustomed to see a warning message in dmesg like below:

[    1.292138] usb 1-10: new high-speed USB device number 3 using xhci_hcd
[    1.440637] usb 1-10: config 1 has an invalid interface number: 3 but max is 2
[    1.440638] usb 1-10: config 1 has an invalid interface number: 3 but max is 2
[    1.440639] usb 1-10: config 1 has no interface number 0
[    1.441153] usb 1-10: New USB device found, idVendor=16d0, idProduct=071a
[    1.441154] usb 1-10: New USB device strings: Mfr=1, Product=2, SerialNumber=3
[    1.441155] usb 1-10: Product: Combo384 Amanero
[    1.441156] usb 1-10: Manufacturer: Amanero Technologies
[    1.441156] usb 1-10: SerialNumber: 413-001
[    2.689600] usb 1-10: Warning! Unlikely big volume range (=32767), cval->res is probably wrong.
[    2.689602] usb 1-10: [10] FU [PCM Playback Volume] ch = 2, val = -32767/0/1
[    2.691127] usb 1-10: Warning! Unlikely big volume range (=32767), cval->res is probably wrong.
[    2.691129] usb 1-10: [10] FU [PCM Playback Volume] ch = 1, val = -32767/0/1

This is part of dmesg from a linux kernel (4.15.0-34-generic, Ubuntu 18.04) running on an i7 6700K PC, connected to a DAC 8 DSD with 2005be Amanero firmware, playing DSD512 via HQP/Roon. I'm not sure this warning can be relevant to the current noise issue but my impression is that it can be a sign of firmware with certain bugs.

 

This is a bug in UAC2 "mixer" controls in the firmware, similar can be seen with some other (XMOS based) DACs too. Usually it doesn't cause any functional issues and if it does, only in volume control. It should be fixed though, but it is not critical.

 

[twluke]

29 minutes ago, Miska said:

 

This is a bug in UAC2 "mixer" controls in the firmware, similar can be seen with some other (XMOS based) DACs too. Usually it doesn't cause any functional issues and if it does, only in volume control. It should be fixed though, but it is not critical.

 

Understood. Thank you for taking your time.

[blue2]

On ‎9‎/‎19‎/‎2018 at 11:16 AM, OE333 said:

 

This problem only appears with very few DACs and there is a solution for it.

If you encounter this problem with your DAC, please contact the T+A service dept. or your T+A dealer. 

 

Thanks for all your responses on here. Good to know only some devices are affected. Can you explain your comment a little more? Does this mean installing a different firmware or some hardware repair?

Thanks

[twluke]

Hi there. This is a small addition to my previous post.

 

On 9/9/2018 at 1:39 AM, twluke said:

-snip-

Among them, the TA_DAC8_ServiceTool, a kind of prefixed serial communication tool specially designed for DAC 8 DSD was, I would say, of a great help for fool-proof proceeding of updating work. No other serial application is required: only to prepare a set of RS232 cable and plug is sufficient for it.

-snip-

 

Shown below is an example of cable setting necessary for updating the DAC firmware to V 2.70.

A combination of USB to RS232 (DB9) serial adapter cable connected to RJ11 (or RJ12) cable with the pin assignment indicated by the PDF manual from T+A. Hope this will be of some help for those who might be interested in the DAC firmware update.

 

[blue2]

@OE333At CEDIA 2018 T+A apparently displayed updates to the MP 8 Multi-Source Player and DAC 8 DSD which it is reported now supports DSD512???

Can you please clarify this statement? What is actually different between the new and previous versions? Can we have our DAC 8 DSD's upgraded and if so what will the cost be?

TIA

[firedog]

14 minutes ago, blue2 said:

@OE333At CEDIA 2018 T+A apparently displayed updates to the MP 8 Multi-Source Player and DAC 8 DSD which it is reported now supports DSD512???

Can you please clarify this statement? What is actually different between the new and previous versions? Can we have our DAC 8 DSD's upgraded and if so what will the cost be?

TIA

I think they are talking about with a Windows computer, not Linux. So not anything really new or revolutionary. 

[blue2]

14 minutes ago, firedog said:

I think they are talking about with a Windows computer, not Linux. So not anything really new or revolutionary. 

Let's not guess and hear from the horses mouth  They have announced it as an update so something must have changed you would have to think! Hopefully Linux support for DSD512!

[OE333]

On 9/22/2018 at 10:47 AM, blue2 said:

Thanks for all your responses on here. Good to know only some devices are affected. Can you explain your comment a little more? Does this mean installing a different firmware or some hardware repair? 

Thanks

 

Our findings were that the "faint pop" does not occur on all systems. It is not a hardware issue and it depends on quite a few things:

1.) the PC system (and maybe other software installed on it) - miska has already shed some light on possible causes

2.) the USB receiver firmware - we did not observe the pops with older firmware and do not observe it with the latest firmware version

3.) pops can also be caused by transmission errors - USB audio uses isochronous transmission and there is no way of error correction - this is by the way independent of the specific hardware (Amanero/XMOS) and also independent of the OS (Win/Linux).

So, if the pop problem occurs or not - this really depends on many factors...

By far the most systems we have checked did not show this problem.

 

[OE333]

6 hours ago, blue2 said:

@OE333At CEDIA 2018 T+A apparently displayed updates to the MP 8 Multi-Source Player and DAC 8 DSD which it is reported now supports DSD512???

Can you please clarify this statement? What is actually different between the new and previous versions? Can we have our DAC 8 DSD's upgraded and if so what will the cost be?

TIA

 

Currently DSD512 is supported from Windows sources only..

DSD512 / Linux is still under development.

So no change, yet.

 

[mav52]

On 9/23/2018 at 12:14 PM, OE333 said:

 

Currently DSD512 is supported from Windows sources only..

DSD512 / Linux is still under development.

So no change, yet.

 

So those updates noted here  https://parttimeaudiophile.com/2018/...-roon-and-mqa/
Are for the MP 8 only.  Nothing planned for the DSD DAC 8 ?

[juanitox]

hi, 

 

since

On 9/23/2018 at 5:10 PM, OE333 said:

 

Our findings were that the "faint pop" does not occur on all systems. It is not a hardware issue and it depends on quite a few things:

1.) the PC system (and maybe other software installed on it) - miska has already shed some light on possible causes

2.) the USB receiver firmware - we did not observe the pops with older firmware and do not observe it with the latest firmware version

3.) pops can also be caused by transmission errors - USB audio uses isochronous transmission and there is no way of error correction - this is by the way independent of the specific hardware (Amanero/XMOS) and also independent of the OS (Win/Linux).

So, if the pop problem occurs or not - this really depends on many factors...

By far the most systems we have checked did not show this problem.

 

 

 

hi ,    since few days i have a problem with my  dac8 DSD.      i'v got a loud white noise sort of fireworks  comming from the speaker .i use windows 10 and HQplayer and this random noise only appears with the DSD dac never in PCM .    it can happen 2 or 3 times in an hour .

do you think my dac is deffective or can it be a driver / softwxare problem ?

[gdpr]

Hi Juanitox

 

It is definitely a software problem, not necessarily a driver problem.  

What I mean is that it can be due to the used combination of filter and  dithering.

 

What's your configuration of HQPlayer for DSD ?

 

Dirk

[Miska]

31 minutes ago, juanitox said:

hi ,    since few days i have a problem with my  dac8 DSD.      i'v got a loud white noise sort of fireworks  comming from the speaker .i use windows 10 and HQplayer and this random noise only appears with the DSD dac never in PCM .    it can happen 2 or 3 times in an hour .

 

This kind of thing usually happens due to USB transfer errors. I'd recommend to check your cable first and try with a different cable. Preferably ordinary USB HiSpeed certified one.

 

[sbenyo]

I also hear white noise in some tracks in DSD only. I use DSD512 with ploy-sinc-xtr filter. I also have ISO REGEN and PPA2 usb card. It starts randomly in the middle of the track. It's not loud but very noticeable.

 

It does not happen in PCM.

 

I found that if I run Fidelizer first and then HQPlayer, the sound is usually clean but I did not test it enough to be sure this completely resolves the problem.

 

[juanitox]

thanks Miska and Dirk ,     my config is a decated PC  ( LPS supply) running windows 10  roon/HQP DSD512 /  INTONA  LPS supply  T+A DSD8  

 

there is so much variable to test , but what i find strange is that the crash only happen with the dac playing PCM TO DSD or DSD never with the PCM DAc.

i will check usb cables /  INTONA  test first..

[juanitox]

57 minutes ago, sbenyo said:

I also hear white noise in some tracks in DSD only. I use DSD512 with ploy-sinc-xtr filter. I also have ISO REGEN and PPA2 usb card. It starts randomly in the middle of the track. It's not loud but very noticeable.

 

It does not happen in PCM.

 

I found that if I run Fidelizer first and then HQPlayer, the sound is usually clean but I did not test it enough to be sure this completely resolves the problem.

 

 

 i think for me it's a different thing it is not a low white noise but a real fireworks  like my speaker was in fire ?

[sbenyo]

2 minutes ago, juanitox said:

 

 i think for me it's a different thing it is not a low white noise but a real fireworks  like my speaker was in fire ?

 

What you describe is extreme and never happened to me. Something with the setup is probably wrong if this happens.

It might be the same root cause that in your case just sounds much worse.

 

Does it happen all the time or just with specific tracks randomly?

I do recommend that you try Fidelizer or something similar to reduce any OS load/limitations to see if it helps.

[juanitox]

Quote

 

Does it happen all the time or just with specific tracks randomly?

 

randomly,  can happen 3 times in a day during one minute  than  it stop alone.   some days i don't have problem .

[sbenyo]

Which filter are you using? Does it happen on the same tracks or completely random ones?

 

You may want to test lighter filters. This may give you a clue it might be related to OS load/optimization.

You should also use Fidelizer. I tried it again and it seems to be removing the white noise I have it specific tracks. I suspect may be related to OS/USB optimizations/load.

 

[Miska]

3 hours ago, juanitox said:

thanks Miska and Dirk ,     my config is a decated PC  ( LPS supply) running windows 10  roon/HQP DSD512 /  INTONA  LPS supply  T+A DSD8  

 

there is so much variable to test , but what i find strange is that the crash only happen with the dac playing PCM TO DSD or DSD never with the PCM DAc.

i will check usb cables /  INTONA  test first..

 

You could go with simplest config, just plain certified USB cable and no Intona, straight from computer to the DAC. That's how I have it. Then you have at least another reference to compare against.

 

I haven't got that effect myself with DAC8 DSD, but with some other DACs I've seen all kinds of effects.

 

At PCM , even when playing at 384 kHz, the data rate over USB is still half lower than what goes through at DSD512. For comparison you could also try for example at DSD128 and see if you still get same behavior or not.

 

Since the DAC8 DSD ASIO driver has fixed buffer size, the buffer's length in time is at lowest when running at DSD512 - just little over 1 millisecond... And exactly lowest at 48k x512 DSD rate where things easily begin to fall apart... Since CPU load at least doubles every time you double the DSD rate, and buffer's length in time halves every time you double the DSD rate, you pretty quickly come to challenging combination. So if you run for example at DSD256 you have at most half the CPU load and double the buffer time...