1st day brand new user of Audirvana. Got everything up and running fine. Running Audirvana on a Windows 11 PC and using a Cambridge Audio CXN100 as my DAC. When I go to set the DSD upsampling I want to use Audirvana gives me a choice in the dropdown of only DSD64 and DSD128. I want to upsample to DSD512. My CXN100 is capable of native DSD512. How do I fix this in Audirvana? Thanks in advance for any help.
You need to ensure you have an ASIO driver installed and playback Audirvana using ASIO. This will facilitate native DSD playback to your Cambridge up to the specified native DSD capability. WASAPI and Kernal streaming only allows DSD over DoP.
I have a CA USB Audio ASIO Driver installed and I have that selected as my Output device. But then when I go to the select the “Forced upsampling type”, DSD is not even a selection. I go back and select UPnP as my Output device and then DSD shows up as a choice in the Forced upsampling, but only as high as DSD128. I am sure I don’t have something configured properly. Perhaps a different driver is needed?
Welcome…
The CXN100 platform design is built around the ESS ES9028Q2M chipset, which has no un-fettered 1-bit PDM (DSD) signal path to output… The chipset decimates all DSD signals to multi-bit PCM for output… I suggest that you up-sample all PCM files using ‘Power of Two’ as this will up-sample those files to the DAC maximum sample-rates of 705.6kHz and 768kHz.
1 Like
Thank you for your reply. However, I do know the CXN100 is capable of native DSD512. I have confirmed it with CA Tech Support, I have upsampled to DSD512 using other SDM modulators successfully and I have played DSD512 music files on a USB flash drive on my CXN100. And the CXN100 displays DSD512 in all of those cases. I have also confirmed the architecture of the ESS ES9028Q2M chip set uses HyperStream 32 BIT 768kHz and a supports native DSD 512. But I will also try upsampling for “Power of Two” as you suggest just to see if I hear the difference. Thank you for your suggestion.
Of course the chipset will accept 1-bit PDM signals, however the 1-bit PDM signal is decimated to multi-bit PCM for output… You are misinterpreting the information or being mislead…
3 Likes
I assume you have the Cambridge USB Audio 2.0 driver installed as well and it is set up correctly? From the manual this seems a bit of a process.
I have read a couple of CXN100/DSD posts on the Roon forum with responses from support which confirms @Agoldenear statements re decimation of the input DSD signal to PCM so the question would be what are you gaining, other than conducting additional upsampling/decimation steps? I would personally stick to PCM upsampling in this case.
2 Likes
I stand corrected. Something for me to follow up on with CA.
1 Like
Nothing to corroborate… They state it here in the CXN100 overview rhetoric regarding the D/A conversion:
1 Like
An example of DAC topology misunderstanding vs upsampling strategy…
1 Like
To be fair, the manufacturers are not exactly open with advertising. ‘Native’ DSD support for a lot of manufactures just means the DAC will accept the file then decimate it. For people like me who have limited knowledge of audio engineering it seems to be designed to mislead. Thanks to the various more knowledgeable posters on this forum I now know I need to raid my loose change jars to see if I can get a T+A HA200 headphone amp if I want to truly determine if my ears prefer PCM or DSD files…. Not sure the loose change jars will cut it😳
1 Like
Get it all in this platform…
and if you want to take the HPA to another level… make a stack…
My recommendation for a less expensive high-performance stack with these components… The Cirrus Logic chipset in the D70 Pro Octo DAC handles DSD properly…
https://www.toppingaudio.com/product-item/d70-pro-octo
https://www.toppingaudio.com/product-item/a70-pro
2 Likes
Here are the upsampling settings (and reasons for them) recommended for ESS chips with HQPlayer by its developer. Note he recommends a 5th order modulator, which would be the B5 modulator in Audirvana.
Regarding the trouble you’re having finding the right settings: Yes, let me have a look at Audirvana’s settings for Windows in a few hours (the PC is busy with something in one of my Linux OSs until then), or I’ll check the remote and see if I can find a way for you to get there.
1 Like
Using the remote app, see whether you can find the following:
Settings screen - Go all the way to the bottom and tap DAC Input. This should bring you to the DAC Input screen.
DAC Input screen - A little way down on this screen you should see Maximum PCM Rate. Tap the arrow and select 768kHz from the next screen that appears. This takes you back to the DAC Input screen.
Back at the DAC Input screen, select the appropriate PCM bit depth, and a little further down the Maximum DSD Rate, which should now allow you to pick DSD512 (or DSD256 if preferred).
Hope this helps. If it doesn’t work, let me know what’s happening when you try.
Are you recommending modulation to DSD? If so, the signal is being decimated in the Hyperstream architecture… Why inject more potential for data corruption?
The Hyperstream architecture is similar to Chord’s DAVE
1 Like
Jud, thank you for helping me out. Got it resolved. Have a great day!
1 Like
The developer of HQPlayer, who is very familiar with the behavior of most DAC chips, including ESS chips, is recommending it, and I was simply passing along his advice. He gives a very brief summary of his reasons in the comment I linked.
There is just simple logic involved here, especially when there is clear understanding of the Hyperstream architecture and how it handles PCM and decimates DSD… There will be no more contextual musical information created in the modulation process to DSD… So just let Hyperstream do its thing and not complicate the process and increase the potential for data corruption…
A better service would be to recommend configuration parameters for r8Brain or SoX when up-sampling using ‘Power of Two’.
1 Like
@Tigerman57 @Djm1960 @Jud
It took me awhile to find my .PDF copy of this presentation… But this will be better…
ESS
“Noise-Shaping-Sigma-Delta DACs”
1 Like
Yes, excellent presentation. I used to enjoy the seminars at RMAF. Sad to see it go.