The Internet fails us or a page does not work. The impression is that the connection is slower than it should and the first thing that many users do is measure the speed. The well-known speed tests that we have already talked about and that can come as a big surprise if we compare the results. Same router, same cable, same computer and disparate results. What are the reasons for these differences?
We cannot attribute the different values to a single reason that offers us a speed test according to the web platform that we use or even within the same measurement service. As a speedometer, a speed test indicates values of upload, download and latency in the connection and the results may vary. And in this post we will see what the reasons for disparate results may be.
Speed test, why and for what
A speed test is a test that we do using a web application (it is the most common) or an app that tests our connection and if it complies with what the operator promises us. We must also bear in mind that to take advantage of high speeds we must have a team that supports the connection and a related cable.
We start from the basis that we are going to carry out the tests with the same cable, same router and same device be it PC, tablet or mobile phone. These are all the parameters that we are going to see appear on the screen:
Latency (ping): with this value we measure the exact time it takes for a data packet to be transmitted, in this case from the device to the server. The name, ping, is the result obtained, a value that corresponds to the milliseconds it takes for the server and device to communicate. And it is an important value, because the lower latency the less time it takes to connect the server and the device, something that can be seen in high-quality video streaming or online games.
Supplier: Another aspect that we will see is the one related to the provider with whom we have contracted the line. Operator that may or may not be the same, especially if we have contracted a low cost dependent on a higher one (Movistar may appear if we have O2 or Orange if we have Jazztel, to name just two examples).
- Download speed: the first measurement that we will see, it measures the number of goals that the computer manages to download in a given time from the server. It is generally measured in megabytes per second (Mbps or megabits per second). Measures the number of megabits the connection can send in one second
- Upload speed: similar to the previous one, but now it is the time it takes for the file to be uploaded to the server. It is also measured in megabytes per second and in both cases the result is usually lower than what the operator promises. What the operator promises is theoretical speed and the closer the results are, the better.
How a speed test works
Once we have these data clear, we must see how a speed test works. Is about download and upload a package of files to a server in a specified time in order to calculate the speed of the connection.
And once we start the measurement, the first thing the platform we use does is determine which is the company that provides us with the Internet connection: the provider we saw earlier. We will see that servers appear that do not have to be nearby and vary depending on our location. Normally, the service automatically chooses the server closest to our position, although some allow it to be changed for another.
You will also provide us with information about the location of the server to which we connect. It is an important aspect to determine the reliability of the speed test, because the further away it is from our position, the worse the connection speed obtained in the test will be.
The next step, the next information that we will see on the screen, once the server is determined, is the latency measurement. It is the time it takes for a packet of files to be sent from the server to the user and vice versa and the result, pings, shows the time it takes to respond. The lower the value the better.
With the latency calculated, the upload and download processes are started. As speedometers we will see that there are maximum speed peaks, peaks, but in the end we will stay with a medium speed throughout the process.
These are all the data that we are going to see on the screen when we start a speed test.
Why do the results vary?
And at this point we obtain a result that can vary depending on the chosen platform and can even change in the same web service. What is the reason if the base is the same? Aspects that we can correct and others that cannot interfere.
The most compelling reason for the numbers to change is the location of the server and to test the images on these lines. In the measurement, only one value will be fixed in the equation: our location. It is important where the server is so it is interesting that the server is as close as possible. If in a test one is 1,000 kilometers further than the other, the values change. And in each test a different server is chosen, at a different distance.
But it is not the only reason that results may change. Thus, if in one test we have had more devices connected and in the other with less, this may affect, as the speed may be reduced due to greater consumption of the bandwidth of the connection to the server. It is ideal that we have almost no connected devices and that in both tests they are the same.
What’s more we will have to use a cable connection and not Wi-Fi. The Wi-Fi signal may weaken and consequently the connection speed may be reduced. And in case it is by cable, we will have to take into account the different categories that offer access at different speeds.
Category | Speed | Download speed |
---|---|---|
Ethernet Cat 5 | 100 Mbps | 15.5 MB / s |
Cat 5e Ethernet | 1,000 Mbps | 150.5 MB / s |
Cat 6 Ethernet | 1,000 Mbps | 150.5 MB / s |
Cat 6a Ethernet | 10,000 Mbps | 1,250 MB / s or 1.25 GB / s |
We can even use an adapter to connect an Ethernet cable and measure the wired connection from the mobile. And there we must also see the speed that the adapter supports.
It can also interfere with the bottom line the computer is performing other tasks in the background such as downloading files or playing a streaming. This can lead to different results.
What’s more there may be server congestion or saturation, which can cause the speed and the result to fluctuate. It is thus ideal to take several measurements and at various times of the day.
All this is up to us, except for possible server congestion. But it can, and is normal, that depending on the web service used the results under the same conditions, are different. Something that is completely normal.
If we want to use 1 Gb fiber we must have a PC or Mac with a Gigabit port and a cable that is at least category 5e
And is that not all speed tests work the same way. Some, for example, process multiple chunks of data simultaneously throughout the test, while others run connections in parallel. This can lead to a different end result.
Can even that the web browser and its add-ons and extensions may be interfering in some way or with the way the service performs the test.
In this sense, it must be clarified that regardless of our operator, practically any speed test can be used. And in order to have figures that are close to reality, take different measurements throughout different hours of the day.