Curious as to how ADS-B data is distributed and shared.
Appreciate the FA site says data may be delayed by 30 seconds or more but am wondering; do FA and Flightradar24 use the same data sources? Reason I ask is that Flightradar24 appears to be real time whereas FA is definitely delayed - unless of course viewing Skyview on my own R-pi.
This discrepancy came to light when I was speaking to a friend in US (I am in UK) and telling her when aircraft would pass overhead, the FA map was too delayed for this to be accurate whereas FR24 appeared to be real time.
I never checked against FA, but I did notice that FR24 is delayed compared to Planefinder. Whether itās deliberate, or just the way their web server flows, I donāt know.
If you have the US/Canada Radar (Live) option turned on in FR24, then that may explain the delay. Unless things have changed, the data from the FAA has a deliberate time delay.
I donāt remember if the US/Canada Radar was on or off at the time, but my observations were made while tracking my sonās flights across Europe. Nothing scientific.
Planefinder was consistently āmoreā real time, and the updates seemed more frequent than FR24. Also, Planefinder seemed to have better coverage at the airport level. I could actually track the planes from/to the gates.
I hope weāre not deviating too much from the acceptable topics on the FA forum.
FA is the only commercial site I currently feed data.
I live around 12 miles from Heathrow and viewing FAās Skyview is naturally spot on. When I recently spent a few weeks in California I took my Pi with me and fascinating to identify aircraft details then see them fly overhead.
I was merely thinking about scenarios where two enthusiasts in different parts of the World could both view live data.
My son is on the move again. This time Iām tracking his flight on FA, FR24, and Planefinder.
I was able to track at the airport level on both FR24 and Planefinder. The FA live map was not loaded at the time.
Planefinder is a split second ahead of FR24 on updates. I now think the difference between the two has to do with the web server flow, not deliberate delay, unless both are delaying by the same amount of time.
FA is behind both. The altitude is 2000 feet lower, and the distance traveled is around 10 km behind. After reaching cruising altitude, all 3 were the same, but FA is still 10 km behind on distance traveled.
The flight is about to land. The FA difference in distance traveled is likely related to the web page update frequency. It appears to update every 30 seconds or so. Also, the page was showing landed, while the live map still showed the plane approaching the runway.
Overall, they are all very good as far as basic tracking goes.
Iām simply using the data displayed on their respective web pages and live maps. Nothing scientific.
The time difference is due to FR24 using future projections and FA doesnāt. FR24 will send the last position (also with a 10-15 second delay) and then project the current plane location on the map about 15 seconds into the future to account for the delays. They are mostly estimating speed and direction. If the plane is turning you will see the plane jump around. This is especially noticeable around airports where airplanes are changing direction, altitude and speed.
FA show the actual position of the plane on the map. So the path is correct and will not jump around. The delay isnāt intentional but due to the time cost to process the data. You will also not see planes flying backwards or into buildings or other planes on FA.
This is an irritating āfeatureā in my opinion. It happens a lot with FR24. I just checked, and for some reason, estimation was enabled and set to 240 minutes. Iām pretty sure I disabled it before.
I do no see this much jumping around on Planefinder, and never saw anything like that with FA.
This is what I referred to as web server flow. Again, another guess in the ball park.
Iām on my way now to guess some lottery numbers for Wednesdayās draw.
Iāve been contributing to FA for many months now, ~2000+ flights/day on an outdoor antenna situated between SFO, OAK, SJC and right next to Googleās private airstrip at KNUQ. I also sent data to FR24, and could see a slight difference in the output from the respective websites and what I was feeding from dump1090 off my RPI system. I recently moved my feed, however, to adsbexchange who donāt take time to censor the feed of all the āspecial peopleā who want their privacy - so whatever adsb receiver sites put in comes right back out without any censorship within a second or two. Itās got to be the shortest delay of any site as a result, which is why I now do my MLAT solutions with their system having completed the migration this weekend. While any site is able to give you the air carriers flight locations, adsb exchange gives info on private (BARR LIST) planes that Iām doing the MLAT for when they squawk 1200/mode C flying through my coverage area (250 nm out). Iām curious what others have seen in feed delay there.
SkyView I am assuming is spot on as it is created locally using oneās own data from the tuner.
The delays on map locations I have not found an issue except when I am speaking live to a friend in US and I am in UK. As they do not have their own ADS-B setup I can only refer to map locations for aircraft and as such if I am letting them know of an approaching plane the FR24 map is more suited for this purpose.
I do however only feed data to FlightAware as I like the overall system and the friendly discussion forum.
Yes, all my references to live maps referred to the internet web maps of each respective tracking site, not the live map on oneās ads-b receiving station. The day a delay is introduced, by whatever means, on my own site, itās game over, Iām abandoning the hobby.
That and MLAT. Hopefully the friendly discussion will continue after 2020.
The 240 min projection has nothing to do with the short term projection. (the 240 min one is for transatlantic flights out of coverage where a great circle towards the destination is plotted at the last known ground speed)
I get the impression FA focuses on having all data displayed synchronous. So FAA data and MLAT data and ADS-B data are synchronized and for that purpose some data will need to be delayed.
This does not happen on fr24, so their mlat data and radar data regularly lags behind their ADS-B data. Also the vector prediction totally butchers radar data and mlat data to a certain degree.
Sorry to say when i listen to liveatc at some airport and want to āwatchā plane tracks i like the āless preciseā site much better so iām feeding both. Only thing i did was install NTP on my pi and redirect all the ntp pools so they donāt get burdened by that strange feeder of theirs.
Sometimes i think about just delaying the liveatc audio and using FA but even then the interface seems to be not quite as nice for that purpose.
The 240 minutes was the setting it defaulted to on its own. I was not tracking a transatlantic flight. The options for the user to select are: OFF, 5, 15, 30, 60, 120, and 240 minutes. I always select OFF.
Yeah and i said the short term projection david.baker talked about is a āfeatureā you canāt turn off on fr24.
The estimation you can set from 0 to 240 minutes is a different feature that you can turn off.
So the ājumping aroundā will not stop regardless of what that option is set to.
I hope this is clearer. (At least that is what i was saying and believe to be true i have not seen the code.)
What you can try though is to disable āAnimate Aircraftā itās in the āMapā options.
Yesā¦thatās another feature I normally disable. I donāt know if these settings are retained by way of cookies on the userās computer, or not. If they are, the may have expired, or deleted by the browser on closing, or by the security software.
I donāt visit these sites very often, and even less compare them. This was only the second time, maybe. I visit one or the other when there is a reason, such as checking somebodyās flight. They are good enough for this purpose, whether the planes jump around or not.
Business users may have other requirements, but Iām not one of those.
Thanks guys, this thread has been of interest to me as Iām not certain if sharing was allowed between the various sites due to the ābonusesā afforded to those that do share data with a certain server.
Iām assuming then there is no issue to whom you share your data with.
The ADS-B data are received by you, so you can do whatever you want with them.
Just the MLAT results you get from FA or others, those shouldnāt be shared with other sites.
For FA and adsbexchange that canāt happen because the feed clients donāt forward MLAT results.
Also if you donāt reconfigure your receiver MLAT results wonāt be even forwarded from it to the feed clients.
Thank you, every day is a learning day.
Iāll research a little more as to how my system is set up?
Of the two sites you mention, I have viewed both on many occasions before actually getting involved this July with my own hardware setup.
Iāts become apparent to me that the various servers donāt all see the same, such is the nature of the whole ADS_B setup and dependent on each sites ability to track high and low level aircraft unless your antenna has a good 360 degree āclearā view of the sky.
For a while I though local Mil traffic switched their ADS_B off as there was just no returns from them unless their over FL100, in which case, ads-b sites further away could pick up on these tracks, and only then did they become visible on the maps.
All changed now since my antenna is better placed, I now see more traffic from the South, but mil flying is still reduced here for now due to essential surface repairs.
So currently you probably have dump1090-fa as a decoder, then piaware as a feed client connects to that to get the data.
FAs Skyview is named confusingly as itās the webinterface of dump1090-fa ā¦
Well i shouldnāt complain tar1090 might also be confusing using the json files from either /run/dump1090-fa or /run/readsb to provide a webinterface
