About 4 weeks ago, my collectd plot of ADS-B Receiver Tracks Seen showed a marked increase in the number of “tracks with single message”. Insofar as I could discern under the recent decline in aviation traffic, there were no other noticeable changes in my system (CPU use, range, etc.). I made no hardware adjustments, including no change to the gain of my LNA. I have noted, however, that the change occurred approximately around the time that piaware 3.8.0 got upgraded to 3.8.1. As far as I could tell from the changelog, there was nothing apparently related to what I’m seeing. See the attached plot, for reference. Has anyone else observed such a change, and/or would anyone have any suggestions as to what I could look into to troubleshoot?
Thanks for your reply. I had wondered that too, but thought I’d ask here before starting to mess around with my settings, which I haven’t changed since last summer. I do have the --fix option specified in both the RECEIVER_OPTIONS and DECODER_OPTIONS option lists in /etc/default/dump1090-fa.
Thanks again @foxhunter. I checked and, yes, I somehow had those two separate --fix options so they got interpreted as --fix --fix, which I understand now does local on-board two-bit error correction. All “fixed” now! This matter is now resolved.
If given once it will do single bit error correction where possible. This is usually an OK tradeoff; it will increase your message rate (and possibly range) but will will slightly increase the number of garbage messages received. Usually the existing message filters are enough to eliminate the garbage.
If given twice (as accidentally happened here) then it’ll also do two-bit error correction. This is a more marginal tradeoff; you’ll get more messages but also significantly more garbage. The rate of garbage messages can be high enough to make it through the message filters regularly. It’s mostly useful for the airspy case where airspy_adsb is doing two-bit correction anyway and you want those results to appear on your local display; or if you’re willing to deal with bogus data turning up regularly.
Messages with 2-bit errors corrected are not fed to FlightAware because the rate of bad messages that it generates is too high to be useful to us.
I’ve removed “–fix” for testing the last few hours. Looks like also the “green” message volume has been decreased during that time. Can you confirm this or is this just by chance?
First section was with a higher gain, middle part with the gain i usually use and the last part is without the fix enabled
