[rb2013]

  Burl would need to add a power supply to a custom Brooklyn card, correct? We're talking a standalone unit, not some kind of add-on? Hopefully that should still be doable <$500... I'm going to hold off on a getting a rednet until I hear more about a smaller option.

Yes it would be a simple board - no DAC or ADC function.  Just taking the BK2 card and feeding it DC power and outputing SPDIF and/or AES.
 
The new Dante Brooklyn II card has the processor built in:
https://www.audinate.com/products/manufacturer-products/dante-brooklyn-ii
 

A single Brooklyn II module provides a complete, ready-to-use Dante interface, and can equip a networked audio device with as many as 64 channels of bi-directional digital streaming. The onboard MicroBlaze CPU includes its own Linux environment, enabling the development of custom embedded applications for device and network control and monitoring. A wide range of interfaces is supported, and Audinate’s comprehensive software toolkit streamlines integration, dramatically lowering time to market. 

 

Power and Simplicity

With a powerful FPGA engine, plus on board micro-processor, Brooklyn II gives a comprehensive application environment to streamline software integration. Brooklyn II offers its own high-resolution, low jitter clocking, or connects to external clock, and supports industry standard sample rates of 44.1, 48, 88.2, 96, 176.4, and 192kHz.
Audinate’s acclaimed networking expertise is built into every Brooklyn II module, with a comprehensive toolkit that gives you a vast and varied array of tools, options, and possibilities. Its simple software development environment gives you maximum flexibility for custom functionality and control.

1. Dante plug and play media networking 
2. Powerful onboard microprocessor — no additional CPU required
3. Mini-PCI module and form factor — fits into virtually any design
4. Sub millisecond latency with +/- 1 microsecond synchronization 
5. Up to 64x64 redundant bi-directional channels @ 48kHz
6. Flexible Interface Support — SPI, I2C, RS232 and configurable GPIO
7. Powerful development environment and toolkit
8. Audinate’s patented Glitch-Free Redundancy 
9. Complete sample rate support
10. Support for 2,4,8, or 16 audio channels per TDM
11. The most interoperable networking technology: Works seamlessly with any other Dante-powered device 
12. Network health, clock and audio quality monitoring
13. Gigabit Ethernet switch chip support
14. Dante works with standard off-the-shelf network switches — easily integrates into existing network switch infrastructure, and automatically discovers other Dante-enabled devices on the network
15. Firmware upgradeable

The Dante API and accompanying Product Development Kit (PDK) allows you to create custom software that can be used for automatic routing, labeling of signals, custom control and monitoring.

Applications

1. Create Dante-enabled products quickly and cost-effectively 
2. Connect to any other Dante-enabled devices as part of larger systems
3. Customize Dante Brooklyn II to differentiate your products with special features and controls 
4. Perfect for use in DSPs, mixers, amplifiers, preamps, wireless microphones and more

Sample Rates	44.1 / 48/88.2 / 96 / 176.4 / 192kHz
Audio Channels In/Out (44.1/48kHz)	Up to 64x64 channels
Audio Channels In/Out (88.2/96kHz)	Up to 32x32 channels
Audio Channels In/Out (176.4/192kHz)	Up to 16x16 channels
Audio Flows In/Out	Up to 32x32 simultaneous streams
Digital Audio Format	TDM, I2S
Sample Bit Depth	16, 24 or 32 bits per sample
Audio Buffering	Up to 2000 samples per channel
Clocking	Onboard word clock or external word clock
Form Factor	Card edge module. 4.5cm x 6cm (1.75” x 2.4”)
FPGA	High performance Xilinx Spartan6 FPGA
Microprocessor	Soft-core Microblaze processor
Clock	High quality, low jitter onboard SiLabs
Ethernet	Standard RGMII/MII interface for Ethernet PHY or switch chip
Power	3.3VDC @ 2W max
Physical Connector	Mini-PCI
Control Interfaces	SPI Master and Slave; GPIO; I2C
Network	RGMII/MII

3.3VDC power 2W max.
 
Maybe @abartels could rig a RPi to act as an interface to the BK2 card for SPDIF out.

 

[ciphercomplete]

A shame I will need to buy a new 2-port NIC to try though.

Cheers

 
A 2 port isn't necessary if you have a extra 1 port laying around.  You can configure windows to use both the motherboard input and a 1 port nic card in tandem.  There are lots of guides for this online.  But, a 2 port nic is the easiest way.

 

[jabbr]

You mean 192k? 

No I mean 176.4kHz as that is a whole divider for the DSD sample rates, the 48kHz range with 192kHz isn't.

DOP uses a 24-bit word at 176.4 kHz which contains 16 bits data and 8 bits for a dsd-marker.

 

[rb2013]

Yes, give or take 0.5

But as I saw you coming with such a reply

I did a listening test specifically for it.
And the result was that I could not bear to listen to 96kHz on Foobar. Sorry if it sounds harsh, but i found it totally uninvolving and muddled, whereas Mqn plays with absolute clarity, great level of detail and very well balanced from deep bass to high frequencies. It is also very, very lifelike e.g. very well noticeable on percussion instruments like woodblocks or rims of drums where it gives a great sense of attack.

I must say I have optimised my OS by, amongst other things, syncing the NTTimerResolution and the size of the USB driver buffer for the F1 which I find greatly improves SQ.

It is a shame forum talk cannot really convey what one is actually hearing.

I might just try a Rednet D16 for a few weeks, as I found a dealer who has a 60-days return policy, as I am much attracted to the AOIP concept.
So maybe the proof has to come from the eating of the pudding. A shame I will need to buy a new 2-port NIC to try though.

Cheers

"And the result was that I could not bear to listen to 96kHz on Foobar."  - You need a better AOIP solution...

 
Give SoX a try at 192k - it's free.
 
But hey if you are married to MQN or HQplayer - or Roon - and abolutely hate ASIO - think it sounds like a AM radio in a fish tank - I can't help you.

 

[wushuliu]

Yes, give or take 0.5

But as I saw you coming with such a reply

I did a listening test specifically for it.
And the result was that I could not bear to listen to 96kHz on Foobar. Sorry if it sounds harsh, but i found it totally uninvolving and muddled, whereas Mqn plays with absolute clarity, great level of detail and very well balanced from deep bass to high frequencies. It is also very, very lifelike e.g. very well noticeable on percussion instruments like woodblocks or rims of drums where it gives a great sense of attack.

I must say I have optimised my OS by, amongst other things, syncing the NTTimerResolution and the size of the USB driver buffer for the F1 which I find greatly improves SQ.

It is a shame forum talk cannot really convey what one is actually hearing.

I might just try a Rednet D16 for a few weeks, as I found a dealer who has a 60-days return policy, as I am much attracted to the AOIP concept.
So maybe the proof has to come from the eating of the pudding. A shame I will need to buy a new 2-port NIC to try though.

Cheers

 
Have you tried the Xmos ASIO? For Jplay the Xmos ASIO is better than KS and sounds great (but not quite as good) for Foobar... I think the F-1 is optimized for the ASIO.

 

[rb2013]

Oh and before I see the post - no DANTE does not do DSD 512 or 784k PCM.  If that is a must - then this is not for you.

 

[wushuliu]

 
Yes it would be a simple board - no DAC or ADC function.  Just taking the BK2 card and feeding it DC power and outputing SPDIF and/or AES.
 
The new Dante Brooklyn II card has the processor built in:
https://www.audinate.com/products/manufacturer-products/dante-brooklyn-ii
 
 
3.3VDC power 2W max.
 
Maybe @abartels could rig a RPi to act as an interface to the BK2 card for SPDIF out.

 
Awesome, so a simple 5vdc and we're good to go...

 

[rb2013]

No I mean 176.4kHz as that is a whole divider for the DSD sample rates, the 48kHz range with 192kHz isn't.

DOP uses a 24-bit word at 176.4 kHz which contains 16 bits data and 8 bits for a dsd-marker.

Just upsample to 192k or get the $1600 D16.

 

[rb2013]

   
Have you tried the Xmos ASIO? For Jplay the Xmos ASIO is better than KS and sounds great (but not quite as good) for Foobar... I think the F-1 is optimized for the ASIO.

+1 on my systems. 

 

[jabbr]

"And the result was that I could not bear to listen to 96kHz on Foobar."  - You need a better AOIP solution...:wink_face:

Give SoX a try at 192k - it's free.

But hey if you are married to MQN or HQplayer - or Roon - and abolutely hate ASIO - think it sounds like a AM radio in a fish tank - I can't help you.

I think ASIO in itself does a decent job. The difference between ASIO and KS using the same player is not that big, maybe even a matter of preference.

No, with the development proces of Mqn I have come to realise that SQ greatly improves by making the computer do less and less. Mqn has come incredibly far with this concept. All kinds of extra processes on the PC actually makes the SQ worse. So for true SQ you have to let go of convenience items that activate the CPU.

You only have to try, Mqn is free and easy to use.
I have been through all these developments in digital audio for many years already, and I find only one tops the rest.

 

[rb2013]

First - have to the DANTE REDNET 3 is rock solid - ran all night with a hiccup or lock.  Not to surprising as this is studio level gear and the totl for Focusrite.  Studios can not put up with bauky USB drivers that lockup or crash.
 
So did some Foobar, DANTE DVS and REDNET Control optimizing last night - gained another 5% improvement in SQ:
 
So here is what I did - removed the Dither in Foobar Output window - dropped the buffer to 50ms.  Not even a trace of sweat!  With USB this low of a buffer would cause drop outs.
 
In the DVS dropped the Buffer Latency down to 5ms, increased the Encoding from 24 bits to 32 bits, and increased the Buffer size to 1024.
 
In REDNET Control - there were no Network events,  Primary Tx B/W: 20Mbps.

 

[wushuliu]

I think ASIO in itself does a decent job. The difference between ASIO and KS using the same player is not that big, maybe even a matter of preference.

No, with the development proces of Mqn I have come to realise that SQ greatly improves by making the computer do less and less. Mqn has come incredibly far with this concept. All kinds of extra processes on the PC actually makes the SQ worse. So for true SQ you have to let go of convenience items that activate the CPU.

You only have to try, Mqn is free and easy to use.
I have been through all these developments in digital audio for many years already, and I find only one tops the rest.

 
I have used MQN for a long time, and now have gone back to Jplay for convenience. MQN is *great*, but unless it’s taken some huge leap in the last year, it is not THAT good. Sounds more like you have spent of lot time dialing in your PC to maximize MQN and in order to try something else, may need to reconfigure. Foobar/Jplay with the F-1 Xmos ASIO should sound relatively grain-free, not harsh. I found Foobar to sound flatter in presentation, but not harsh or unlistenable.
 
IOW, if I had to choose between the F-1/Axmos ASIO/Foobar combo or MQN/KS/something else, I would choose the former. The F-1 does far more for sound quality than any of the software players. Especially with Fidelizer running.

 

[jabbr]

Have you tried the Xmos ASIO? For Jplay the Xmos ASIO is better than KS and sounds great (but not quite as good) for Foobar... I think the F-1 is optimized for the ASIO.

With the coming of JPlay 6 I find the sound much strained, not relaxed. Since then I have stepped away from JPlay.
XMoS ASIO doesn't differ from Thesycon ASIO as I believe the XMOS-driver is a 'small' version of the full Thesycon. I prefer the ability set the USB and ASIO buffer sizes in the full Thesycon driver set, also for the F1.

I don't think hardware favours one driver implementation over another. Maybe the default buffer sizes of the XMOS-driver set are better for ASIO than for USB, but as said I prefer the ability to set them myself.

 

[jabbr]

Just upsample to 192k or get the $1600 D16.

Over here the price difference between a Rednet D16 and a Rednet 3 is just under $200.

And upsampling from the 44.1-range into the 48-range is a very, very bad thing to do.

 

[rb2013]

I think ASIO in itself does a decent job. The difference between ASIO and KS using the same player is not that big, maybe even a matter of preference.

No, with the development proces of Mqn I have come to realise that SQ greatly improves by making the computer do less and less. Mqn has come incredibly far with this concept. All kinds of extra processes on the PC actually makes the SQ worse. So for true SQ you have to let go of convenience items that activate the CPU.

You only have to try, Mqn is free and easy to use.
I have been through all these developments in digital audio for many years already, and I find only one tops the rest.

I'm totally good with Foobar and SoX.  Besides if the SQ gets any better I may never leave my sound room:
 
Like this guy - remember the movie Brainstorm