Hi folks,
I’ll receive my Trinnov Nova next week and this gear has a Dant audio interface.
It is also possible to install a Dante Audio virtual card in Windows to allow the system to send audio over IP from the computer to the Nova.
I think that could work with Audirvana.
Someone here has an experience with Dante Audio virtual card and audirvana?
++
Food for thought… you don’t need Dante protocol to network the Trinnov box
Configuration and Setup and Networking explained in the manual…
Dante is limited to 192kHz… Its doubtful that Dante protocol support will ever be integrated into Audirvāna, if not just for the licensing fees… Merging Technologies Ravenna protocol and the Zman interface will be the better system as it supports PCM sample-rates to 384kHz and DSD256.
This question is best put to Trinnov Audio.
Hi,
- I’ve tested the Nova during 3 weeks and you are right, the Nova can be managed from Trinnov interface without Dante protocole.
- My request is not to integrate Dante protocole into audirvana.
- My goal is just to use a virtual sound card from audirvana. As this virtual sound card is not free I hope that someone has already tested it with Audirvana.
https://www.getdante.com/products/software-essentials/dante-virtual-soundcard/
I understand the nature of the Dante virtual sound card…
Note the bold text from the Virtual Soundcard User guide:
The latest version of Dante Virtual Soundcard supports the Core Audio (Mac OS X), Steinberg ASIO (Windows), and WDM (Windows) audio interfaces, and can be used with any supporting audio application. Once you install Dante Virtual Soundcard on a computer and connect it to the Dante network, you can:
Start and stop the application
Lock and unlock the application
Change the device latency (the minimum latency a receiving device should apply before playout)
Select the number of audio channels available
Select your preferred Ethernet network interface
Change the audio sample rate, audio encoding bit depth and device name (using Dante Controller)
https://dev.audinate.com/GA/dvs/userguide/pdf/latest/AUD-MAN-DanteVirtualSoundcard_4.5.x-v1.0.pdf
It is not as simple as you would like to believe, and using Audirvāna for a Dante audio-distribution network is a compromised use scenario, except for the inherent QoS guarantee of the peer to peer protocol…
Again, it is highly doubtful that Audirvāna will support Dante protocol.
I read the DVS documentation and at this time I don’t see any blocking point to work with Audirvana using ASIO.
I suggest that you contact Audinate in regard to application support…
- I found a free 30 days trial version of Dante Virtual Soundcard and I’ve started my trial today.
- I started DVS with these following configuration: Asio and 2x2 channels.
- Audirvana is able to enumerate my virtual sound card and send PCM flow to it.
- Dante controller app show my DVS source and I must route it to the Nova receiver.
- Waitting for receiving the Trinnov Nova to continue…
I’m curious… Why do you need the DVS with Audirvāna and why add another layer of software in the transmission of the digital-audio signal? 
1 Like
A software layer is mandatory between Audirvana and the audio stream like UPnP, ASIO, USB driver
- My goal is to connect Audirvana to my Trinnov Nova.
- The Trinnov Nova doesn’t have an USB interface, only optical and coax digital input.
- I want to keep my PC far of the Nova.
If the Dante solution doesn’t satisfy my needs for quality and stability, I’d look into using another gear via USB.
A driver is mandatory in your case (ASIO), however DVS is another layer of unnecessary operational overhead, running on top of ASIO…
You are investing in the Trinnov system for some reason beyond the average… I presume so, in the spirit of achieving the best possible performance from your playback system listening environment.
This streamer will complement your application of the Nova system that many Audirvāna users implement for network UPnP playback:
1 Like
Thank for your answer. If I understand, you suggest using a new hardware device with its cables to avoid a software layer that doesn’t alter the audio stream, I think it is not efficient.
Also, If I understand well, ASIO is not a software layer but an interface to allow communication between the application and the driver and in some case, from the application to the hardware bypassing the kernel. Here, Audirvana is the application and DVS the driver.
And I’m not very confidante with UPnp, I have opened several cases on this forum about UPnP issues without a solution from Audirvana or the brand:
- White noise when playing a radio with a 3DLab via UPnP
- Stopping playback with Lumin U2 via UPnP - #17 by schoeppi
- I also have an Eversolo where some issues occur but I have not yet open a subject here.
I want to test Dante out of curiosity and to identify some limitations. If the quality is there and there are no bugs, Dante could be used on my system.
I’m going to ask some technical questions and I’m looking forward to reading your answers, which will be very useful to me.
I received the Nova yesterday and was able to play audio from Audirvana to the Nova via Dante protocol.
So far I haven’t had any problems and the quality is very similar to UPnP. If I switch to the next track, the track switch without latency.
In my point of view, the main limitation of the Dante protocol is that the frequency must be defined before playing audio and must be the same between audio interfaces which want communicates together.
So I’ve chosen a frequency of 96kHZ for the Dante virtual sound card and the Nova, and I’ve used R8Brain to fix the mismatch between the source and Dante frequency.
I know that R8Brain is a very nice algorithm, I’m confident of the quality of the conversion from 48kHz to 96kHz, but I wonder if the conversion from 44.1kHz / 88.2kHz to 96kHz is a good thing!
44.1 is fine. 88.2 of course is not something you would do for the quality purposes for which upsampling is ordinarily done, but it should not have negative effects.
You are adding the extra layer of DVS processing overhead, in concert with Audirvāna system level operations and CPU cycles and System RAM… In this scenario you are interfering with the efficient flow of the digital-audio signal produced by the Audirvāna audio-engine architecture, to the output bus, so that Audirvāna ‘bit perfect’ output quality becomes moot… I agree that Dante protocol provides a solid level of transmission QoS due to the peer to peer nature of the protocol as compared to UPnP.
Using the ‘Power of Two’ up-sampling strategy, will produce the proper logical result for the Nova…
Example:
44.1kHz → 176.4kHz
48kHz → 192kHz
88.2kHz → 176.4kHz
96kHz → 192kHz
OR… Create a ‘Custom’ sample-rate conversion strategy
Example:
44.1kHz → 176.4kHz
48kHz → 192kHz
88.2kHz → 176.4kHz
96kHz → 192kHz
176.4kHz → 176.4kHz
192kHz → 192kHz
352.8kHz → 176.4kHz
384kHz → 192kHz
705.6kHz → 176.4kHz
768kHz → 192kHz
I don’t know how a custom up-sampling strategy will affect the playback of 1-bit DSD files when the ‘DSD streaming method’ is set to ‘None, convert to PCM’.
Hi @Agoldnear, you’re right about the extra consumption of DVS compared with a direct USB connection to a DAC. On the other hand, I don’t think this is a problem with a sufficiently powerful PC.
I may have misspoken with my story about oversampling (perhaps because of my poor English).
The Dante network must be set to a fixed sampling frequency and this cannot be changed once configured.
The solution I found with R8Brain is a custom configuration like:
- 44.1kHz → 96kHz
- 48kHz → 96kHz
- 88.2kHz → 96kHz
- 96kHz → 96kHz
- 176.4kHz → 96kHz
- …
I wonder if non-multiple frequency oversampling is a good thing, for example 88.2kHz to 96kHz!
Translated with DeepL.com (free version)h
You are forcing conversions to illogical 96kHz results of the fundamental sample-rate when converting some files.
If you pick 176.4kHz and create a custom strategy, at least all of your 44.1kHz, 88.2kHz, 176.4kHz and 352.8kHz files will play correctly… If you have DSD files that will get converted to PCM, these will get decimated correctly to 176.4kHz…
You are creating an ugly playback scenario… Typically a Dante system is working up to 192khz when recording live, it is not flexible for the scenario you are applying it to, unless your music library is composed of 48kHz, 96kHz or 192kHz files… You may get a more stable network transmission, but at a price in hassle-factor and ultimate sound-quality, as compared to Zen Stream
Any computer today can do those multiplications easily without been buggy
1 Like
It’s not about the calculations… it’s the results of the calculations…
[44.14kHz x 2.17489805 Fs = 96kHz]
[44.14kHz x 2 Fs = 88.28]
The extra samples created in the illogical 44.14kHz → 96kHz conversion math result are superfluous and create noise in the anti-aliasing filter at the Nyquist Frequency (Fs).
The 44.14kHz signal is converted to 96kHz by inserting zeros (0’s) to increase the sample-rate…
In the case of illogical down-conversion to 96kHz, truncation distortion is created in the decimation process.
1 Like
Though we cannot access it from the US currently, this Canadian site shows graphs of downsampling results from many different software products. You’ll see that many of them, perhaps the majority, including R8Brain used in Audirvana, produce essentially flawless results.
1 Like
Yes… But not comparatively… The real assessment is in the quality of the perceived output we audition… You can’t hear a spectral plot… 
It’s the non-linearities that aren’t measured that make the differences tangible.
(Edit:) Note that the SRC down-conversion results spectral plots in the Infinitewave database are the product of logical ‘in-family’ decimations.