We have all heard about testing speed on devices like laptop etc and i have noticed that the speed varies for devices like the google pixel phone i have on the 5G wifi i get 30Mbps but on my laptop on the same network i get 101 Mbps... I am wondering now how can ae do this test for the TV sets? Use the browser app on the TV, correct?
Normal sites like speedtest.net are not relevant for Stadia. The only one that comes close is https://projectstream.google.com/speedtest
If your smart TV doesn't have a modern web browser, then there's no real way to tell beforehand what kind of speeds you are getting from the TV.
However, the only way to get Stadia on your TV (other than plugging a computer in via a display cable) is for the Chromcast ultra that comes with the Founders Edition. This will be no reliant on the TV for internet, so it doesn't have to be a smart TV. Just use the HDMI connector (and preferably the Ethernet adapter)
You can try using the built-in speed test in the Netflix app I suppose.
Settings -> Get Help -> Check Your Network
Honestly though, don't worry too much about speed tests... just try it. If it works well, great. Only if it doesn't do you have to start worrying about fixing something. If for whatever reason you can't get an Ethernet cable to where your TV is then I would highly recommend a good set of powerline ethernet adapters which are going to (generally) be way better for ping than wifi.
I’d really like to see the latency results added to the test. Similar to what Netflix’s fast.com has. With all the hoopla about latency in the media for stadia, none of us actually knows what ours is. Just our download speed results which is a completely different thing. A ping time should give us a better understanding of what to expect for input lag.
I'm not sure that the latency result would actually be that much use. As I understand it (although I could be misunderstanding the technology involved here), the performance is only being measured to an M-Lab server present in some Tier 1 data centre, so you're not really getting a proper end to end measurement to the Stadia server.
On the flip side, perhaps I am wrong and there is a lot more data analysis being made that is from your browser to the M-Lab server in conjunction with additional measurements made between the M-Lab server and the Stadia server(s)?? I don't know if that is encompassed into the Green/Red measurement that projectstream provides. All I know is that based on the network performance I have the projectstream values are way off from the values reported by fast.com or speedtest.net. So any values reported probably aren't that accurate to begin with, and if there is that much difference on download speeds, I wouldn't be very confident in latency measurements either.
projectstream: 169.338 Mbps,
fast.com: 1.5 Gbps
speedtest.net: 549.10 Mbps
That’s awesome. I’d never tried the three different ones to see what the differences were. I suspect that the distance the signal travels to the server makes a big difference but just switching from isp to isp can make a difference:
I occasionally do service provider backbone testing for work but I usually just run test set to test set. The test ramps up #of packets of different sizes until it fails and then averages the results(rfc2544 is one of the tests). I don’t generally test using any of the web based speed tests but when I have tried out if curiosity fast.com is usually the only one that shows numbers close to what I get using the dedicated test sets.
The latency I’d like to see would be from the home to the google game server I’d be connecting to but I also am unsure of how the Stadia magic works so maybe this type of latency is a non issue. I’d just figured it would give an indication on what kind of latency I I expect on button presses. I’m no network engineer, just someone who knows enough to be dangerous