Would this setup work?

I run audirvana core player on a synology 224+. This nas has two ethernetports. I assigned the second portone for use to communicate with my pliniuis tiki upnp streamer. This works wel. The 2nd nas port goes into a switch.

I now like to take it a further step by, instead of the going into a switch, connect the 2nd port to a mediaconvertor-> fiber cable → mediaconvertor → plinius tiki. So the plinius tiki will by ‘directly’ connected to the nas. Kind of how the direct player port on a melco works.

Will this work? Can the plinius obtain via dhcp an ipadress, i cannot set a static ip-adress.

Which device assigns IP addresses?

You won’t profit from the added fiber connection, since the media conversion (for example FLAC to RAW PCM) is done by Audirvana on the NAS. Any jitter that is introduced into the signal will pass through the fiber conversion untouched. To make matters worse, the fiber conversion might introduce additional jitter because of the electrical noise of the fiber converters.

This is probably the worst setup thinkable in terms of sound quality.

My internet modem assigns the ip-address.

The current setup is:

  • Internetmodem
    • switch 1 with plinius stream
      • switch2 with nas
        Where switch 2 is connected to switch 1.

I’d keep this setup. That works.

If you want to improve sound quality, the best option is to invest in better power supplies for the NAS and for switch 1.

All my switches and internetmodem or on a liniear power supply.

So does does the nas which also has 18GB memory and 2 SSD disks. I recently switched from a 220+ to 224+ which made a very noticeable change.

Switch 1 a sfp switch, switch 2 connected via sfp.

So I already made a lot upgrades.

Then you are already as good as you can get.

I am trying to create a dedicated player port like a melco has.

Buy a Melco……

I’ve had both copper and fiber connections for lengthy periods, and prefer the sound quality with fiber.

Your statement about passing along jitter and possibly making it worse isn’t accurate with regard to 10G and faster Ethernet, copper or optical. Both have specifications that require not only low jitter levels, but also that any incoming jitter is reduced to a specified level. This is why my entire optical Ethernet connection between computer and endpoint is 10G spec. I’m therefore assured of a very low level of jitter in the signal throughout that connection.

An additional assurance against jitter is that my USB connection is also optical. There’s complete galvanic isolation, no electrical interference transmitted from computer or endpoint to the DAC’s clock.

There are certainly other ways to minimize jitter. I don’t think we want to be too absolutist about any particular method always being good or bad.

Re Read: in the proposed setup

I think it would work. You would want the cables and switches to be 10G or higher spec, for the reason I mentioned in my response to @jmtennapel . You would also want to pay careful attention to power supplies, to ensure they have low levels of leakage current and other noise.

Then the question becomes whether this versus some other method gets you the lowest levels of jitter and electrical noise for the least money.

Right, so if for example you use a couple of these switches as fiber converters:

TeraDak T-S212 media converter. This is the mediaconvertor I am using on my streamer. Other network appliances are mikrotik.

My question is about use one netwerkport of the nas as a dedicated direct connection to the plinius tiki networkplayer. If this is fiber or copper shouldn’t matter in this case and is another discussion.

It’s gigabit (1G), and no jitter specification that I’m aware of. So no way to know whether it would be an improvement or not. (I’m not sure why you mentioned two media converters if it’s immaterial whether you use copper or fiber?)

Putting the second networkport in another subnet and spinning up a dhcp server on the nas for this subnet was the solution to make this setup work.

This topic can be closed.

1 Like