Objective-Fi Optical Networking & SFPs

Hi Guys, this is an objective topic to look at the assertions made by some that SFPs make a sonic difference in audio systems. I'd like to use real science and a real scientific approach in this one. This means, we may have a belief one way or the other, but we should go into this with an open mind and follow the available objective evidence. 

@jabbr's thread on optical networking is what made me think of creating this topic. That thread is much more subjective and I know others have objective comments they'd like to put forth in this discussion. 

Here is Jabbr's thread - 

I'd be curious if anyone has done any bench testing using an analyzer of some kind to see if the SFP's generate any more or less noise than a regular eth nic does. It certainly makes me wonder why almost no manufacturer of high quality dacs has gone down the fiber direct route to dac yet. I believe there are a few that have used the fiber direct to dac connection like Meitner and Bel Canto but they are the only ones that come to mind.

It seems like a no brainer to a simpleton like me to try and use a fiber direct solution to a dac due to fiber cables baked in electrical isolation but I'm clearly missing why it hasn't caught on yet. I mean in the IT world fiber usage is everywhere and has been for many, many years so there is nothing really untested or new about the technology itself.

Short of the electrical isolation part of fiber I see no other reason to use it for home audio purposes unless one stores there computer, nas or streamer in a neighbors house hundreds of feet away or a few km away in the next town over from the listening room. I also see zero reason to use it for speed reasons as 1gb regular nics are already super overkill. And then I've seen some going 10gb for audio which can't be for anything but bragging rights or just out of curiosity as a project in ones spare time, which of course is perfectly fine if that's what interests the person doing it.

I would certainly love to see it catch on with dac manufacturers but it seems the industry  just ain't there yet or perhaps ever will be

1 minute ago, cjf said:

I'd be curious if anyone has done any bench testing using an analyzer of some kind to see if the SFP's generate any more or less noise than a regular eth nic does. It certainly makes me wonder why almost no manufacturer of high quality dacs has gone down the fiber direct route to dac yet. I believe there are a few that have used the fiber direct to dac connection like Meitner and Bel Canto but they are the only ones that come to mind.

 

It seems like a no brainer to a simpleton like me to try and use a fiber direct solution to a dac due to fiber cables baked in electrical isolation but I'm clearly missing why it hasn't caught on yet. I mean in the IT world fiber usage is everywhere and has been for many, many years so there is nothing really untested or new about the technology itself.

 

Short of the electrical isolation part of fiber I see no other reason to use it for home audio purposes unless one stores there computer, nas or streamer in a neighbors house hundreds of feet away or a few km away in the next town over from the listening room. I also see zero reason to use it for speed reasons as 1gb regular nics are already super overkill. And then I've seen some going 10gb for audio which can't be for anything but bragging rights or just out of curiosity as a project in ones spare time, which of course is perfectly fine if that's what interests the person doing it.

 

I would certainly love to see it catch on with dac manufacturers but it seems the industry  just ain't there yet or perhaps ever will be

EMM and Bel Canto use audio over optical connections, not Ethernet fiber connections. 
 

Lumin’s X1 has an SFP slot that I used during my review. Good stuff. 

8 hours ago, cjf said:

I'm clearly missing why it hasn't caught on yet

From my POV:

1. It doesn't make a SQ difference

2. It adds unnecessary cost

3. It adds unnecessary complexity

4. Everyone knows what an RJ45 Ethernet cable is

5. WiFi offers just as much isolation while providing more than enough throughput on 5G / 80VHT

32 minutes ago, plissken said:

WiFi offers just as much isolation while providing more than enough throughput on 5G / 80VHT

That’s a great topic for another thread. 
 

WiFi devices that let the user control the channel width are great. Heck, my UniFi stuff goes to VHT160, but I have nothing that can use that. VHT80 is great as long as the environment is great. 

7 hours ago, The Computer Audiophile said:

That’s a great topic for another thread. 
 

WiFi devices that let the user control the channel width are great. Heck, my UniFi stuff goes to VHT160, but I have nothing that can use that. VHT80 is great as long as the environment is great. 

VHT 80 is recommended for Greenfield installs. Especially if you can use DFS channels.

2 minutes ago, plissken said:

VHT 80 is recommended for Greenfield installs. Especially if you can use DFS channels.

DFS will stop working in the presence of radar though correct?

8 hours ago, plissken said:

 

From my POV:

 

1. It doesn't make a SQ difference

2. It adds unnecessary cost

3. It adds unnecessary complexity

4. Everyone knows what an RJ45 Ethernet cable is

5. WiFi offers just as much isolation while providing more than enough throughput on 5G / 80VHT

Per Chris' post:  "I'd like to use real science and a real scientific approach in this one. This means, we may have a belief one way or the other, but we should go into this with an open mind and follow the available objective evidence."  Can you provide any objective evidence?

2 minutes ago, The Computer Audiophile said:

DFS will stop working in the presence of radar though correct?

Correct,  it will channel hunt and hop. But you should be able to see DFS alerts in the controller.

1 minute ago, kennyb123 said:

 

Per Chris' post:  "I'd like to use real science and a real scientific approach in this one. This means, we may have a belief one way or the other, but we should go into this with an open mind and follow the available objective evidence."  Can you provide any objective evidence?

Sure. Would you take a ADC capture of a recording where I swap out cables during playback?

On 6/15/2020 at 8:38 AM, The Computer Audiophile said:

Hi Guys, this is an objective topic to look at the assertions made by some that SFPs make a sonic difference in audio systems. I'd like to use real science and a real scientific approach in this one. This means, we may have a belief one way or the other, but we should go into this with an open mind and follow the available objective evidence. 

Great topic! Thanks for posting this.

@ray-dude has put forth the hypothesis that SFPs that draw lower power will sound better.  Thus far the subjective evidence seems to back this.  I'd love to understand why.

5 minutes ago, plissken said:

 

Sure. Would you take a ADC capture of a recording where I swap out cables during playback?

That won't address the SFP question though. 

7 minutes ago, plissken said:

 

Sure. Would you take a ADC capture of a recording where I swap out cables during playback?

No, because none of your claims referenced audio.  For example you claimed that "WiFi offers just as much isolation while providing more than enough throughput on 5G / 80VHT".  I'd love to see proof.

Correction, your first claim did reference sound quality.  Your test would be acceptable only if you proved that your methodology was capable of detecting known differences that impact sound quality.  In other words, add noise to ethernet until the output changes to see if your test can identify this.

Thank's for the pointer to this topic Ken.  My hypothesis is that switching to the laser is causing a correlated pulsing on the reference voltage and/or ground planes that is having an indirect impact on the DAC.  

My basis for this is that in previous experiments with coax SPDIF signals to my Chord DAVE, that particular DAC seemed to have a strong susceptibility to parasitic signals in the 2-3GHz band.  Copious ferrites on the dual coax connectors to the DAC had a surprising benefit to SQ.  IIRC Rob Watts (designer of the DAC) attributed the impact to noise floor modulation in the DAC (I would have look up some old notes to confirm).  It is easy to replicate this by having a WiFi basestation configured for 2.4Ghz and bringing it close to this DAC, or having  music server on WiFi that is close to the DAC.  

All above is hypothesis or experiential.

48 minutes ago, The Computer Audiophile said:

That won't address the SFP question though. 

Why wouldn't it? You can only hear a difference if the output of a DAC is altered. I'm offering a way, just like Alpha Audio on YouTube did when they tried six different Ethernet Cables.

55 minutes ago, kennyb123 said:

 

@ray-dude Thus far the subjective evidence seems to back this.  I'd love to understand why.

This is the objective-fi sub-forum.

Just now, plissken said:

 

This is the objective-fi sub-forum.

Yes it is.  Hoping someone is able to provide some objective insight.

@kennyb123 What objective data would you accept?

7 minutes ago, plissken said:

...just like Alpha Audio on YouTube did when they tried six different Ethernet Cables.

And that was the silliest thing I ever saw. They (and you via your extracting the audio from the video) were expecting comparison of Ethernet cables to be valid via microphone recording through some unknown ADC, then encoded/compressed into a YouTube video. Unbelievable--and surely not "objectively" valid in the least 

Just now, kennyb123 said:

 

Yes it is.  Hoping someone is able to provide some objective insight.

Ok. so if I spin up my ADC, connect my DAC to it's MIC inputs, and capture 1Khz with both SFP+ and make available for download would you evaluate those?

If current circumstances didn't prohibit it: If I came out to you on you're setup with my playback computer with my dual SFP+ Solar Flare card and did this with you blind, would you evaluate that way?

7 minutes ago, plissken said:

 

Why wouldn't it? You can only hear a difference if the output of a DAC is altered. I'm offering a way, just like Alpha Audio on YouTube did when they tried six different Ethernet Cables.

You would need to switch SFPs if the hypothesis is that SFPs sound different. 

7 minutes ago, Superdad said:

 

And that was the silliest thing I ever saw. They (and you via your extracting the audio from the video) were expecting comparison of Ethernet cables to be valid via microphone recording through some unknown ADC, then encoded/compressed into a YouTube video. Unbelievable! 

I didn't make those rules. But the comments on that video read just about the same as the comments for your eR. So are they wrong? Why do you get to sit in judgement?

All the sudden your technical understanding is acceptable why ADC into a Mic, transcoded over to AAC 32/44.1 on YT is a problem. But I point out the same issues with your eR and my technical understanding isn't. You've got a brass set of em', I'll give you that.

Just now, The Computer Audiophile said:

You would need to switch SFPs if the hypothesis is that SFPs sound different. 

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.

9 minutes ago, plissken said:

 

Ok. so if I spin up my ADC, connect my DAC to it's MIC inputs, and capture 1Khz with both SFP+ and make available for download would you evaluate those?

 

If current circumstances didn't prohibit it: If I came out to you on you're setup with my playback computer with my dual SFP+ Solar Flare card and did this with you blind, would you evaluate that way?

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.

Here's a picture for you.  See the "procedure working" step?  See the how the results that don't align only lead to constructing a new hypothesis.  You ignore these things.