Thank you for reply and all the information - sorry I hadn't meant to be critical of the site - more a reflection of my frustration.
I have read your reply and appreciate all the hard work you have put into investigating issues raised.- thank you. Will spend more time rereading your reply and hopefully at some stage all will make sense.
Just a note in the tests I selected Melbourne being my closest city while you have selected Sydney
Parado, sorry for not responding quicker. Ask regulars to the forum, I'm normally much faster at responding. I've had family visiting and now I'm out visiting family. Besides the few awesome, dedicated bad asses in the forum I'm the only one here. When I finally take a few minutes off each year, some questions may not get answered the way this forum intends to help people. I apologize because I saw and read your question as soon as it came in... I just wasn't able to respond. Couple that with my daily email regimen and it doesn't take long before your post/message/email gets piled on top of. It usually weighs on my mind and I do eventually get to many of the messages I may miss while I'm away but it can be extremely difficult (impossible) to get it perfect.
After looking at your results and reading your thread I don't see a problem. If TestMy.net is saying 8.1 Mbps... you're getting that speed. It's better than your supposed to get, right? TestMy.net is known by many to be the hardest speed test to get a favorable score on, this is by design. Tests (of any kind) shouldn't be easy if you're trying to gain useful information on a subject. Usually we hear it the other way around, "TestMy.net speed is much lower than the other speed test I use..." -- if iPrimus shows a lower speed on their test it could be due to the route to the server. Even though it's within your ISP doesn't mean that there isn't an issue with that server, the route to that server or the test itself.
I just did the iPrimus, Telstra and TestMy.net speed tests all against Sydney AU. (not my normal connection but it still proves my point)
Telstra was funny, even though for a large portion of the test my computer was reading only 8 Mbps... then down to far less (trailed off HUGE from the middle to the end), the test itself never indicated any of that happened. It just ignored the bad parts of the result, as ookla speed tests are well known to do. If we were graded like that in school, everyone would have doctorates and would learn nothing.
iPrimus -- looks exactly the same as Telstra's test, they're both ookla tests. They both have similar designs and server selection. But iPrimus scored far below the Telstra speed test result.
TestMy.net is exactly as I would expect. Shows everything that I saw on the reading coming off my computer while it was happening. Sorry it doesn't make people feel as good or want to share their scores as much... but it's the truth. I'm here to help you with the truth, sometimes it hurts. In those instances, if you listen, it will help. According to the true test, the connection I'm on doesn't fair as well to Australia as the other guys would want me to believe.
Even a multithread speed test on TestMy.net (targeting only Sydney, AU) can only achieve 25 Mbps on this connection.
That's all she's got!
Funny thing is, after the initial tests I ran both of those tests (iPrimus and Telstra) at the exact same time... and they returned impossible results. Somehow I showed more bandwidth on the Telstra test while the other test was running. Think about this -- I drew more bandwidth, away from the test... yet the test was able to display significantly more bandwidth than I was supposedly capable of in the original test result.
What?! A combined 44 Mbps against Australia on a US connection that has a hard time pulling 25 Mbps out of Dallas. Uh, no. Do the same on TestMy.net and you'll see that the speed is split between the two results... aggregate the scores and you'll arrive at the same speed you'd expect to see off a single test. This is what any tech would expect to see in that situation, because TMN is a real test and has no vested interest in your score being one way or the other.
Here's a Dallas result just now on this connection.
Sure, I hit 40 Mbps... briefly. Would you call that 40 Mbps or anywhere close? Other tests algorithms may make it appear that way... their motto - keep the best, ignore the rest.
So, that iPrimus test also performed far lower on my end. Not a reliable resource. My main point however is that you really should stop wasting your time with all the BS speed tests out there. You've finally found the real one.
Parado got a reaction from mudmanc4 in Speed test providers variations