There are several online speed tests available, and different testing methodologies – and some are more accurate than others.
What are you measuring ?
- Many of the online speed tests are based in the USA, so are measuring the speed of the total distance between the testing server in USA and your PC. The more “hops” between you and the testing server, the more potential problems you could be including in your results.
- Many of the speed tests only test the download (aka downstream) speed – which is usually much faster than the upstream speed. VoIP in particular requires adequate speed in both directions, with upstream often becoming the bottleneck.
- Note also that some devices/programs report bits-per-second (Kbps or Kb/s where the lower case “b” indicates bits), but other devices/programs report bytes-per-second (KBps with capital “B” indicating Bytes). There are 8 bits in one byte, however when you add packet headers and other overheads a 1/10 ratio provides a quick and easy ‘rule of thumb’.
- Also be aware that all the cable users in a street (actually on a segment of the cable) are sharing the cable’s bandwidth, and depending on what other users are doing the actual downstream speed might be only a small fraction of the maximum. The up side to cable is that when there are few users, the speed can be much higher than ADSL.
- There are locally-provided online speed tests for a few local ISPs but if you are with a different ISP we recommend timing uploads and downloads to your ISP’s FTP or web server is best for measuring the actual speed of your connection to your ISP. (Bearing in mind the Windows download speed indicator is notoriously inaccurate).
- Pick a file which will take a couple of minutes to download
- Download a file, then check the download manager for average speed.
- Upload a file, then check the download manager for average speed.
You may need to perform this procedure several times to get an average speed; and/or to test different conditions (such as with QoS off and QoS on).