How to test your connection for a live-streamed workshop

Your internet speed looks fine, but a live workshop can still freeze, lose audio, or disconnect. A single speed result does not tell you how a connection behaves over time or while other people are online.

Use a repeatable test sequence to check upload and download speed, response time, jitter, and short interruptions. The Speed Test measures the first set of results, while connection and load tests help explain whether the connection is steady enough for your workshop.

Test the connection the workshop will use

Use the computer and network you plan to use for the workshop. Put the computer in its usual place, connect the same way you will on the day, and close video calls, downloads, cloud backups, and streaming on other devices if practical.
Open the Speed Test and run a baseline. Note the download speed, upload speed, ping, and jitter. A speed test exchanges test data with our server to measure the connection.
Repeat the speed test two more times, waiting about a minute between runs. Keep the device in the same place and use the same network. Write down the results instead of relying on the best one.
Open Connection Stability and let it run for a few minutes. Note whether response time changes, requests are lost, or the connection drops out.
Run the Bufferbloat Test to see whether response time rises while the connection is busy downloading or uploading. Compare the grade and delay change with your idle results.

If you will present over Wi-Fi, test on Wi-Fi rather than relying on a wired result. If the workshop will use a wired connection, test that same cable and port. A result from another computer or another room may not describe the connection you will actually use.

What the results mean

Upload speed is much lower than expected

Upload speed affects how well your camera and microphone can reach other people, especially if you are sharing slides or presenting video at the same time. Compare it with the workshop platform's current requirements for your chosen video quality. Leave room above the stated minimum because other devices and normal network changes can use part of your upload capacity.

Download speed matters when you receive other presenters' video, slides, and shared content. A high download result does not make up for a weak upload when you are the one presenting.

Ping is high or changes between runs

Ping is the time it takes a test request to reach a server and return. Higher or inconsistent ping can make interactive conversation feel delayed. The speed test measures a route to its server, not necessarily the exact route to your workshop platform, so use it as a baseline rather than a guarantee.

Jitter is high

Jitter means the delay varies from one request to another. Large changes can make live audio and video harder to deliver smoothly, even when the average ping seems reasonable. Compare several runs and check whether the variation appears at the same time each time.

Connection Stability shows lost requests or dropouts

Lost requests or gaps during a several-minute test suggest brief interruptions that a short speed test could miss. Repeat the stability check once with the network otherwise quiet. If the losses remain, test a wired connection or another location to help separate a Wi-Fi problem from a wider internet problem.

The Bufferbloat grade is low or delay rises sharply

Bufferbloat is extra delay that builds when a connection is busy. A download, upload, or backup can make a call feel laggy even if a speed test reports a good peak speed. A poor grade is a reason to test again while the usual household devices are active, then consider router settings or reducing competing traffic.

How to make the connection more reliable

Pause large transfers. Stop cloud backups, game downloads, and device updates during the workshop. If other people need the internet, ask them to avoid large uploads and downloads while you present.
Move closer to the router. For Wi-Fi, reduce the distance and the number of walls between the computer and router. On Windows 11, check the connected network under Settings › Network & internet. On macOS, check the selected network under System Settings › Wi-Fi. Browsers cannot read Wi-Fi signal strength in dBm, so judge Wi-Fi by the connection results and whether they improve in a different spot.
Try Ethernet. Connect the computer directly to the router with an Ethernet cable if possible, then repeat the same tests. If Ethernet is steady but Wi-Fi is not, focus on the wireless link, router placement, or interference.
Check for a VPN or proxy. A VPN can change the route and add delay. If your workplace allows it, compare results with and without the VPN, then follow the organization's rules for the workshop. Do not disable a required security connection.
Restart the router and computer. Restarting can clear a temporary fault. After both devices reconnect, repeat the tests from the same location and compare the results with your notes.
Contact your internet provider if the problem remains. Share the times of the tests, wired and Wi-Fi results, and any repeated dropouts. A single slow result is less useful than several consistent readings.

Test under realistic conditions

A quiet-network baseline answers whether the connection can perform when little else is happening. A second run during the time of day your workshop will take place reveals more about congestion. If several people will use the same connection, repeat the stability and bufferbloat checks while normal household or office traffic is active. Do not run a large transfer just to create load if it could disrupt someone else's work.

Browser tests have limits. They measure how this device and browser exchange data with a test server at that moment. They cannot guarantee the workshop platform's route, measure Wi-Fi signal strength directly, or predict every change in internet traffic. Treat repeatable results as evidence for planning, then join the platform's own test call or preview if it offers one.

Keep a short record of the device, connection type, room, time, and results so you can compare a change without guessing what was different.