Troubleshooting 'Other' data

Well the comparison was between keithma, suffolk3000 and myself, who are all close enough to overlap and largely see similar traffic. The difference is he is on the coast in an area with flat terrain and no obstructions and is able to take advantage of it with a pretty much ideal installation. I’m terrain limited in most directions albeit not too badly, whereas he is horizon limited. Here’s a direct comparison:

I don’t know what “other” actually represents on that table, since by definition they don’t have positions associated - if they did they would be either under mlat, or ads-b. The leaderboard is ranked by unique aircraft seen in a day not positions, so I would expect him to see more simply by covering a larger area. It’s probably not valid to compare messages received between a US and European receiver since there are so many more messages transmitted in Europe.

There are not too many planes left that didn’t upgrade to ADS-B.

And yet you have more far away planes.

I have two receivers, you’re looking at the one which is in my loft using a Pi3B, a homebrew aerial and a FA dongle. This is my new one.

Your ‘other’ positions are 22% of your total. Looking at mine right now, it’s 31% of total and I’ve not even started tweaking yet. That’s with -w 5.

I think this being almost completely green is good, yes?

Oh, I got no problem with those, if you have 31% more power to you. My observation was more of an academic note, nothing personal.
I just don’t know how useful or not those “other” are to FA. Maybe they are…

Nope
caius
caius

me
keithma

I’m as keen as anyone else to get rid of the ‘other’ figures. I don’t see how they can be useful at all.

Planes, not positions.

Hmm, how are you determining that? Not sure where that can be seen.

It’s OK, I didn’t take it personally - We’re all trying to squeeze the best performance out of our systems so all comments are useful. If I can do anything to improve the signals I’m receiving, I’ll do it.

There’s still plenty of GA aircraft here that do not have ADS-B. Remember that UAS doesn’t exist in Europe at present, so most of those aircraft are mode s only. There are still a noticeable proportion of commercial flights that are not ads-b equipped yet as well, though the number is steadily reducing.

Yes - it means that there are virtually no single message tracks making it past the filter. That’s good because the likelihood of receiving only a single message from an aircraft is fairly small. If the graph is that clean with w 5, you won’t gain anything by increasing w further.

I found that I still get some single tracks with -w 4. Here’s what it looked like earlier today when I tried it for a bit:

1 Like

Apologies, I miscalculated both our figures (I blame being up since 03:30 this morning).
You’re on 18% and I’m on 23% ‘other’ positions. That’s better.

I can get more if needed :innocent:

An interesting thing I have noticed with “other” aircraft…
I’m in a terrain constrained area, and I get what appear to be partial decodes from aircraft quite a bit.
While staring at tar1090, I have noticed Other planes, that are on the edge of my receiving range pop up either with a position, or just some info, only to show up completely valid a few minutes later as they enter the range.
I’ve seen this behavior quite a few times, especially obvious when I get a position hit in purple and the plane shows up on the map, or an altitude, and then a plane comes into view from that direction, altitude is the same, or the like, turns blue and shows as normal adsb traffic.
I don’t know if these planes are transmitting Mode-S, ADSB, or UAT, all at once, or if partial adsb decodes also show up as Other.
They are labeled as Mode-S, ADSB Version none, but suddenly aren’t Mode-S once they are in range.
Maybe they transmit in some other mode until they are near the airport?
I really have no clue what is going on.
I can only assume I’m getting a partial signal reflection, or bounced signal.

“Other” positions have nothing to do with single messages from my undrstanding.

I have a few single track messages, but i am good with my setup. I am using the blue FA stick. Maybe gain is a bit too high, but with an amount of 5% of messages > -3dBFS i am satisfied.


“Other” aircraft are ones which don’t have a position, so it’s common for aircraft to appear as mode-s only when they first come into range. If they have ADS-B, then quickly they will get position data and change. If they don’t, then there may be a short delay before they get a position from mlat.

It’s possible that they remain as mode-s only: perhaps there aren’t enough mlat stations recieving it simultaneously, it’s unable to calculate a fix for some other reason. These are still counted as “other”, but are completely legitimate. They are real planes, just without a position.

The problem is that bogus data also gets counted as “other”, because it counts ICAOs without a position. These never get updated with an mlat or ADS-B position, and they inflate the stats on the FA page because they aren’t filtered out.

It’s quite unlikely, but not impossible that you will receive only one message from any particular real plane, because there transmit so many. If you get one, and it’s real then you are highly likely to get more. This means that the majority of single message tracks are not real, which is why minimising them is good.

3 Likes

I seem to be getting considerably poorer performance since cutting the antenna (i.e. from 22/11)

It’s Thanksgiving traffic maybe.

Also, did you put a can below it as shown in the thread?
That’s an important factor as well.

Can you show a photo of where you have placed the antenna?

Install the graphs for some more data on signal strength and other stuff: GitHub - wiedehopf/graphs1090: Graphs for readsb / dump1090-fa / dump1090 (based on dump1090-tools by mutability)

I’m not convinced it is much worse.
If you compare with your nearest stations they show the same sort of drop after the 22nd.
Looks more like winter traffic levels.

This is before I cut the antenna, but haven’t changed anything else.

For the graphs1090 data: https://imgur.com/a/43SpVOj

That’s a helpful observation - thank you.

You can cut the can as well to around 65 mm so the whole length from top to bottom is 136 mm.
(careful it will have very sharp edges if you cut it)

In that location every bit you can go further up probably helps, so propping it up higher would be a good idea.

I’ve looked at your range and it indeed seems like it decreased a bit after you cut the antenna.
Antennas can be very tricky.
You didn’t change the gain did you? Should be at -10 with this setup.

Put the coax cable away from the power supply cable. The shielding on these very cheap coax cables probably isn’t too good.

Your main problem is obviously the antenna location, just not very much to work with.
Depending on how long the antenna is now, you can try cutting a few mm more.