Taiko Audio SGM Extreme : the Crème de la Crème
- Thread starter CKKeung
- Start date
Quick question to All Extreme users
What is the router you use ?
I was comparing the ubiquiti edgerouter with edgerouter sfp and the one with sfp is a better sounding.
tried to look why is that and the only significant difference I could find was the 2 clocks in SFP version.
There is only one clock in regular edgerouter.
All proposals for ultimate Router are welcome.
what are you using currently with your Taiko Extreme?
I am also curious what Emile and Taiko Team is using.
What is the router you use ?
I was comparing the ubiquiti edgerouter with edgerouter sfp and the one with sfp is a better sounding.
tried to look why is that and the only significant difference I could find was the 2 clocks in SFP version.
There is only one clock in regular edgerouter.
All proposals for ultimate Router are welcome.
what are you using currently with your Taiko Extreme?
I am also curious what Emile and Taiko Team is using.
I’m using my Internet Service Provider router and a Paul Pang modified switch, while waiting for the Taiko new one come out.
Router: Ubiquiti UniFi Dream Machine Pro
Switch: Ubiquiti UniFi USW-24-POE Gen2
Wifi: Ubiquiti Networks UniFi 6 Long-Range
Do note we have a relatively busy network environment compared to your usual home network:
Last edited:
--FTTH on ISP Nokia modem.
--Netgear Nighthawk WiFi/Router.
--PFBuffalo switch
--Sonore OM
All devices are on LPSs--JS-2, LPS1.2, Farad3s. Approximately one meter of copper in network (not counting PCBs ). The network gear is on a dedicated line from the panel, opposite phase from audio gear. Fiber network is audio only.
--Netgear Nighthawk WiFi/Router.
--PFBuffalo switch
--Sonore OM
All devices are on LPSs--JS-2, LPS1.2, Farad3s. Approximately one meter of copper in network (not counting PCBs ). The network gear is on a dedicated line from the panel, opposite phase from audio gear. Fiber network is audio only.
Thanks to all ! More ifo form other users is welcome.
I think Router is a big part of the SQ.
This is also why we are getting different results.
Planning to test some of the solutions to make sure we know what is going on.
So far
1. Edgerouter X with pinkfaun ultra clock is a king
2. Edgeriuter X sfp
3. Edgerouter X
4. Cisco Linksys EA 4500
need more devices to try,
I remember Emile tested Nighthawk one time and it was a disaster.
But it looks Marcus has very good results with that.
I think Buffalo with PF clock , following NH is a big part of his success.
I think Router is a big part of the SQ.
This is also why we are getting different results.
Planning to test some of the solutions to make sure we know what is going on.
So far
1. Edgerouter X with pinkfaun ultra clock is a king
2. Edgeriuter X sfp
3. Edgerouter X
4. Cisco Linksys EA 4500
need more devices to try,
I remember Emile tested Nighthawk one time and it was a disaster.
But it looks Marcus has very good results with that.
I think Buffalo with PF clock , following NH is a big part of his success.
@Kris the guy who developed the Fidelizer software has a modded router, BTW. I had some communication with him, but I was not able to determine the actual potential of this product, so I did not purchase it. Happy to hear from others better able to evaluate this product re: benefits.
www.fidelizer-audio.com
Introducing AirStream, Audiophile Wireless Router from Fidelizer Audio | Fidelizer.Audio
Today I'd like to introduce a new product I've been working for over a year now. It's a wireless router with modification that can bring high fidelity audio in the air for the first time.
www.fidelizer-audio.com
...@Kris, if I had better network kung-fu I would have used the Ubiquity. I have the X with the SFP. I tried in vain to get it going on my copper network before fiber was added, and I could not get it to work correctly.
Now the only network traffic is if I stream Qobuz, or the WiFi commands from Taiko app. It's my "Extreme" vlan.
Are you still without a WiFi "attachment"? You were hardwired, as I recall.
Now the only network traffic is if I stream Qobuz, or the WiFi commands from Taiko app. It's my "Extreme" vlan.
Are you still without a WiFi "attachment"? You were hardwired, as I recall.
I use #2
From my QNAP main network switches to Edgerouter X SFP: Useind a Keces LPS at 12v on Edgerouter
Tried both fiber in/copper out to Extreme and copper in/fiber out to Extreme
Right now I'm fiber in/copper out to Extreme
Waiting to hear what Emile and the Taiko team come up with/recommend.....
Not using a separate router at all, Kris. I'm going directly from one of the four provided RJ-45 ethernet sockets on the internet service provider's modem/router directly to the Extreme. I do however use a linear power supply instead of the ISP SMPS wall-wart for the modem/router.
Steve Z
steve williams
Site Founder, Site Owner, Administrator
I do exactly the same. I don’t use any switches unless of course Emile releases one.
I am testing both options with and without Wifi.
without is great but there is touch less control of the sound. Not sure why.
I want them both So I regulate this with Buffalo switches and connection between them.
I have dedicated fibre line just for music room direct from network provider server.
15 km single fibre…..no swiches nor any active things on the way to main server.
Transparent modem that is actually just optical fibre to Copper etherner module.
I must agee with Emil expression that this is a „ soup of noise”
What to add and what to disconnect is a very big question.
It looks I can adapt the SQ to any scenario now.
Its amazing that using only switches, modem, router and their supplys you can get any kind of sound.
From unacceptable to perfection ….. And all in between.
Having Taiko switch will be a cure for this disease I hope.
I'm using a Nighthawk router with one band dedicated to my system. I believe I'm using a different model than the one Emile tested. I'm using S Boosters on both the router and modem.
steve williams
Site Founder, Site Owner, Administrator
Fantastic.
I plan to borrow some of the routers from different sources .
Quite sure that the number of cores and Ram size makes a big difference .
just like it makes a big difference when using 10 cores insted of 1 for the server processing.
I also suspect none of us is close to what Emile is hearing at Taiko as he is using
Ubiquiti Dream Machine Pro.
This is a true beast for what we need.
And the name of that router is Dream Machine , so we know what to expect.
I plan to borrow some of the routers from different sources .
Quite sure that the number of cores and Ram size makes a big difference .
just like it makes a big difference when using 10 cores insted of 1 for the server processing.
I also suspect none of us is close to what Emile is hearing at Taiko as he is using
Ubiquiti Dream Machine Pro.
This is a true beast for what we need.
And the name of that router is Dream Machine , so we know what to expect.
Ubiquiti UniFi Dream Machine Pro
Processor Quad ARM Cortex-A57 Core at 1.7 GHz
16 GB eM Storage
4GB RAM that is DDR4 .
Most of what we use is single core, 512 mb ran ddr2 !
Not sure if Emile tried the most popular Edgerouter X SPF In comparison to his Dream Machine ?
Processor Quad ARM Cortex-A57 Core at 1.7 GHz
16 GB eM Storage
4GB RAM that is DDR4 .
Most of what we use is single core, 512 mb ran ddr2 !
Not sure if Emile tried the most popular Edgerouter X SPF In comparison to his Dream Machine ?
Yes, the dream machine works better here, but there's more to it then just the device.
We had several connectivity issues with packet drops and sometimes poor throughput performance and/or high latency on wireless devices using a mixture of switches / wireless access points. Largely because this is a large building and we're using part of the neighbouring building now too totalling to about 2200 m2 to cover. The solution to this was using all Ubiquiti Unify devices and 3 Ubiquiti Unify Long Range access points. We now have a very well performing network, all devices have near 100% maximum throughput, low latency and no packet drops. Part of the ability to diagnose this effectively is centralised management as Unify provides when using all Unify devices. You can see where things go wrong and make adjustments. Surprisingly here now 5G wifi works better then 2.4G, 2.4G can even cause poor Roon remote performance, and that IS audible. In our previous facility in Hengelo I had a single Wifi Router running at 2.4G only and spaced it far away from the listening room. That does not work well here at all.
Right now there is NO difference here in SQ when unplugging the network cable after starting a track with TAS, while we have continuous network utilization, several internet radio devices streaming music all day etc. The obvious question here is IF in fact this continuous utilization is actually SQ positive, a constant "noise soup" where the deviation from using a wireless remote or streaming a music file is very small relative to the continuous normal utilization. To test this theory I have created a dedicated "music" VLAN, which, much to my surprise, sounded worse, with a sizeable uplift in SQ when unplugging the network cable. A logical next step would be to deliberately create a load while listening to music.
I have been pondering this for a while now.
It is true that disconnecting the network cable in Taiko’s system did not make a difference. It is also true that I was shocked that it did not make a difference. Everywhere else I have tried that in a resolving system, the difference has been quite obvious. But I’ve only tried that in small home networks before the Taiko system.
That raises the question Emile touched on. What if the problem with our networks is that in a small home network the occasional packet processing causes more interruption than in a busy network? Imagine a network device sitting mostly idle and receiving packets every few seconds. Each packet has to “wake up” (not literally) the network device. On the other hand, in a busy network, where tens of thousands of packets are processed every second, the occasional audio related packet makes relatively no difference.
To try to give an illustrative example - in a small home network you are riding a car that has to stop on a stop sign every few seconds; in a busy network, you are riding on a highway with a constant speed. Your “highway” ride will likely be much smoother.
Haven’t had time to do more testing in my system due to other priorities, but an interesting experiment would be run a traffic generator that will keep constant traffic flow and check if my network tweaks, such as Buffalo switches with PinkFaun ultraOCXO clock and Sean Jacobs DC4 LPS make the same difference, or less, or none.
You can also think about this as loading the network like a class A amplifier that is almost always working at high capacity. I guess we can call it a Class A network
. It's a very interesting idea, and the concept is not new to high end audio.
It is true that disconnecting the network cable in Taiko’s system did not make a difference. It is also true that I was shocked that it did not make a difference. Everywhere else I have tried that in a resolving system, the difference has been quite obvious. But I’ve only tried that in small home networks before the Taiko system.
That raises the question Emile touched on. What if the problem with our networks is that in a small home network the occasional packet processing causes more interruption than in a busy network? Imagine a network device sitting mostly idle and receiving packets every few seconds. Each packet has to “wake up” (not literally) the network device. On the other hand, in a busy network, where tens of thousands of packets are processed every second, the occasional audio related packet makes relatively no difference.
To try to give an illustrative example - in a small home network you are riding a car that has to stop on a stop sign every few seconds; in a busy network, you are riding on a highway with a constant speed. Your “highway” ride will likely be much smoother.
Haven’t had time to do more testing in my system due to other priorities, but an interesting experiment would be run a traffic generator that will keep constant traffic flow and check if my network tweaks, such as Buffalo switches with PinkFaun ultraOCXO clock and Sean Jacobs DC4 LPS make the same difference, or less, or none.
You can also think about this as loading the network like a class A amplifier that is almost always working at high capacity. I guess we can call it a Class A network
. It's a very interesting idea, and the concept is not new to high end audio.Thank you Emile. Amazing how many aspects play a key difference.
I will test busy network also. Will connect gaming server to my audio network.
I agree with 5 G wifi. It could be very nasty but if properly done you gain lightning speeds And sharper ,
better controlled sound.
