the last mile — sonore opticalRendu

aKnyght

aKnyght

Active member
Joined
Jan 18, 2018
Messages
1,052
Location
saṃsāra
previously, my diy roon ROCK server was connected directly to my DAC via USB as described in this thread. however, with recent system changes it was necessary to move the server down to the basement networking closet. consequently, in order to keep feeding the DAC via USB which provides asynchronous and ultra high-resolution streams (DSD512) it was necessary to add a streamer located at the rack.

i opted for the sonore opticalRendu which eliminates 100% of the upstream electrical noise via optical isolation. it is a quite small roon-ready endpoint that is a bridge between the server and the DAC. it is connected to the network via ethernet optical fiber and to the DAC with a USB cable.

before, my roon ROCK server was connected to the network with optical fiber which also eliminated all upstream electrical noise. however, the signal subsequently passed to the DAC via USB unfortunately included some of the electrical noise generated by the server running all the roon software. now that noise is also completely eliminated and the digital stream entering the opticalRendu is 100% noise free.

absent a DAC with an on-board optical fiber input, it is not possible to get a cleaner asynchronous signal to the DAC. unfortunately, as of this writing and to my knowledge there are only a very few DACs offering fiber input (lumin X1 + P1, linn klimax DSM and all MSB models via its proISL input).

not to over-hype this solution but the results were immediately apparent and significant. versus my prior configuration: a huge soundstage with much more air around the instruments; increased instrument resonance / decay; and overall a richer and a way more holographic sound.

in any event, i have waxed poetic before about ethernet fiber and optical isolation. suffice it to say that i find the opticalRendu a relatively inexpensive method of eliminating all electrical noise from a audio stream at the most important point in the digital chain — just prior to delivery to the DAC. thereby, obviating the need for expensive audiophile networking equipment and/or specialized music servers.
___________________________

note: given that my server runs roon ROCK, i chose the roon-only version of the opticalRendu. i also got a small green computer LPS to power it. also, small green computer offers a 30-day no questions asked return policy. however, i knew it was staying in my system after 10 minutes.

;)
 
aKnyght said:
previously, my diy roon ROCK server was connected directly to my DAC via USB as described in this thread. however, with recent system changes it was necessary to move the server down to the basement networking closet. consequently, in order to keep feeding the DAC via USB which provides asynchronous and ultra high-resolution streams (DSD512) it was necessary to add a streamer located at the rack.

i opted for the sonore opticalRendu which eliminates 100% of the upstream electrical noise via optical isolation. it is a quite small roon-ready endpoint that is a bridge between the server and the DAC. it is connected to the network via ethernet optical fiber and to the DAC with a USB cable.

before, my roon ROCK server was connected to the network with optical fiber which also eliminated all upstream electrical noise. however, the signal subsequently passed to the DAC via USB unfortunately included some of the electrical noise generated by the server running all the roon software. now that noise is also completely eliminated and the digital stream entering the opticalRendu is 100% noise free.

absent a DAC with an on-board optical fiber input, it is not possible to get a cleaner asynchronous signal to the DAC. unfortunately, as of this writing and to my knowledge there are only a very few DACs offering fiber input (lumin X1 + P1, linn klimax DSM and all MSB models via its proISL input).

not to over-hype this solution but the results were immediately apparent and significant. versus my prior configuration: a huge soundstage with much more air around the instruments; increased instrument resonance / decay; and overall a richer and a way more holographic sound.

in any event, i have waxed poetic before about ethernet fiber and optical isolation. suffice it to say that i find the opticalRendu a relatively inexpensive method of eliminating all electrical noise from a audio stream at the most important point in the digital chain — just prior to delivery to the DAC. thereby, obviating the need for expensive audiophile networking equipment and/or specialized music servers.
___________________________

note: given that my server runs roon ROCK, i chose the roon-only version of the opticalRendu. i also got a small green computer LPS to power it. also, small green computer offers a 30-day no questions asked return policy. however, i knew it was staying in my system after 10 minutes.

;)
Click to expand...

This is really a great write-up, aKnyght, and describes well the benefits of using an optical system to connect server to network bridge.

Just some more detail of what aKnyght is referring to with respect to noise components: the power supplies of standard "networked" computing devices, e.g. NUCs, computers, generic Ethernet switches, etc., etc. that are all powered with SMPS create a class of current referred specifically to as "high-source leakage impedance current". This current adds noise to the signal ground plane of bridges, streamers, DACs, etc. and results in a very specific type of jitter known as "threshold jitter". This class of jitter, much like clock phase noise, results in timing errors that are quite audible as perceived by the human brain in our perception of "music" It's the brain that converts the sound of air waves of varying pressure into what we perceive as "music", and this is where these timing errors perceived by the brain can result in "disengagement".
 
aKnyght said:
previously, my diy roon ROCK server was connected directly to my DAC via USB as described in this thread. however, with recent system changes it was necessary to move the server down to the basement networking closet. consequently, in order to keep feeding the DAC via USB which provides asynchronous and ultra high-resolution streams (DSD512) it was necessary to add a streamer located at the rack.

i opted for the sonore opticalRendu which eliminates 100% of the upstream electrical noise via optical isolation. it is a quite small roon-ready endpoint that is a bridge between the server and the DAC. it is connected to the network via ethernet optical fiber and to the DAC with a USB cable.

before, my roon ROCK server was connected to the network with optical fiber which also eliminated all upstream electrical noise. however, the signal subsequently passed to the DAC via USB unfortunately included some of the electrical noise generated by the server running all the roon software. now that noise is also completely eliminated and the digital stream entering the opticalRendu is 100% noise free.

absent a DAC with an on-board optical fiber input, it is not possible to get a cleaner asynchronous signal to the DAC. unfortunately, as of this writing and to my knowledge there are only a very few DACs offering fiber input (lumin X1 + P1, linn klimax DSM and all MSB models via its proISL input).

not to over-hype this solution but the results were immediately apparent and significant. versus my prior configuration: a huge soundstage with much more air around the instruments; increased instrument resonance / decay; and overall a richer and a way more holographic sound.

in any event, i have waxed poetic before about ethernet fiber and optical isolation. suffice it to say that i find the opticalRendu a relatively inexpensive method of eliminating all electrical noise from a audio stream at the most important point in the digital chain — just prior to delivery to the DAC. thereby, obviating the need for expensive audiophile networking equipment and/or specialized music servers.
___________________________

note: given that my server runs roon ROCK, i chose the roon-only version of the opticalRendu. i also got a small green computer LPS to power it. also, small green computer offers a 30-day no questions asked return policy. however, i knew it was staying in my system after 10 minutes.

;)
Click to expand...

I also run Roon using the Sonore opticalRendu with an opticalModule in front of it (both on SGC linear power supplies). I have had a very similar experience. Very satisfied with the sound, particularly for the total investment.
 
have you tried comparing the ultrarendu vs the optical rendu? My ultrarendu also sounds great, I use for that a special build Bach LPS.

This optical solution sounds interesting, I need to cross 10 metres of distance with a network cable. Light does not have Electromagnetic distortion at all. On the other hand does a well shielded network cable also block electromagnetic distortion? Because in the latter case a signal from copper does not have to be converted to light and as a rule of thumb: the less conversions the better.

On the other hand, audio is sometimes difficult to understand, especially when a network comes into play. Why does an Intel Nuc as Roon server build in a fanless aluminium case sound better then the same board in the original case?
 
Alkyogre said:
have you tried comparing the ultrarendu vs the optical rendu? My ultrarendu also sounds great, I use for that a special build Bach LPS.

This optical solution sounds interesting, I need to cross 10 metres of distance with a network cable. Light does not have Electromagnetic distortion at all. On the other hand does a well shielded network cable also block electromagnetic distortion? Because in the latter case a signal from copper does not have to be converted to light and as a rule of thumb: the less conversions the better.

On the other hand, audio is sometimes difficult to understand, especially when a network comes into play. Why does an Intel Nuc as Roon server build in a fanless aluminium case sound better then the same board in the original case?
Click to expand...

I don't know what you mean by "electromagnetic distortion".

Optical cables are not susceptible to RFI and EMI.

A network cable, if it is a shielded one with shields connected at both ends, will allow the passage of high-source leakage impedance current. This is not EMI. The impact of high-source leakage impedance current is threshold jitter, which causes timing errors. An optical cable will not pass high-source leakage impedance current, which means no threshold jitter.

Both types of cable are susceptible to clock phase noise from the el-cheapo clocks in generic networking devices.
 
My DAC/PRE takes copper Ethernet connection. I opted to use the etherRegen unit to collect network connection via optical connection, and use the highly regulated Ethernet 100 connection to the Bricasti. I also consider having a non-galvanic connection close to the Ethernet endpoint to be useful.
 
Great write up. We tried the optical connection on the X1 and felt it was just “different”, not sure it was better. Cleaner, more detail. But more musical? Don’t know.

The MSB ProUSB/ProISL is our favorite of all MSB inputs - by far.

Optical certainly eliminates a lot of problems and I hope more companies adopt it.


Sent from my iPad using Tapatalk Pro
 
Puma Cat said:
I don't know what you mean by "electromagnetic distortion".

Optical cables are not susceptible to RFI and EMI.

A network cable, if it is a shielded one with shields connected at both ends, will allow the passage of high-source leakage impedance current. This is not EMI. The impact of high-source leakage impedance current is threshold jitter, which causes timing errors. An optical cable will not pass high-source leakage impedance current, which means no threshold jitter.

Both types of cable are susceptible to clock phase noise from the el-cheapo clocks in generic networking devices.
Click to expand...

I mean distortion from all different sources, other electric items like microwaves, tv’s the amp itself ,the power cable which is located next to it, wifi signals etcetera

I was wondering, because you are talking about jitter. This prevention of jitter was exactly the reason why this ultrarendu was chosen, because it has USB 2 output. I thoughf that SP/dif, digital coax and toslink could cause jitter because it is sending ones and zero’s to the dac and USB works different with data files instead where it is reconstructed to the same packages before it is converted to analog.

Of course then it can again cause new jitter, but at this stage the network cable/fiber cable is already passed.
 
Alkyogre said:
...I was wondering, because you are talking about jitter. This prevention of jitter was exactly the reason why this ultrarendu was chosen, because it has USB 2 output. I thoughf that SP/dif, digital coax and toslink could cause jitter because it is sending ones and zero’s to the dac and USB works different with data files instead where it is reconstructed to the same packages before it is converted to analog.

Of course then it can again cause new jitter, but at this stage the network cable/fiber cable is already passed.
Click to expand...

correct ...this type jitter is not an issue with an asynchronous DAC receiving a USB data stream. here, the DAC takes at-rest data (i.e. cached) and uses its own internal clock for conversion timing. therefore, any jitter that could be cause by upstream devices is irrelevant since those clocks are NOT used in the conversion process.

as you point out, there is of course the issue of the accuracy of the DAC's internal clock but that is a different concern.
 
aKnyght said:
correct ...this type jitter is not an issue with an asynchronous DAC receiving a USB data stream. here, the DAC takes at-rest data (i.e. cached) and uses its own internal clock for conversion timing. therefore, any jitter that could be cause by upstream devices is irrelevant since those clocks are NOT used in the conversion process.

as you point out, there is of course the issue of the accuracy of the DAC's internal clock but that is a different concern.
Click to expand...

Along these lines I've yet to understand why people feel that re-clocking the ethernet-transmitted signal (sometimes multiple times) prior to it reaching a streamer/player that caches the data and reclocks it itself is a reasonable approach. Assuming there is a reasonable clock in the streamer/player.

Other than the inevitable "I tried 2 EtherRegen switches between my server and my player and 2 sounded better than 1." Next of course is 3 switches.

As to USB - I've never owned a USB dac so have 0 experience. At a Nagra demo however the Nagra rep was feeding the HD Dac with a Macbook and a cheapie usb cable. When asked he said 'with this dac it doesn't matter.' But obviously that's just one dac and I have 0 reasons to doubt that people with usb dacs hear improvements as they 'improve' the data stream.
 
Alkyogre said:
I was wondering, because you are talking about jitter. This prevention of jitter was exactly the reason why this ultrarendu was chosen, because it has USB 2 output. I thoughf that SP/dif, digital coax and toslink could cause jitter because it is sending ones and zero’s to the dac and USB works different with data files instead where it is reconstructed to the same packages before it is converted to analog.

Of course then it can again cause new jitter, but at this stage the network cable/fiber cable is already passed.
Click to expand...

Um..no. With all respect, not sure you're grasping what I'm referring to.
 
aKnyght said:
correct ...this type jitter is not an issue with an asynchronous DAC receiving a USB data stream. here, the DAC takes at-rest data (i.e. cached) and uses its own internal clock for conversion timing. therefore, any jitter that could be cause by upstream devices is irrelevant since those clocks are NOT used in the conversion process.

as you point out, there is of course the issue of the accuracy of the DAC's internal clock but that is a different concern.
Click to expand...

Not exactly accurate, aKnyght. Your comments above do not address the problems of threshold jitter. Threshold jitter is the result of noise on the signal ground plane due to high-source impedance leakage current. This causes timing errors, it is not the same type of jitter we're historically used to thinking of with digital.
 
Bart001 said:
Along these lines I've yet to understand why people feel that re-clocking the ethernet-transmitted signal (sometimes multiple times) prior to it reaching a streamer/player that caches the data and reclocks it itself is a reasonable approach. Assuming there is a reasonable clock in the streamer/player.
Click to expand...

It's because...it works. ;)
 
Puma Cat said:
Not exactly accurate, aKnyght. Your comments above do not address the problems of threshold jitter. Threshold jitter is the result of noise on the signal ground plane due to high-source impedance leakage current. This causes timing errors, it is not the same type of jitter we're historically used to thinking of with digital.
Click to expand...

hence, the benefit of optical isolation in eliminating all upstream electrical noise from the signal.
 
aKnyght said:
hence, the benefit of optical isolation in eliminating all upstream electrical noise from the signal.
Click to expand...

Yes, sir! ;)

I'm busy getting breakfast ready at the moment but I'll put up some references in just a bit, by someone who knows more this than I do because...they did it professionally for a living for 40 years. Namely, John Swenson...who designed the UltraRendu, OpticalRendu, EtherREGEN, and the Uptone ISO Regen reclocker.
 
it isnt that easy.....no matter how good that last clock in the chain is, it is meant to work for the dac and not to clean dirt from upstream devices. it gets overstrained.

i am all in for double or tripple switches with top clocks and all the tweaking.
and reclocking. its all about it.
 
u-sound said:
it isnt that easy.....no matter how good that last clock in the chain is, it is meant to work for the dac and not to clean dirt from upstream devices. it gets overstrained.

i am all in for double or tripple switches with top clocks and all the twaeking.
and reclocking. its all about it.
Click to expand...

There's no..."dirt." And clocks don't get.."strained."
 
You must log in or register to reply here.
Back
Top