Objective-Fi Optical Networking & SFPs

5 minutes ago, kennyb123 said:

 

We've gone round and round on conducting a proper test.  I'm sorry - but I have to ask you this:   what was your educational background?  You don't seem to be able to understand how to properly conduct experiments.  Nor do you seem to understand the need for experiments to be done properly.

Ok, let's put up a hypothetical:

You've just met a person that says they, from a standing position, can jump up and clear a 10' high bar.

Do you design an elaborate test or do you bring out a tape measure?

I've a follow on question: What about my proposed testing rig is problematic? Chris asked a question, part of it would be how can we get a setup that would allow to test for this.

10 minutes ago, plissken said:

 

Ok, let's put up a hypothetical:

 

You've just met a person that says they, from a standing position, jump up and clear a 10' high bar.

 

Do you design an elaborate test or do you bring out a tape measure?

That you see this as analogous to the differences we are hearing is not at all surprising to me.  Your tests are constructed to take measurements without the necessary precision and you can't even see this.

This in an objective sub-forum.  Your approach would earn you an F and lots of derision at any engineering school.

Just now, kennyb123 said:

 

That you see this as analogous to the differences we are hearing is not at all surprising to me.  Your tests are constructed to take measurements without the necessary precision and you can't even see this.

Which tests? The ADC capture? If so why isn't it precise?

Human evaluation? Why don't you trust your ears?

Just now, kennyb123 said:

 

 Your approach would earn you an F and lots of laughs at any engineering school.

You want to know what would really get laughs at an engineering school? You're ability to hear differences between SFP+ modules.

18 minutes ago, plissken said:

 

Solar Flare adapter has an easy to use Teaming GUI. I can put an FS and Finisar SFP+ module in side by side and simply disconnect at the switch while music is playing.

 

Also Windows 10 features by default SMB multichannel if we want to dispense with teaming the NIC.

That wouldn’t isolate a noise SFP 100%. 

Just now, The Computer Audiophile said:

That wouldn’t isolate a noise SFP 100%. 

Why not. We are unplugging the cable. We could instead unplug the adapter.

8 minutes ago, ray-dude said:

FWIW, at the engineering schools I went to the reaction would have been "No way!  Let me hear that...  Huh...  Wow...  What the hell is going on?  Cool, let's try to figure this out!"  This is the exciting stuff!  I hope that is the spirit of this thread.

I've zero issue with that. I would even be willing, when this covid madness is over, set this up single blind and see you demonstrate this ability.

5 minutes ago, plissken said:

 

I've zero issue with that. I would even be willing, when this covid madness is over, set this up single blind and see you demonstrate this ability.

 

 

37 minutes ago, plissken said:

 

Why not. We are unplugging the cable. We could instead unplug the adapter.

You would need to unplug the adapter. 

Just now, The Computer Audiophile said:

You would need to unplug the adapter. 

No problem. I do that all the time during my day job 🙂

@ray-dude Which in your list is SFP capable? Are you using FMC converters to last leg it over RJE?

Audio System

Paul, I have a couple different components that are SFP capable:

The  StarTech PEX1000SFP2 PCIe optical Network card in my Taiko Audio Extreme

(2x) Ubiquiti EdgeRouter X SFP's

Sonore opticalModule

I recently sold my Uptone Audio EtherREGEN, but I was experimenting with the SFP port on the ER in my Taiko Audio SGM Extreme review (part 4 for those subjective findings)

I currently have some TP Link multi-mode SFPs connecting my two EdgeRouter X SFPs.  ERX1 is my router/gateway and connected to my ATT fiber ONT.  It has a audio subnet that I connect by copper to my opticalModule. I run fiber from the opticalModule to my Extreme.  I've run both the Planet Tech SFPs and Finisar SFPs on my audio net.

I have the rest of my home network (including WiFi) on ERX2, galvanically isolated from ERX1.  ERX2 is configured as a simple switch, but I do have a couple VLANs for my WiFI (guest network, IoT network, etc)

With the Planet Tech SFPs, I preferred copper NIC on the Extreme (fiber to the opticalModule+PlanetTech to copper to the Extreme).  With the Finisars, I'm back to preferring the optical NIC, esp. now that I have my audio network isolated from my home network, but it isn't a clear preference.  

FWIW, subjectively, reduced network traffic to the my music server NIC does have an impact.  This is contributing to my hypothesis of laser switching-induced electrical noise being the mechanism in play.

All of the above is with streaming TIDAL content to Roon to HQP to DAC.

The network part of my system is very much in flux right now, because of what I've been hearing.  There are a couple of permutations I'd like to try to see if it helps develop a hypothesis for a what a core mechanism could be.

Here is my wish list for things to try if I had proper equipment in quarantine with me:

* Probe power rails on the SFP modules and look at spectrum of noise generated on same by network traffic.  If that spectrum is the same, my hypothesis can filed in the "oh well" cabinet.  If different, then the question shifts to whether that difference is audible (and if so, how).

* Build an extender and wire to SFP cage outside of NIC/FMC/Routers/etc.  Rig so that I can swap between powering SFP externally or via the cage leads. If there is variability with different class power supplies, that would generally support the laser-switching-noise hypothesis. If not, back to the "oh well" cabinet we go.

* Get access to OEM data sheets for SFP modules and find the lowest power consuming unit available...give it a listen, get another subjective data point

@Superdad without asking you to divulge anything proprietary, did you or John look into any of these things when you were designing and prototyping the ER? (and if that is poking at things that you consider proprietary, my apologies in advance)

23 minutes ago, ray-dude said:

 

@Superdad without asking you to divulge anything proprietary, did you or John look into any of these things when you were designing and prototyping the ER?

Hi Ray: 

We don't include any SFP transceiver with the EtherREGEN and have not dissected and measured ground-plane noise or looked at the signal integrity eye-pattern for various fiber-optic modules.

One thing I will point out--which I think you might find interesting--is something sort of obvious but overlooked:

All Ethernet RJ45 copper connections go though both a set of transformer cores (typically 2~8 tiny wound round cores, though for EtherREGEN we use 12-core-per-port magnetics) and PHY transceiver circuitry (mostly these days built into the Ethernet switch chip, but for some switches a discrete PHY chip, per port or grouped, is used--ala the old Cisco Catalyst 2960s, likely one of the things that made them special).

And the signals for SFP cages are always wired directly to the switch chip via the processor's SGMII interface (or even QSGMII or XGMII for much higher speeds--though those are not via SFP cages). Aside from that direct interface perhaps not activating as much circuitry/generating ground-current noise/clock-threshold jitter inside the chip, SGMII is also an LVDS interface, and that carries some advantages.  

Lastly, most good switches will dedicate a 3.3V regulator to provide power to whatever SFP transceiver is plugged into the cage (as we do with an LT3045 just one centimeter behind the EtherREGEN's SFP cage).

--Alex C.

P.S. By the way, @plissken's first name is Mark not Paul. (And over at ASR he is Jinjuku; various other names elsewhere...)

6 minutes ago, Superdad said:

P.S. By the way, @plissken's first name is Mark not Paul. (And over at ASR he is Jinjuku; various other names elsewhere...)

Mark, my apologies!  I have no idea how "Paul" settled into my brain, but certainly no disrespect was intended.  Thank you Alex for the gentle correction (as well as the info that you shared)

3 hours ago, Superdad said:

One thing I will point out--which I think you might find interesting--is something sort of obvious but overlooked:

All Ethernet RJ45 copper connections go though both a set of transformer cores (typically 2~8 tiny wound round cores, though for EtherREGEN we use 12-core-per-port magnetics) and PHY transceiver circuitry (mostly these days built into the Ethernet switch chip, but for some switches a discrete PHY chip, per port or grouped, is used--ala the old Cisco Catalyst 2960s, likely one of the things that made them special).

And the signals for SFP cages are always wired directly to the switch chip via the processor's SGMII interface (or even QSGMII or XGMII for much higher speeds--though those are not via SFP cages). Aside from that direct interface perhaps not activating as much circuitry/generating ground-current noise/clock-threshold jitter inside the chip, SGMII is also an LVDS interface, and that carries some advantages.  

Lastly, most good switches will dedicate a 3.3V regulator to provide power to whatever SFP transceiver is plugged into the cage (as we do with an LT3045 just one centimeter behind the EtherREGEN's SFP cage).

 

@Superdad any chance you can summarize this for me?  It sounds like you are saying that one interface has an advantage over the other, but I’m unfamiliar with some of these terms so I’m not able to interpret which is better and why that’s so.

10 hours ago, Audiophile Neuroscience said:

 

 

What could possibly go wrong? 🙄

 

I didn't know Mark still offered his mobile blind testing debunking service. IIRC it used to come with some kind of monetary incentive.

Yep my $4k to your $1K. It's amazing how quickly people can be fanned back from home plate.

As the OP of this thread, I ask that the nonsense be stopped. This thread is in the Objective-Fi sub-forum. Please stay on topic and strive for objective information. 

When I receive my Finisar Long Reach modules and 15 meter SM OS2 cabling what kind of testing do you want to see? Objective testing that is.

Do you want to see if 1KHz is affected? Do you want me to record at track into my ADC and swap cables? Do we want to consider a get together when appropriate?

I'm asking for what empirical, debiased, method is acceptable. I'm spending my own $$ to move this conversation along. What DAC? Would the Schiit Modius suffice with it's Unison USB UAC2 implementation? Would 24/192 suffice? If 24/192 then what tracks would you like?

1 hour ago, Superdad said:

 

Some of those might be an advantage favoring the use of optical, some of them might instead favor--or at least influence interaction (and sonic differences)--towards copper.

 

Ah maybe that's why I was having difficulty determining if you were saying one was better than the other.  Appreciate you sharing those insights. 

Just waiting on the Finisar for the client end to show up: