Canadian TV, Computing and Home Theatre Forums banner

Which speed test is the most accurate?

20K views 21 replies 8 participants last post by  reets 
#1 ·
My wife and I recently got a Rogers bundle with Ignite 100u + Select TV + basic home phone. Last night I ran a test* through Rogers' SpeedCheck and got ~115Mbps down. This agreed with speedtest.net which - like SpeedCheck - uses Ookla. :smile

When I used a different - supposedly more-accurate - test at speedof.me or testmy.net, I got ~30-40Mbps down. :frown

Does anyone know:
- why I'd get such different readings from different tests;
- which test(s) I should consider (more) "definitive" (than others); and
- which non-Rogers test(s), if any, Rogers considers "definitive"?

Thanks. :)
__________________
(*I was testing my office PC wired to an Asus RT-N66U in repeater mode. My wife's laptop - wireless, but in the same room as the Rogers CGN3 gateway - doesn't get over 40Mbps according to SpeedCheck.)
 
#2 ·
speed tests can have multiple issues when testing. testing with the rogers server will tell you the speed to the server itself usually within the local network. those will usually give you the advertised speeds and sometimes could be skewed by whats called downstream or upstream burst. when you get a fast connection and they peg that as the speed.

the 30-40mb on an external server is quite good, but these are affected by time of day usage and test server load.

i hope that makes sense.

also make sure to test via a wired connection, wifi can have too many variables and the results can be all over the place.
 
#3 ·
greenshadow said:
... testing with the rogers server will tell you the speed to the server itself usually within the local network. those will usually give you the advertised speeds ...

the 30-40mb on an external server is quite good, but these are affected by time of day usage and test server load.

i hope that makes sense. ...
Thanks for the reply and, yes, it does make sense. :)

... also make sure to test via a wired connection, wifi can have too many variables and the results can be all over the place.
It seems strange to tie the advertised speed to a wired connection given the # of wireless devices in the average home (e.g., we have two laptops, two tablets, two smartphones and one "smart" BD player). :confused:

Anyway, thanks again for your input. :)
 
#5 · (Edited)
DSgamby said:
The advertised speed will also be the highest. ...
I was getting faster-than-advertised speeds last night. :D Whether the speeds remain consistent is another story...

... With Wireless, the speed can be affected by outside stuff including interference.
I can understand that if the receiving device is some distance from the transmitter, but my wife's laptop (for example) is less than 10' away from it. :confused:

Meanwhile, although my desktop is wired to a router in the upstairs office, that router connects wirelessly to the same transmitter (gateway) in the family room. And despite the wireless connection and the greater distance, I still get faster speeds than my wife does on her laptop. :confused:
 
#10 ·
The Speedtest at Xfinity is good as it tests IPV6 as well as IPv4. I have Rogers 100 Mbps and usually see about 140 Mbps down, 10 Mbps up on both protocols. All Xfinity's test servers are in the USA so the ping is a bit long at around 34 milliseconds.
I use Ethernet to a Dlink 868 and a Cgnacsmr modem in bridge mode.


Sent from my iPad using Tapatalk HD
 
#12 ·
Neither speedof.me or testmy.net had particularly accurate download speed measurements and fluctuated significantly between tests (anywhere from 120 Mbps and 270 Mbps). Rogers Speedcheck and Xfinity showed a steady download speed of ~315 Mbps. This matches up to my real world observations when I'm downloading a large file stored on a server with a fat, fat pipe.
 
#14 · (Edited)
What speeds are you getting with your wife's laptop hooked up directly to the modem? Are you using a good quality CAT5 or CAT6 cable? Any high CPU or disk activity when the test is run?

Also, download a speedcheck app for your mobile (Ookla has one) and see what it returns. Your mobile's hardware should be capable of handling 100 Mbps down for the test to be accurate.
 
#18 · (Edited)
Direct Connect to ROUTER...NOT via WiFi:

First time thru https://www.speedof.me HUNG trying to do UPLOAD test saying I needed to Disable ADBLOCKER for [JUST] their website....[Download Speed agreed with others listed below]. So I Disabled ADBLOCKER and Reloaded Webpage....and it immediately HUNG and could NOT be Reloaded and Restarted...perhaps it's waiting forever for their AD's to LOAD????? Based on other problems reported above, let's BLACKLIST it....

First time thru https://www.verizon.com/speedtest HUNG trying to measure UPLOAD Speed....Reruns worked, but DOWNLOAD Speed appears inflated by 2-10 Mbps and Real Time Activity Display seems very jerky. Also only ONE choice for Test Location. I do NOT Recommend it for use....

Three Tries....Three FAILS: https://www.cnet.com/internet-speed-test

============================================================
ALL of following reported ABOUT THE SAME D/L ~ 69 to 71 Mbps and U/L ~ 5 to 6 Mbps Speeds.
And slower [55-65 Mbps] when Test Site is further away:

Most Useful:
Speedtest.net by Ookla - The Global Broadband Speed Test [Many Locations to chose from.]
[I like their new I/F...and EASY to select far away WORLD locations to see my REAL WORLD I-N Speeds.]
COMCAST XFINITY Speed Test [Only a few locations....nearest 500-mi away.]
[NOTE DIFFERENCE: 70.9/5.9 Mbps for IPv4 and only 66.4/5.5 Mbps for IPv6]

Of lesser Utility:
https://www.speakeasy.net/speedtest [Only a few locations.]
https://www.megapath.com/speedtestplus [Only a few locations. Packet Loss Test Results said "UNKNOWN".]

Not very useful:
https://www.charter.com/browse/content/speed-test [MY NEW ISP: Only ONE location...500-mi away.]
https://www.timewarnercable.com/en/support/speed-test.html [MY ISP: Only ONE location...and they don't tell you WHERE.]
www.att.com/speedtest [Only ONE location...and they don't even tell you WHERE. Separate test for 300+ Mbps.]
 
#19 ·
I still get faster speeds than my wife does on her laptop.
That could easily be due to the WiFi chips in the laptop. Some devices are very limited in speed due to hardware restraints. Their built in antennas don't help either. It's not surprising that the router to router speed is good. They generally make a much higher speed connection than portable devices. Two high speed AC routers will outperform almost any portable device for WiFi throughput, especially if they are the same make and model.
 
#20 ·
^ This makes sense. :)

Last night, I tried hard-wiring my wife's laptop to the Rogers gateway but I couldn't get the connection to work. Same thing with my laptop. (Wireless connections were disabled on both laptops.) I'll have to tinker with some settings this weekend and try again.

I also ran the Xfinity speed test and got numbers comparable with Rogers' SpeedCheck.
 
This is an older thread, you may not receive a response, and could be reviving an old thread. Please consider creating a new thread.
Top