Speed test result vs actual download speed

[gcosta56]

Hello,

I installed my Starlink antenna a few days ago to eventually replace my sluggish ADSL, but I'm encountering a bandwidth issue and I'm wondering if any of you have faced this problem.

In summary, my bandwidth as estimated by services like speedtest.net or Starlink's internal app ranges between 150 and 280 Mbps.
However, all my real download tests are capped at 4.5 MB/s (36 Mbps).

My antenna is installed on the roof with a clear view to the north.
To eliminate any issues with my hardware, I've just reset the Starlink router. I've separated the 2.4 GHz and 5 GHz networks. The Ethernet adapter is connected.

1st test: I connect my phone to the 5 GHz network. The Starlink app test gives me good speed. The same goes for advanced mode (see photos).
I do a download test on a friend's FTP server who has fiber, and I get a speed lower than 4 MB/s (less than 32 Mbps, see photo).

2nd test: I connect a laptop to the Starlink Ethernet adapter and redo a speed test on speedtest.net (speed greater than 200 Mbps), then a download test in FTP and also in HTTP on test-debit.free.fr. Here too, my real speeds are capped at 4.5 MB/s.

For now, Starlink Support isn't helping much by suggesting I switch to a priority subscription.
In fact, I'm just looking to have a real bandwidth that is roughly consistent with what I observe with their measurement tool.
Have any of you noticed such a phenomenon?

Thanks in advance

[RicochetStarlink]

Yes. We have. All of us... and there's no fixing it. That's just the way it is...

TL; DR;

Internet speed tests (e.g., Ookla Speed Test) are generally pretty accurate. They measure the maximum bandwidth available at a given moment.

However, the actual speed that you observe with a particular transfer operation is only partially dependent on the available bandwidth. In fact, the transfer speed is affected by an enormous number of factors.

• The capacity of the server that is transmitting the data
• The speed of the Internet connection that that server uses
• The distance that the data is traveling on the Internet
• Network congestion in...
  • The server itself
  • The server's Internet connection
  • The Internet in general
  • Your Internet connection
• Anything else affecting performance on either end of the connection... or in the middle

All this translates down to two basic things

1. Actual transfer speed will NEVER be as high as measured transfer speed. Sometimes it will be a small percentage. That's just the way it is.
2. Actual transfer speed at a given moment during a specific transfer activity will be different than the transfer speed experienced on a previous or subsequent moment during the exact same transfer activity.

...which leads to the answer to your original question.

> "In fact, I'm just looking to have a real bandwidth that is roughly consistent with what I observe with their measurement tool.
> Have any of you noticed such a phenomenon?"

Yes. We have. All of us... and there's no fixing it. That's just the way it is...