Help me to better understand the digital terms I2S and Ethernet Renderer

Steve Williams

Site Founder, Site Owner, Administrator
I must live in the digital "stone age" as I am now reading about things about which I have no idea how they function or why we even need them. Reading about the newer DACs especially the MSB Select ll as recently acquired by Mike Lavigne and Priaptor, I have to admit that I am totally clueless about the following 2 things which I have been reading more and more of here and that is

1. Can someone give me a crash course on I2S and why and when it is used and why I think I am reading that it is sonically better than connecting USB

2. Ethernet renderers. Is ethernet connection hardwired the way to go and if so what does an ethernet renderer do

Over the past few years I have been listening to vinyl over 90% of the time but I know my digital side of things is showing it's age (although it still sounds pretty good IMHO) and to that extent I just added 2 new CMS Black Diamond filters to my existing rack in the anticipation of a new DAC and a server/streamer.

http://www.whatsbestforum.com/showt...Steve-Williams&p=454963&viewfull=1#post454963

So if there any of you out there who can give me and others a crash course on I2S as well as connecting ethernet and what the heck does an ethernet renderer do?

All of these seem to add one more layer of complexity in getting the digital files from the server to the DAC. So my question is, are these really necessary or just another means to an end as both seem to add not only more complexity but so also more cost
 

Mike Lavigne

Member Sponsor & WBF Founding Member
Apr 25, 2010
12,473
11,369
4,410
I must live in the digital "stone age" as I am now reading about things about which I have no idea how they function or why we even need them. Reading about the newer DACs especially the MSB Select ll as recently acquired by Mike Lavigne and Priaptor, I have to admit that I am totally clueless about the following 2 things which I have been reading more and more of here and that is

1. Can someone give me a crash course on I2S and why and when it is used and why I think I am reading that it is sonically better than connecting USB

i2S is a proprietary link developed by MSB to connect their transports to their dacs. and other manufacturers are using it for CD transports. as MSB uses it it satisfies the requirement for SACD to not be an open source, as it only can link with an MSB dac and pass the dsd stream from SACD. so it's just an interface, but used by different manufacturers in different ways.

it has nothing to do with USB really, it's more that at their best transports can sound great, but it's just not that simple and it's a fluid situation and we are not at any black and white end point where one way has won the battle. and, of course, higher resolution than redbook or dsd64 will sound quite a bit better than CD's and SACD's.....so there is that to consider.

2. Ethernet renderers. Is ethernet connection hardwired the way to go and if so what does an ethernet renderer do

as I explained to you in the PM a few days ago, and Renderer is a CPU inside a dac which allows a dac to be on a network. as opposed to either a USB or SPDIF interface which most dacs use and require a server of some sort.

Over the past few years I have been listening to vinyl over 90% of the time but I know my digital side of things is showing it's age (although it still sounds pretty good IMHO) and to that extent I just added 2 new CMS Black Diamond filters to my existing rack in the anticipation of a new DAC and a server/streamer.

http://www.whatsbestforum.com/showt...Steve-Williams&p=454963&viewfull=1#post454963

So if there any of you out there who can give me and others a crash course on I2S as well as connecting ethernet and what the heck does an ethernet renderer do?

All of these seem to add one more layer of complexity in getting the digital files from the server to the DAC. So my question is, are these really necessary or just another means to an end as both seem to add not only more complexity but so also more cost
 

wisnon

Well-Known Member
Dec 12, 2011
3,525
635
1,200
i2s is the internal communication protocol of Dac chips. It is originally only meant for very short interconnect distances, like a few inches.

Developers have piggybacked on it and extended the range with related protocols like LVDS..."low voltage distance signalling" forms of i2s and run that over HMDI cables, adapting the pins to the protocols. A popular standard is ithe PSAudio HDMI i2s. It can also be run over ethernet cables and other types like RJ45?
 
Last edited:

asiufy

Industry Expert/VIP Donor
Jul 8, 2011
3,711
723
1,200
San Diego, CA
almaaudio.com
As wisnon put it, I2S was an internal connection/bus, that was extended for use externally as well. MSB extended it further by called it "Pro I2S", adding, among other things, bi-directionality, so the DAC can enslave the transport and vice-versa. ProI2S can also carry clock signals along with the data.
I2S has no standard connector, and some manufacturers use RJ-45/CAT6 or CAT7 cabling/connections, while some other manufacturers use HDMI. Since they're not generally compatible among manufacturers (because most brands extend I2S beyond the original norm), don't expect to hook up two pieces from two manufacturers via I2S.
MSB uses ProI2S for their transports, because it provides the link they need to transfer hi-res content (SACD/Blue Ray), and also allows the DAC to enslave the transport to its (Femto) clock.

As for the Renderer, it's quite like Mike described. It's like miniaturizing a simple server and stuffing it inside the DAC, connected directly to the DAC "core", bypassing any and all external connections (SPDIF, AES, USB, even I2S). Being such "direct" connection has benefits, and that's why, more often than not, the Renderer on the MSBs are the best sounding inputs. That direct access to the DAC, and the fact that MSB's filters are optimized for 16/44 content, makes a beefy, upsampling server, irrelevant and unnecessary, as you can make do with any NAS on your network to do file serving/Roon Core duties.


cheers,
alex
 

wisnon

Well-Known Member
Dec 12, 2011
3,525
635
1,200
I guess you just connect the MSB to the NAS.

I also figure MSB connects the internal renderer to the Dac module by original i2S, ie short optimized distances.

The Sonore signature rendu uses LVDS over HDMI using the PSaudio standard pinout and so can feed i2s to about 10 dacs that designed for that standard, ie PSAudio, Rocna, Lampi, etc.

Ethernet proper has packetized data stream like USB, so they have their tradeoffs too.
 

Armsan

Well-Known Member
Jan 28, 2016
216
226
175
Portugal
Please take a look at http://pinkfaun.com/index.php/produ...ridge-st-stereo-uitvoering-d-a-converter.html
In a simplest way i2s is the language of all DACs. You may considered it as the lowest layer. I know this a very simplest view but the Renderers use the next up layer. The USB is one step up, sort of speak. So yes, from my experience, and in my systems along many years, the USB provides de worst SQ, then the Renderers and yes, i2s the best by far.
Cheers.
Armsan
 

asiufy

Industry Expert/VIP Donor
Jul 8, 2011
3,711
723
1,200
San Diego, CA
almaaudio.com
2? You only need 1. What I meant is that, on MSB DACs, out of all the inputs, the Renderer will usually sound the best. Not that you'd need multiple Renderer inputs :)

Prices are on the website. When you buy a DAC, you get a certain number of inputs included in price. So usually it'll be included in the price of the DAC, as we generally recommend people get the Renderer input when they purchase a DAC.

But if buying later on, it's $1995, for all 3 DACs (Analog, Reference and SELECT).

The Renderer module is currently available, and I've listened to it extensively in the SELECT and Analog DAC. What's forthcoming is the Renderer module that is compatible with MQA and Roon. In any case, if you already have a Renderer input on your SELECT, you'll be upgraded to the new one once it comes out. That's why I always say you don't need to lose money in digital, if you pick the right brand/products.


cheers,
alex
 

Ken Newton

Well-Known Member
Dec 11, 2012
243
2
95
Steve,

There are two primary considerations behind the various streaming data-source/transport box to DAC box interconnection schemes. The first consideration is jitter rejection. Jitter is produced both by the system clock generator, and also by the electrical properties of the interconnection. The other consideration is ease of computer file/network server connectivity.

The very best jitter rejection schemes locate the master clock generator in the DAC, and then slave the data-source to the DAC's master clock. This is opposite to how the familiar S/PDIF scheme works, which locates the master clock in the transport and slaves the DAC to the transport's clock. This subject area is called, flow control. Ideally, we want the DAC to have flow control. In that way, any jitter introduced by the interconnection between the data-source and DAC can be completely rejected once the data enters the DAC via what's called re-clocking. Re-clocking is an very simple step whereby the incoming data is re-aligned in time instant to the DAC's master clock.

Asynchronous USB, for example, enables the DAC to be clock master and the computer to be clock slave, which is it's claim to fame. However, USB is not necessarily an easy to use interconnection scheme, as driver installation is sometimes required. I don't know if there are any special driver software requirements for ethernet audio. Since I know next to nothing about about audio over ethernet, I won't attempt to speak on it.

As was mentioned earlier, I2S is a 3-wire interface bus created by Philips and intended for transferring digital audio data and clock signals between chips soldered on the same PCB. Basic I2S is comprised of a serial streaming data-wire, an bit-clock signal wire and a word(sample) clock signal wire. A few high-end vendors have taken I2S and extended it to connect streaming data-source boxes to DAC boxes. While this eliminates the jitter inducing process of seperating S/PDIF clock and data signals via PLL, it's not clear to me whether it also supports DAC-as-clock-master with data-source-as-slave type of flow control. If not, that would limit I2S as an interface for lowest posssible jitter.
 
Last edited:

bibo01

Member
Nov 26, 2013
201
1
16
Re. network renderer by MSB, the old/present model runs on as UPnP network client. One significant issue is that it bypasses the DSP engine. It means that employing bass management and EQ through JRMC is not possible.
The new/future model, by making it a Roon Ready device, allows the user to employ all the features of Roon, including convolution and other signal DSP.
 

sbo6

VIP/Donor
May 18, 2014
1,661
594
480
Round Rock, TX
The .pdf does a good job of explaining what a renderer is, "“Media renderer” is a term from the UPnP AV protocol which specifies how devices connected to a LAN can cooperate to play a media file stored on one device (a “media server”) on another device (a “media renderer”).

To clarify some of the above posts' explanations, a renderer isn't a CPU inside a DAC, it's typically a modular device that takes network - based input (music) and outputs via USB. An example is the microrendu and new ultrarendu which perform 3 tasks: 1) Accept data in via ethernet 2) Run an OS on a simple FPGA which 3) converts and reclock the input data to USB and outputs to a DAC.

As Wisnon said quite well, i2s is a communication protocol of DAC chips designed for very short distances (<6"). Companies like MSB have been adopting their own flavor extending past 6". I believe others have used it via HDMI and other connector interfaces.

As far as pros vs cons:
Pros:
Renderer is a dedicated, purpose built unit (opposite of a music PC/Mac designed to do many tasks not just music data - specific)
Renderer often includes common programs accepting input for DLNA, HQPlayer, ROON, etc. (e,g,: 1 stop shop for your favorite music interface)

Cons:
Requires a network accessible computer/device to provide music via ethernet
Extra cables, extra box(es) for additional PSU
 

Kal Rubinson

Well-Known Member
May 4, 2010
2,360
697
1,700
NYC
www.stereophile.com
Thanks Mike

I read that I2s has been around for over 20 years and developed by Philips semiconductor

Correct. It was developed by them as a standard for communications between audio chips/devices with a component. Its use for linking audio components was first implemented by Audio Alchemy. It was further developed by Sonic Frontiers and their partners as I2Se with additional lines/features. Its original appeal was that it maintained the clock separate from the data and, at the time, provided an improvement over S/PDIF.
 

JackD201

WBF Founding Member
Apr 20, 2010
12,308
1,425
1,820
Manila, Philippines
Correct. It was developed by them as a standard for communications between audio chips/devices with a component. Its use for linking audio components was first implemented by Audio Alchemy.

Yes that is my recollection. AA was first to take it out of the box.
 

About us

  • What’s Best Forum is THE forum for high end audio, product reviews, advice and sharing experiences on the best of everything else. This is THE place where audiophiles and audio companies discuss vintage, contemporary and new audio products, music servers, music streamers, computer audio, digital-to-analog converters, turntables, phono stages, cartridges, reel-to-reel tape machines, speakers, headphones and tube and solid-state amplification. Founded in 2010 What’s Best Forum invites intelligent and courteous people of all interests and backgrounds to describe and discuss the best of everything. From beginners to life-long hobbyists to industry professionals, we enjoy learning about new things and meeting new people, and participating in spirited debates.

Quick Navigation

User Menu

Steve Williams
Site Founder | Site Owner | Administrator
Ron Resnick
Site Co-Owner | Administrator
Julian (The Fixer)
Website Build | Marketing Managersing