I just set up an rpi and started tracking yesterday (LGAT-136481). I need to experiment with antennas and antenna placements, so I’ve been watching my local feed closely and comparing it to the FA and FR24 public sites, so that I can see the aircraft that my feed doesn’t see and draw conclusions about the antenna. That’s how I noticed something weird: my feed sees a lot of traffic that never shows up on the FA website.
First thing, I checked the feed. Sure, I see the traffic, but am I actually forwarding it to FA? The answer was categorically yes. Next question, is it “odd” traffic, e.g. military, unknown identity etc? Yes, sometimes, but most of the time it’s airliners on scheduled flights, I have consistent and regular messages from them (i.e. mesg=(hundreds) and age=0), I can see them on FR24, but they still don’t show up on FA.
So to the question: How come? Are any feed data intentionally filtered, or how else can flights not show up on FA, although a feeder has seen them and fed them to FA?
Military have the ability to disable tracking by civilians, for obvious reasons.
There are also agreements regarding “do not track” aircraft in various countries. Some feeders adhere to these agreements, others don’t.
Additionally, there are also temporary random ADS-B IDs for those that want a little more privacy.
Note that outside of controlled airspace, many aircraft are not require to transmit any transponder signal.
@Jon: Yes, all those are valid cases, but I am talking about airliners on scheduled flights showing up on my feed and not on FA. But now I have a hunch of what might be going on.
I was just watching RYR1991 pass south of Athens. It appeared both on my feed and on FA, but on my feed it was some 35nm ahead of FA. Obviously my feed displays data in real time, while the FA display is supposed to be about one to one and a half minute behind real time. However, with this flight going at 448kt, the time lapse between my feed and FA was a whopping 4 min 42 sec, maybe give or take a few seconds for the difference between airspeed and ground speed[*].
With such a delay, it is possible that the flights I saw on my feed but did not see on FA, simply appeared on FA long after I stopped looking for them. I have to check again tomorrow when there is more traffic.
Hold on, I have an invisible flight right now (2020-09-18, 21:08Z): AZA3WU (AZ722, ICAO 4ca9bd). I just saw it on my feed pass about 2nm north of LGAV on a 92-some heading, then lost it. FR24 shows it made a long left turn over the Gulf of Evia to final on LGAV 21R. On the FA front page map it is nowhere to be seen; neither in its real position nor lagging anywhere behind. Then again, if I search for it, it is there with its full track as a live flight. And it has also appeared in the flight’s history.
Conclusion: the feed data is recorded correctly, but the map on the FA index page is faulty in that it does not show all recorded flights within its boundaries.
.
[*] Does “Speed” on the dump1090-fa web interface and on the FA website refer to air or to ground speed?
@obj: Thanks for the clarification. On the subject of flights not showing up on the FA main page, have I found a bug or is there some intentionality involved there? That AZ flight last night was really nowhere, even not when I zoomed out to include its entire flight path on the FA map. I did that while it was on its downwind leg, so still very much airborne and should have shown up on the map.
I don’t have much knowledge of the workings of that bit of the website, but if it’s tracking OK on the underlying flight page then I’d guess it’s just a display problem specific to the front page map. Maybe see if refreshing the map helps when you next see it.
The second paragraph of my original post says “most of the time it’s airliners on scheduled flights”. That’s a verbatim quote, not a paraphrasis. Of course I could have typed it twice in bold all caps, but I did not anticipate someone who missed it not once, but twice, to then also complain about my lack of clarity. I apologise for this lack of foresight on my part.
Easy, Jon is one of the good guys and was attempting to assist. You’ve been provided several things that could help explain what you are seeing (or not seeing) - perhaps there is a bug. What more do you want?
That’s what it has to be, but it can be just anywhere between the raw data database, the webserver and the javascript in the browser, so only someone with access to all three parts and everything in-between them can do any meaningful troubleshooting.
I did last night; it was the first thing that crossed my mind before I went on to document the event in my original post. It’s a pity that the flight landed a couple of minutes later; had it been in transit, I would have had more time to compare its live status page against the FA index page and play with reloding and zooming in and out. Next time I see it I’m better prepared.
If you have that installed, you will be able to go through the flight history of your own installation to compare against the main site. That way you wont need to scramble to do it in real-time.
I don’t doubt he’s a good guy who’s only trying to help, but so am I.
The only valid and relevant explanation so far is obj’s “display problem specific to the front page map”. And yes, it looks very much like a bug. What more I want is to either rule out a bug or find it and fix it, which can be done if we play together and try to verify and document the bug’s existence. I mean, as long as I am the only one to have noticed this, it could be down to my browser or my network. If others can see aircraft on their local feeds that do not appear on the FA main page map, we would have a confirmation, and then it would make more sense for FA to spend time debugging it.
I had no idea about it; thank you for the tip. Now I have it installed. It works as advertised and it looks good.
Nevertheless, what I need to compare when I notice a “ghost flight” is the map on the FA main page with that flight’s live page on FA itself. Only the latter can guarantee that FA has actually received the flight’s data and is therefore able to display the flight on the main map. Comparing against my own flight history would leave the possibility open that my feeder captured the data, but FA never received it (let’s not forget the feed is UDP).
