Focusrite REDNet

[Aleg]

The Audinate Via application is also a nice bonus, as you can use it to stream pretty much anything from a Mac or PC. Don't know about the SQ.

 

There are very few viable alternatives to do that. It does require a wired ethernet connection, however, and that may be a problem for some.

 

There was someone, for example, asking about streaming Digital Concert Hall to the microRendu. It is really difficult... Via makes it possible to stream anything to a Dante device on the same network. Pretty cool...

 

Via is also limited to a maximum of 48kHz sample rate. Not very useful IMHO.

[Aleg]

I figured that's where I was heading, gldngate.

 

Thanks a lot for taking the time to try to help me out.

 

Joel

 

If I compare yours with gldgate's settings, it appears as if your RedNet settings are mapped onto the RedNet itself instead of your Tascam.

I would think you connected the incoming channels of RedNet to the outgoing channels of the RedNet.

I would clear the matrix and follow gldgate's example

[Aleg]

None, of the above. It's the "Return of the Trolls", again!

[Aleg]

Be very cautious with the Rednet / AES67 / Ravenna stuff. It's a different animal altogether and may introduce a host of other issue. I'm all for competing technologies, but what I've seen and used in this area has serious issues. Works fine in a studio, but home is another story.

 

 

Hi Chris

 

Can you enlighten us with your experiences in using RedNet / Ravenna / AES67 gear that made you give such a strong warning?

 

I'm very much interested in the technology but like to learn from experiences by others!

 

You can post your answer in this thread or the other you refered to, I follow both.

 

Cheers

[Aleg]

Currently, Windows or Mac only compatibility is a show stopper for me. Are there any plans for Linux compatibility ?

 

Lack of wi-fi compatibility for the Via application (which provides the ability to stream sound from a PC connected to the network, and looks fantastic on paper) is also a limiting factor.

 

This is Pro-Audio gear, not a DIY or consumer device.

As far as I'm aware Linux is a no-go area in Pro-Audio.

 

W.r.t. the Wifi issue:

"the practical limitations of current wireless technology (802.11a/b/g/n) render reliable performance unachievable. For this reason Dante software such as Virtual Soundcard will not recognize wireless connections for audio data."

[Aleg]

Today, yes, but perhaps this technology will be used someday for home audio.

 

Linux, by the way, is also used in pro audio.

 

I believe you about Linux, but never encountered it and find them Pro-people very Apple and Windows minded.

 

Audinate, the manufacturer of the Dante protocol and devices, says Wifi is not inherently impossible, just that the Dante protocol requires real time transport protocol, which cannot be supported reliably on Wifi.

The use of this rttp protocol is required for Dante to achieve clock synchronisation over the network, which again is required to achieve synchronicity of distributed devices over a network, as in recording studios with separated performance and recording rooms or in other large distributed audio systems.

 

When using AOIP in a consumer environment it will be mostly a single point-to-point connection, like the way the Rednet is used by the people here. I suspect the rttp is less of a requirement in those consumer settings, and a Dante-like protocol using rttp might not be strictly necessary for consumer use (though in multi-room applications I can imagine it might be) .

So who knows how an AOIP for consumer use will look like?

 

But at the moment no open/non-proprietary consumer-level protocol exists, while these Pro-Audio protocols do exist and have the AES67 interoperability protocol to enable mix-and-match of different devices and brands and have proven to be reliable and produce excellent sound when using quality devices.

I'm awaiting delivery of a Rednet myself, so I'm very curious to hear for myself how it compares to an optimised USB chain and optimised top quality software player.

[Aleg]

Hi Guys - Some people have asked about the issues I've experienced with AoIP, so I'll try to give a quick example.

 

I had a Ravenna based device for a while. When I first started using it my entire wireless network would stop working. The Ravenna device wasn't even on a wifi network and nothing in the data path was wifi, yet it brought down the wifi network. After several hours of research I figured out Ravenna was flooding the network with packets and I had to enable IGMP snooping on my wifi access points that had this option. I don't expect Joe Sixpack the audiophile to figure out why his wifi goes down when playing music over wired Ethernet. In addition, enabling IGMP snooping didn't fix the issue for another user I know.

 

I also had to replace a seemingly good gigabit Ethernet switch with a managed switch (without needing to change any settings) to get anything to work.

 

Things like this need to be worked out for the home environment if AoIP is going to be viable at home.

 

I really hope AoIP takes off for home audio. Another option is always a good thing.

 

Hi Chris

 

Thank you for your reply.

 

Yes AOIP has the possibility to flood your netwerk, but only when it is set to do Multicasting instead of Unicasting. And yes, when doing Multicasting it can bring down the available bandwidth of any attached Wifi-sections of the same subnet. But Multicasting isn't the default setting in the apps that I know, maybe it was in yours. Multicasting is only used in specific situations for large networks.

So the easy solution is to switch of Multicasting. Or, as you found out, if not possible to switch of configure thw IGMP snooping of your switches.

 

So I agree, when using AOIP networking overa LAN/WAN, it does require that you have some advanced knowledge of network managment (Duh ....., doesn't seem to unreasonable ) and it is not just your plug&pray solution.

BUT, as I said in a consumer environment people are using it in a single point-to-point connection, i.e. Computer-->CAT-cable-->RedNet. This doesn't require managed switches, doesn't introduce issues with Multicasting or IGMP snooping.

 

So there is no issue with the AOIP technology as you stated, it is a lack of sufficient knowledge with people dabbling with it.

If you look at the support section of the Audinate/Dante website they give all kinds of introductory videos, documents, FAQs, optimisation guidelines, troubleshooting tips.

It is really easy to obtain the knowledge required if you want to go into the Pro setup (which isn't required as I just said).

And Audinate/Dante also explicitly warns about network flooding when using Multicasting, tells you what it is/how it works and also tells you how to prevent it.

 

So I think at some point in time AOIP technology will come down to the level of 'ordinary Joe', but it is now used only in Pro-environments.

So I agree people should stick to the direct computer-->device setup for now, if they are not prepared/able to dig a little deeper in networking technology (which really isn't all that difficult as they show on the Audinate/Dante website) when using larger setups.

 

But the advantage of the current Dante/Ravenna/AES67 protocols is the interoperability it offers, as opposed to the closed/proprietary developments in current consumer applications.

[Aleg]

Ouch, yeah Dante/AES67 expects you to do QoS ... my 2p

....

 

Not true if you use direct computer-->device connection. See my previous post and posts by others explaining their setup.

 

No, need to frighten people.

But yes, this is a Pro-audio device capable of being used in complex audio networks. But it can be used in simple setup as well.

[Aleg]

...

 

Overkill for home audio. Aside from the networking issues (which can be dealt with) is the lack of a universal i.e. universally compatible plug and play virtual soundcard app which would connect any software to any DAC.

 

That and lack of Ethernet input DACs.

 

....

 

Dante comes with the Dante Virtual SoundCard which is ASIO and WDM capable. wDM with some limitations, but use the Virtual ASIO cable and you're back in business.

So you can use any software with the Dante DVS.

 

The Rednet devices output SPDIF in several connector, so you can connect any SPDIF Coax or AES capable DAC. No need to dabble with USB anymore.

No need for ethernet DACs either, though I think they will come soon enough

[Aleg]

Many things can be used in complex networks but for our purposes here there is a need to work in simple networks as well so I stand by my comments. There are specific situations where the complexity is needed -- really not for the current generation of home audio products nor users. Moving back to point to pint connections is a poor step backwards which I don't encourage.

 

If you use a Rednet D16, it has 2 RJ45 ports.

Connect one to your computer and one to your LAN. No difficult switch configurations required.

 

If you use a Rednet 3, it only has 1 RJ45 port.

Get a second NIC on your PC so your PC has two ethernet ports, and connect one to the Rednet 3 and the second one to the LAN. No need to use difficult switch configurations either.

 

So yes, if you want to get into the pro-level network setup, yes you will have to do pro-level network management as well. But it can be easily avoided by going another route.

And BTW, many cheap (semi-managed) SMB switches offer QoS settings already (not sure if they are using 4 queues though) and cost only little more than unmanaged switches.

[Aleg]

I've been reading this and related threads with curiosity but also some puzzlement. I've been somewhat familiar with RTP since the early days of IP telecom (which I did some work with back then), and with internet protocols and their Ethernet implementations pretty much since their start. I can't understand the benefits of AES67 over UPnP/DLNA for single-ended consumer applications. In both cases, packet payloads need to be decoded, buffered, and rendered as accurately clocked bit streams by some device for presentation to a DAC. The protocol delivering the packets to the rendering device just needs to take care that packets are not dropped or delayed beyond what the device's buffer can cope with. With those provisos, SQ will only depend on the timing and electrical properties of the connection from the device to the DAC. So, why do we need AES67 for (high-end) consumer use? It's of course possible that pro devices like those from RedNet are just better renderers (better clocks, better electrical properties) than consumer alternatives, but that would not have anything to do with the choice of Ethernet protocol.

 

I'm not familiar with the details of the DLNA/UPNP protocol, apart from bad experiences when using it and I kept far away from it eversince, but others with more knowledge about it condemn it to burn in hell for ever.

 

I can imagine that the content of what is transfered in the IP-packages requires a different kind of rendering at the receiver end. We already know from experience that more CPU activity at the renderer that is connected to the DAC, is bad for sound quality. So that could also be an explanation apart from the possibilities you already raised.

[Aleg]

Here you find a tutorial for setting up a switch for RedNet.

 

This might probably work for Ravenna too.

 

In the latest release notes of the Ravenna Virtual Sound Card it says it only does MultiCast at the moment and UniCast will be added at a later point. This is probably what Chris Connacker used.

This then also requires that IGMP-snooping is configured to prevent drowning the network.

Yes, the manual for the Dante configuration can be used for QoS if you have managed switches that allow for that. The best manual is from audinate itself which explains how to setup IGMP snooping from pg 58 onwards in this manual: https://www.audinate.com/sites/default/files/PDF/advanced-dante-networking-avnw-2015-audinate.pdf

 

But all this IGMP snooping and QoS configuration is not needed for AOIP in a home setting when using Dante RedNets and keeping to the default Unicast settings.

 

Ravenna has still some work to do on their virtual sound card.

It would be best if the Dante DVS would act as a proper AES67 device, it could then be used for any AOIP that was AES67 compliant. Unfortunately Audinate has no plans to make the DVS a proper AES67 device. There is a possibility that Audinate will upgrade their PCIe cards to proper AES67 devices, which could then be used to front any other AES67 device.

 

Cheers

[Aleg]

Merging has two versions of the Ravenna Virtual Sound Card (well more than that actually), but mainly one for the professional side, still in multicast, with much more streams supported, and a consumer one, now operating in unicast (or multicast optionally).

 

There are definitely plans for releasing the unicast support in the professional Virtual Sound Card at some point in the future, however Merging is currently addressing other priorities.

 

AES67 is one of them, Merging released at the beginning of the year a free AES67 Virtual Sound Card that will allow connecting to Dante as soon as their AES67 support will be released, however only at 44.1/48 kHz for the time being. But Merging is following this thread very closely.

 

On thing that is important to notice though, is that Ravenna is the only open protocol supporting sampling rates up to DXD/384kHz, and more important DSD. The currently available Ravenna Virtual Sound Cards, both using multicast for the professional world and unicast for the consumer world do support DSD up to 11.2 (for now), as standard in the protocol, without cheating with DoP. These VSC are available for both Mac and Windows and soon for Linux.

 

AoIP is a complexe beast with multiple heads, that exists for decades on thousands of flavors, however Merging is committed to release usable technology, products and drivers, based on open protocols, allowing for multiple manufacturers to be OPENLY INTEROPERABLE at ANY RESOLUTION on VARIOUS MARKETS.

 

So, Aleg, thank you very much for your very important post, and you're perfectly right Ravenna has still some work to do on their Virtual Sound Card, but this work is actually being done, and very actively. We're progressing step by step, adding the necessary features for each market and applications as they're required, but the motto is simple, networked high resolution for all, and we're not that far.

 

Cheers,

 

 

Hi dbrulhart

 

Thank you for your postings that gives us more insight in the Ravenna products from Merging. Very interesting developments indeed.

Is it correct that the "Merging RAVENNA/AES67 Virtual Audio Device - STANDARD Installer" is only available for OSX, or am I overlooking the proper link for Windows?

I immediately applied for the temporary download access of course, but fail to see the Windows version of the Virtual Audio Device.

 

Thanks for your help.

[Aleg]

..... will allow connecting to Dante as soon as their AES67 support will be released, .....

 

Hi dbrulhart

 

As far as I know (and I have it first hand from Audinate) the AES67 support has already been available since January 2016 for Dante devices based on their Brooklyn II card, so there shouldn't (have to) be any issues on that front.

 

Cheers

[Aleg]

...

You discussed a bit about Merging's intentions for the multi-platform Ravenna Virtual Sound Card software and drivers which they spent a lot of time developing, and which to-date only operates with their hardware.

 

You wrote "Merging is committed to release usable technology, products and drivers, based on open protocols, allowing for multiple manufacturers to be OPENLY INTEROPERABLE at ANY RESOLUTION on VARIOUS MARKETS." Are you saying that Merging is going to begin licensing their VSC and drivers to other manufacturers, including those who might OEM newer AES67 modules (such as Covelez BACH)?

 

....

 

Hi Alex

 

W.r.t the Merging Virtual Sound Card, the Merging website states the existence (and actually allows download of) their software.

The Merging Technologies RAVENNA/AES67 STANDARD Virtual Audio Device edition is free of charge and intended for owners of any RAVENNA and/or AES67 multi-cast device.

 

It has (as of yet) limitations that would prevent wide-spread use, (such as support for 44 & 48 kHz only, but there exists a Premium version for use with the Merging HORUS and HAPI, so who knows what can be negotiated). But I can find only the OSX version, though IIUC there should also be a Windows version in existence and Linux version under development.

 

Comparisson of Standard vs Premium version of Merging Sound Card software:

Mac OS X: RAVENNA/AES67¹ Virtual Audio Device Specifications:

 

PREMIUM edition: Bundled free for owners of a Merging Network Interface

 

STANDARD edition: Free with any RAVENNA/AES67 Hardware compatible device. Downloadable here

 

I would also consider this a giant leap if their software would become generally available or licenseable for other parties with AES67 products.

 

I hope to hear more from dbrulhart about the Merging SOund Card s.w

[Aleg]

Audinate does not recommend using EEE:

 

https://www.audinate.com/resources/faqs then filter switches.

 

So? I don't understand the point you are making?

 

EEE reduces responsiveness of the switch required for the Real Time AOIP.

 

Audinate has several requirements for the switches to be used in pro-audio AOIP settings (but not everybody uses Pro-Audio level networks):

 

What is the minimum requirement for switches in a Dante network?

 

All Ethernet switches are capable of working with Dante. However, please be aware that there are some features on some kinds of switches that will allow you to build larger and more reliable Dante networks.

 

While Gigabit switches are recommended, 100Mbps switches may be used in limited scenarios.

 

For channel counts of 32 or more, Gigabit switches are essential. QoS is required when using Dante in networks that have 100Mbps devices. QoS is also recommended for Gigabit switches on networks that share data with services other than Dante.

For lower channel count (<32) applications, a 100Mbps switch may be used as long as it supports proper QoS, and QoS is active. The use of 100Mbps switches without QoS is not recommended or supported.

What features are important when purchasing a switch?

 

Dante makes use of standard Voice over IP (VoIP) Quality of Service (QoS) switch features, to prioritize clock sync and audio traffic over other network traffic. VoIP QoS features are available in a variety of inexpensive and enterprise Ethernet switches. Any switches with the following features should be appropriate for use with Dante:

 

Gigabit ports for inter-switch connections

Quality of Service (QoS) with 4 queues

Diffserv (DSCP) QoS, with strict priority

A managed switch is also recommended, to provide detailed information about the operation of each network link: port speed, error counters, bandwidth used, etc.

https://www.audinate.com/resources/networks-switches

[Aleg]

Hi Aleg: Thanks for pointing that out. Of course OS X only and 44/48 only won't yet attract other DAC manufacturers, but perhaps Merging will eventually go further. I am just not sure why they would give away their software IP, seeing as they are the ones who developed it and will have to keep it up to date through OS evolutions.

 

.....

 

My feeling is that Merging, being one of the first partners the designers of Ravenna turned to, is probably the one party among all partners, that is best suitable to develop and maintain this kind of software.

 

They may have 'donated' a scaled down version to help along the adoption of the Ravenna protocol, that critically depends on the availability Virtual Sound Card software. They (the Ravenna consortium) have to develop and generally release/license software that allows access to all the features and high-end capabilities (352/384 & DSD), otherwise they will not create enough of a difference to get alongside Dante.

 

They promote Ravenna by emphasising these capabilities over the current limit of 192kHz in Dante (though I believe that 192kHz limit only to be a designers choice and not an inherent limitation of the protocol), so Ravenna will have to deliver on that and in an easy accessible way.

 

I think this year will show interesting developments in the AOIP world, though my feeling is that in the professional audio world things don't happen as fast as in the world of consumer high-end audio.

 

cheers

[Aleg]

Slim Devices had products on the market for years before that. I was enjoying streaming over LAN and wifi using the Squeezebox more than 10 years ago. They had the audiophile Transporter back in 2006. Way ahead of the times they were. I am just glad it is coming back around.

 

Yes, nowadays you see several solutions like that and like what JPlay has developed (the dual PC setup) a few years ago.

The Dante/Ravenna/AES67 protocols are far more robust however, they have to be to be of any value in the pro-audio world. I also think that renderers that have been built for the purpose (unlike the JPlay and NAA type solutions that use general purpose computers) have a far better chance of creating exceptional sound quality than those formentioned solutions using general purpose computers as renderer.

 

We will have to wait, see and listen, but the reported exceptional quality of the two Rednet devices makes it very promising to me.

[Aleg]

....

Regarding "AES in to the Mutec (from the Rednet) and then USB out" I believe that the Mutec only has USB input.

 

 

No, definitely Mutec has USB input AND output. It is a very special feature of Mutec compared to other DDC devices.

[Aleg]

Hi Aleg:

Can you please clarify that? The Mutec has only one USB jack--a 'B' type, for connection to a computer. There is no indication (I skimmed the manual) that it can act as a host to output directly to a DAC via USB. Some rather different circuitry (and s/w) would be required for it to output to a DAC. Don't know why one would want to do that anyway, but I do want to be sure I am understanding the capabilities of the flexible Mutec.

Thanks,

--Alex C.

Hi Alex

 

Of course. Taken from the Mutec MC3+USB product description:

Audio streams received via USB will always be converted to five audio outputs allowing multiple devices to receive the same audio simultaneously. Upon selecting one of the audio inputs, the signal will be converted into a USB audio stream to be sent to the computer.

 

I see now one could read my previous reply as havin simultaneous input and output. That is not what I meant. I tried to say what is in the quote from Mutec. So it is "either or", but USB is bi-directional.

 

Cheers

[Aleg]

...

On that note, what do folks think about why DLNA managed to get such an early hold in home network audio? Was it lack of alternatives? Was it better coordination or OS s/w support by the consortium? Was it because the the big CE and TV companies embraced it? I honestly don't know because I have pretty much eschewed UPnP/DLNA the whole time. But I think we AOIP enthusiasts (and AES67 supporters) might do well to look to the past successes and failures of DLNA to chart a course.

...

 

I think this was just the first network application in A/V, so lack of alternative and all CE manufacturers just put it in their appliances.

 

From a user perspective, it was just there and allowed network access in a fairly simple way. Simple device discovery. Simple 'unicast' flows. Big hurdle was finding combinations of server, renderer and controller software that actually worked.

On the software side, being just for playback and not recording purposes, things can be made a bit simpler than for Pro-audio, like automatic channel mapping, automatic following of sample rates.

Interesting times to see how AOIP develops into the consumer market.

[Aleg]

It's perfectly clear that the so called "Merging RAVENNA/AES67 Virtual Audio Device - STANDARD Installer" has been released to allow anybody interested in trying AES67 to be able to do so for free and easily. We limited this version of the driver to the exact AES67 specifications, being 44.1/48kHz, 48 samples framesize, etc... purposely to fit this specification and NOT make it unusable in certain conditions or markets.

 

....

 

Hi Dominique

 

Thank you for your introduction and detailed explanation of the VSC.

 

I don't have access to the literal text of the AES67 standard, so maybe you could clarify a question that came up in my head.

 

Is the AES67 standard limited to the sample rate and frame size you mentioned (because that is the common denominator between all current protocols), or does it specify the minimum requirement that needs to be supported (but can be higher as well when both sides support those higher level specs)?

 

Just curious how it is meant to function.

 

Cheers

[Aleg]

So I picked a Rednet 3 up to try with my Out of Your Head software, since that software requires a single sampling rate output. I'm hoping the Rednet adds sampling rate support as promised anyway.

 

I see that there is an AES connector (DB25) possible. Why are people choosing the D16 over the RN3 if this is the case?

 

Just my 2 cents:

- D16 is based on their later Brooklyn II board which supports the 176.4kHz sample rate and also is the (only) board by Dante which supports the AES67 interoperability protocol.

- D16 has two ethernet ports instead of only 1 with RN3. Two ports allow for easier setups in consumer environments if people want to avoid using (and configuring of managed) switches.

- D16 has a 1U height instead of the 2U height from RN3.

- RN3 has a lot of optical input but those are only ADAT capable and not SPDIF, so not of much use to the consumer user.

 

cheers

[Aleg]

Thanks. I understand RN3 also supports Dante? And would I need optical input if I'm using Ethernet in?

 

RN3 is Dante.

No you wouldn't, but some wanted to use optical output which didn't work fo them because it is ADAT and not SPDIF

[Aleg]

Thanks, I understand now.

 

Would AES67 confer any additional advantages over what's used in the RN3?

 

Not in a direct setup like PC-->RN3-->DAC.

AES only comes into play when you get other non-Dante devices in your network that you would want to stream to your Dante device.

Not likely to happen very soon, but I like to keep my options open