jrtcjrjr

Here is an odd case... I am getting pretty much the same speed scores when testing with OOKLA and with testMy.net..... Something has changed. Has the OOKLA methodology been corrected?

I just tried (and tried again). Using the Express Download (~25mb) I get a message, waiting for testmy.net. My Internet is working fine. While the testmy.net waits, I opened another tab and conducted a WOW speed test. After it finished the testmy.net page was still waiting. Any ideas of what I might be doing wrong here?

Sorry it has taken so long to get back to this... In response to: What percentage does it show when it hangs? What's the test size? I honestly don't know the size. The first two passes (very small sizes) fly by so quick I cannot read then. The third hangs with a simple message about preparing... I will try again and see if I can get more information. The hang-up seems to be in the file preparation phase as opposed to the download phase. I didn't mention this earlier, I am having this problem using Chrome and OSX. Using Chrome and Windows 10 I do not have any problem. My Apple computer is connected via Ethernet but the Windows computer is wireless. Using my wireless router there has always been a significant degradation in speed compared to the wired connection.

After my ISP (WOW) upped my throughput I haven't been able to finish a testmy.net download test. Two small downloads complete in the blink of an eye but then on the third, larger download the test hangs. Any suggestions? I used to test frequently but stopped. After WOW bumped (doubled) the speed I thought I would come back to verify. WOW's speed test shows something near the expected result but I've been unable to verify using testmy.net. Thanks John

From the Speedtest.net site: Download Speed Your computer downloads small binary files from the web server to the client, and we measure that download to estimate the connection speed. Based off this result, we choose how much data to download for the real test. Our goal is to pick the right amount of data that you can download in 10 seconds, ensuring we get enough for an accurate result, but not take too long. We prevent caches from throwing off results by appending random strings to each download. Once we start downloading, we use up to four HTTP threads to saturate your connection and get an accurate measurement. Throughput samples are received at up to 30 times per second. These samples are then aggregated into 20 slices (each being 5% of the samples). The fastest 10% and slowest 30% of the slices are then discarded. The remaining slices are averaged together to determine the final result. Why do we discard certain results? We want to ensure we're giving you the most accurate assessment of your connection's maximum sustained throughput. Here's how we do that: Outlying 10%: Since we're measuring data transported over HTTP (via Flash), speed can be affected by a few things, such as potential protocol overhead; buffering due to the many layers between our application and the raw data transfer; or throughput bursting due primarily to CPU usage. To account for these variables, we initially drop the top 10% and bottom 10% of our slices as outliers. Test Ramp-Up Period: We keep the default test length short to improve user experience, but in doing so, the ramp-up period can take up a significant portion of the beginning of the test. In consideration of that, we also drop another 20% of the bottom result slices. ----- Speedtest is being much more direct in explaining how the results are calculated - as compared to only a few months ago. Good for them. My observations from comparing Speedtest.net and testmy.net are first, the results cannot be rationally compared. Testmy.net results are calculated using the complete test while speedtest.net results are form a 60% slice of the total test. I also see a difference in how Speedtest is (seemingly) implemented by various ISPs. Speedtest.net tests are observed to take from 3 to 8 seconds depending on which ISP hosted test I use. As has been already noted, there is typically a surge in speed during the first few seconds of download. A Speedtest result that is obtained after only a few seconds provides me with a result which averages 150% of the throughput I pay for. That speed in not sustained, however. Any test, testmy.net or Speedtest.net that takes 7 seconds or more shows the result of speed settling back to or below my paid for throughput. My opinion is that various uses of the Internet impact how we view speed. Using testmy.net I can see the chart illustrating the ramp-up of connection - the subsequent speed burst - the typically sustained throughput - and finally, the tailing off as the download is completed. The early burst in speed is certainly appreciated when simply browsing or checking email. The snappiness provides a warm and fuzzy feeling. When I am streaming video for an extended period then the sustained throughput at the speed I pay for is important. It is no wonder that most ISPs use Speedtest.net as their official throughput test. It usually makes them look good.

It would be a good idea to connect your PS4 via Ethernet and check the throughput. In my home I can see throughput degradation ranging from 10% - 50% depending on where I make the connection. Sitting next to my cable modem/router I can achieve 90% - 100% of throughput (compared to a wired connection). A wall, a floor, and/or other wireless devices such as telephones can cause serious degradation of the signal - and throughput. You might also want to use a wired connection to baseline test your PC and/or laptop so as to eliminate any PS4 issue. Using the process of elimination it becomes easier to identify where the problem lies. If you eliminate potential problems on your side of the ISPs modem it is always easier to receive support as opposed to excuses. Speedtest.net uses the OOKLA flash-based Internet speed test. OOKLA has a unique way of measuring speed. First, they drop the bottom 10% and top 10% of throughput measurement as they perceive those to be outliers. Then they drop another 20% of the bottom as they consider it represents a ramp-up. Effectively, they are reporting back on 60% of the time the test is run. Also, the Flash-based tests are relatively short duration. When I run a large file test here (testmy.net) I see a ramp up, a surge will above the speed I pay for, and then a fairly consistent throughout at the expected throughput, followed of course by a ramp-down. Running a small-file test (akin to the size test Speedtest.net utilizes) I can get exaggerated throughput results - much like I can see from Speedtest.net. If a Internet user is consistently uploading/downloading short bursts then the Speedtest.net would seem correct or accurate. On the other hand if you are streaming video it is more accurate to look for the throughput that is sustained - as you might see when doing a large file test on testmy.net.

I am watching for degradation in throughput. Right now, my throughput is usually 10-15% above my plan rate when auto testing and 2.5-5% above plan when testing large files - when I have no other devices/people taxing the system... Trying to convince any of the ISP techies that I have a problem (or potential problem) is a mountain not worth climbing, yet.

It might be worth your time to research OOKLA and how they measure throughput. OOKLA (Speedtest uses the OOKLA FLASH based speed test) drops out of the calculation the top and bottom 10% - claiming they represent outliers. Then they drop another 20% off the bottom - assuming those values represent ramp-up. That leaves the 30% to 90% group for the measurement. I find the testmy.net large file size test to provide the most representative evaluation of my service provider. I see the ramp-up, then the surge (that my provider utilizes) and then the leveling off at my paid for throughput rate. Aside from all that it seems bad form to begin a conversation by calling people liars. OOKLA is showing the speed during what they perceive as the optimum 60% of the throughput speed. I suspect you are comparing apples and oranges.

I like to run both the auto-size tests and the large file size tests. Sure the auto-size tests provide a quicker result but there is something to learn from the large file test. My 50/5 plan is typically exceeded by 10-15% increase above expected levels when running the auto-sized tests. That gives me a 'feel-good' while the large file tests show my IP surges throughput for a short duration of 5-7 seconds and then settles into... my plan rate of 50/5. If the household is simply doing a bunch of browsing the surge response gives everyone a sense of great performance - better than expected. If there are two or more HD video streams the consistent 50 Mbps download is more than sufficient. Until late afternoon and into the evening, that is. Lots of (WoW) customers in the neighborhood likely the cause for the drop in performance. I also find it useful to run tests when connected via ethernet as well as wireless. With wireless my throughput drops 10-20% even when I am only few feet from the router. Tracking down interference sources has proven difficult but I have not given up.

You have my thanks. Using testmy.net I have been able to isolate recurring problems with my Internet connection. This was something I could not accomplish using the more common FLASH based Internet speed test tools that are so commonly used. I imagine there may be other small issues but the big problem of inconsistency was identified as a cable splice adding a couple feet of coax cable on the end of the line drop to my house. As I had reported, the splice was left exposed to the environment and when weather conditions were poor (wind and rain), my connections suffered with a 15 - 40% degradation. I applied a temporary weather proofing over the spliced area and today, with wind a rain, I have been able to maintain a consistent D/L speed. Again, thank you. This is a good and useful tool.

Sorry I haven't been able to post a picture of the cable... I know there is a video camera around here somewhere. Related comment about the cable splice... Our weather has improved, the sun is out, and my D/L speed is back in the ~50-55 Mbps range as compared to wet conditions when it drops to ~25-40 mbps. That is with all the toys and devices connected and with what I might call a normal load. This morning I put the problem cable area under shelter and as we expect some rain tomorrow, I should be able to tell if the problem is a poor fitting. I do see recurring patterns in D/L throughput. More often than not when running a normal test the throughput drops significantly at about 5 to 5-1/2 seconds into the download, then picks back up. I also ran a few 100 MB and 200 MB D/L tests and in each saw a pattern of speed ramping up to ~50-55 Mbps and then dropping after about 5-7 seconds and then holding steady at a hair under 40 Mbps for the test duration. Being the suspicious type, I wonder if the ISP throttles back at a point where they think most speed tests are completed. Using the ISPs OOKLA speed test my connection 'appears' to get near 100 Mbps on D/L and I am convinced that is not true. If it were true I do not think I would ever see a pause when streaming video, certainly not to the extent where the video provider is recommending dropping down in quality. I've not mentioned it before but I do like your site. I understand why the ISP wants to judge speed only on their WAN as it is what they can directly control. But, you make the correct assumption in stating we buy Internet service for the peer to peer functionality/capability.

I do have a busy and complex home network. (That is my excuse for running so many tests. I have tried by the process of elimination to identify the home grown problems.) I have also tried using a java based Internet speed test. It was closer in line with testmy.net than the many Flash based tools. I have discovered a uncovered splice in the inbound cable a foot from the house. I have always been aware that Internet speed suffered on rainy days, like today. The test scores today are 15-20% lower than what I received before the rain came. Oddly, the rain seems to have no impact on the OOKLA Flash based test scores. Thank you for your response. I will try the MercurY test.

Mine is likely a dumb question... Why are Flash Internet tests inaccurate? I do not doubt they are. Right now, several such tests claim my connection down is 80-95 Mbps and testmy.net is showing ~40 Mbps. Judging from video streaming, I believe the 40 Mbps. And so, my question - why are the Flash tests so wildly inaccurate? Given your experience, I suspect you know the answer.