I’m finding the Observations very slow to load lately and i was just wondering if it’s a network issue or server overload.
What would be the chances of putting a button on the observation page to go right to the ones that need Vetting for signals, that way you would eliminate the need to load all of the observations only to filter out the Un-Vetted ones.
Just a thought.
Karl.
From time to time there are performance issues, we doing our best to solve them.
You can filter the ones that need vetting, in the observations page by hitting the Show More Filters button:
And in the Rated Artifacts filter, choosing the option Non Rated:
This way it will show the observations with waterfalls that haven’t been vetted.
NOTE 1: There is a rate for the observation in general and a rate for the waterfall. Observation is rated automatically as unknown(Non Rated), good, bad or failed based on some parameters, while waterfall is rated manually as unknown(Non Rated), with signal(of the satellite), without signal(of the satellite)
NOTE 2: Some observations are rated automatically as good when they have a specific mode(modulation), for example FSK, and there are frames uploaded from the station for this observation. In the majority of the cases this automated rating is correct but there are cases of “noise frames” where the uploaded frames are demodulated noise and not from satellite signals. When you filter with Non Rated these good-rated observations will also show up as they don’t have a manually rated waterfall. If you don’t want to vet these, you can filter them with the status filter by disabling the good observations:
Use also the rest of the filters to load quicker the observations page, for example filter by Satellite or by Start/End Time, for example if you have completed vetting until now (2024-11-15 8:00 UTC), then use that datetime in the Start Time filter in combination with the others ones for loading only the newer ones to vet.
NOTE 1 As in most cases it takes around 10-15 minutes for an observation to complete, in the example above I would set the Start Time filter to 2024-11-15 7:40 UTC (20 minutes before the datetime) as there are maybe some observations not completed or without uploaded data.
Here are three techniques I follow for quicker and easier massive (re-)vetting:
Satellite filter. This makes it easier as in each waterfall you expect to see a specific type/format of signal and if there are multiple observations from different stations at the same period you can expect the signal to be almost at the same place in waterfall. If the satellite has different transmissions you can go one step further and filter them either by Transmitter Mode or by Transmitter uuid:
(Re-)Vet per observation status, choose the status/rate of the observation as shown in the screenshot above. In this case you can expect consistent results in case of good or bad rated observations and you are prepared to find the false rated ones and just change the status to opposite. For example if you filter to re-vet only the good ones, you expect all the waterfalls to have signal, so you are prepared, while using the arrow keys on the keyboard to navigate through the observations, to hit the “b” button for rating the bad ones in case there was a false rating, similarly for the bad rated ones and “g” button for changing the rate to good.
Similarly to the second technique, in case of re-vetting good rated observations I use also the Results filter to separate the good-rated observations that have data uploaded from the ones that don’t:
Thank You Fredy. I’ll have a play around and see which one works the best for me. 