Post by: Larry on September 07, 2014, 07:36:30 AM
If I use my Airport Express: I have a speed test of download speed of 28.82mb/s and an upload speed of 9.24 mb/s in the back bedroom, and in the back of our salon the download speed is 31.20 mb/s and upload speed of 10.14 mb/s - beside the router it is (Site 3) 23.66 mb/s and 9.68 upload.
We use super fast fibre to connect - in London though we had 76mb/s download, and we miss it.
Post by: Xairbusdriver on September 07, 2014, 02:31:51 PM
And how often do you get the same values, ±5%? Speed tests are fun but the results are rarely duplicated. There are so many variables even in your own LAN that they are good for general, long-term service estimations. They are generally not much good for claiming any records! 
Then, there are the various 'engines' doing the measuring; some use Java, some use Flash®, some use neither, some you may not be able to determine. Next comes selecting the server to download from and upload to; how are those selected? By you or the software? If it's done by the software, are those servers selected to show the best, fastest speeds? Would that be the tests done by your ISP? 
Another variable is the size and load on your particular 'wide-area-network'. Over here, the cable companies setup networks mainly lumping together several houses on a street or block. If all the users in all those houses are surfing at the same time, everyones speeds will be lower. That's one reason cable companies don't guarantee up/download speeds. It depends on the time of day, the data being used (multi-player games, streaming video, email, etc.), and the phase of the Moon. OK, maybe the Moon's phase is not involved, but I don't think you can prove it's not!
Most of the time, we only notice when the speed slows enough to get some spinning 'pizza's' or some stuttering videos. Other than that, we just adapt to what we get, even if it is seriously over-priced! I just deleted some screen shots of speed tests from the cable company and a company that supplies the software to them. Even though the tests were done within 15 minutes of each other, the numbers were much different. Can't even remember which was faster.
<This Wikipedia page> has a nice chart ("Bit rate (Mbit/s)") showing the 'stated' speeds of the different WiFi methods usually available in your business/house; look near the bottom (of the unsorted list) to see the 802.11... speeds. Note that, until 802.11ac, Gigabit Ethernet, which has been around for decades, was always faster than WiFi!
An aside:
FIOS is great, if/when it becomes available. A good friend was using DSL and complained for weeks that he was getting only ~1.5Mps when he had been paying for 3Mps. The last technician asked why he was paying for 3Mps since he was almost 11 miles from the switch?! The 'normally accepted' distance is 10 miles and then only 1.5Mps is guarantees! The friendly phone company had called a few weeks earlier and asked my friend if he knew that FIOS was available in his "neighborhood"? Well, no. How much would that cost, anyway? Of course, the cost would be more but my friend decided to go for it! That last technician I mentioned also advised my friend that, yes, FIOS was in the "neighborhood" but it certainly wasn't planed any streets even close to my friends house! Surely a simple, sales-promotion, misunderstanding...
My friend canceled the phone company completely and is now surfing with the cable company. "Fool me once, shame on me. Fool me twice, shame on you!"
Unfortunately, the cable company is also a monopoly, so enjoy the speed while you can. That's my advice to him, anyway.Maybe you could move to London? I hear it's no longer the most expensive city in the world to live in!
Post by: Paddy on September 09, 2014, 09:36:42 PM
Post by: Larry on September 13, 2014, 01:28:48 AM
QUOTE(Xairbusdriver @ Sep 7 2014, 08:31 PM) <{POST_SNAPBACK}>
And how often do you get the same values, ±5%? Speed tests are fun but the results are rarely duplicated. There are so many variables even in your own LAN that they are good for general, long-term service estimations. They are generally not much good for claiming any records! Then, there are the various 'engines' doing the measuring; some use Java, some use Flash®, some use neither, some you may not be able to determine. Next comes selecting the server to download from and upload to; how are those selected? By you or the software? If it's done by the software, are those servers selected to show the best, fastest speeds? Would that be the tests done by your ISP? The fact remains with two of us using the net at the same time - in the evening the service is diabolical upstairs, and all the soap from Utility Warehouse is just a mirage. Can't wait to leave them but tied into contract for another year.
Post by: Xairbusdriver on September 13, 2014, 10:01:36 AM
QUOTE
in the evening the service is diabolical upstairs
Are you saying the speed is better if the only variable you change is the time of day? In other words, the speed "upstairs" is better during the daytime even with both of you using the same web sites/apps.My only thought is that there are more folks/neighbors on your local network at night, probably including more kids home from school. If your connection is not directly to the ISP, you'll be sharing it with several others. Each person connected can cause the speeds to slow for everyone. At least that's the way it happens with a shared connect as our cable service is wired.
Speeds can also decrease when there are "dirty" connections, even if they are not at your house. A "dirty" connection will cause data packets to be re-transmitted which slows the actual down/up loads to take longer. Same thing can happen if your WiFi links is weak; many data packets will need re-transmitions and the result is a slow down. You may not see this slowdown when simply moving a file, but it may be painfully obvious with a streaming video when the picture freezes or pixelates.
WiFi is convenient since it doesn't need running wires around the house. The disadvantage is that its speeds are not usually as fast because of weaker radio signals; they drop off rather quickly with increased distance. Then there are the barriers of walls, especially those with metal inside. A floor, of course, is simply a much thicker "wall" and often have metal pipes and ducts. Those kinds of "disruptions" won't be helped by faster connections to your house, unfortunately.