Thanks for taking the time to reply. As of what commit did the change take place by chance? (save me some time routing through it). The new method is not coherent on what’s really going on. This other rig’s single track messages are not out-pacing the good decodes with this setting, guaranteed. It would be nearly impossible whilst running -w 3
I need to trust my guages and I don’t trust this one:
vs. the one I trust (also -w 3):
Thanks mate
You misunderstand the graph.
It’s displayed stacked.
It’s not displaying lines, the value of single message tracks is stacked on top of the other value.
It was displayed stacked before (without making sense because one type was contained within the other number which is no longer the case.)
And i’ve checked the averaging, it’s giving the same graph without averaging when looking at the longer timeframes.
(which means there is no scaling error in the averaging)
Switched to -w 1:
Same rig:
I guess what I’m saying is that I’d prefer them unstacked so the axis labels correctly reflect the number of bad messages. It’s sort of like stacking ambient air temp on top of on EGT readout otherwise…Ambient air temp would read 2000 degrees where it should be 30. Hope that makes more sense on what I’m pointing out. I know I’m slow and all. 
By the way, none of this is to detract from the awesome job you do on these. They are absolutely fantastic. Also sorry for derailing the tar1090 so badly.
It seems no one is able to find threads on this forum.
Using google works better than the internal function mostly.
Graphs for dump1090 -- my version with install script
Well i don’t care for the exact number of bad messages, i care for the exact number of actual tracks with more than one message.
The red should be kept low … that’s all there is to say about it in my opinion.
If you want accurate numbers for both, you need to make them proper line graphs.
And i didn’t want to perturb the style too much.
Also the stacking is somewhat logical as the total height gives you a total now (not really a relevant total …)
It’s a matter of taste really.
I’ve found another way to eliminate some more bad stuff.
Was still getting some bogus aircraft showing up from time to time, always a bit error in the 24 bit ICAO address.
Excluding messages from the whitelist scoring where a one bit error was fixed within the ICAO 24 bit address should help the whitelist even further.
(allow running with -w 2 or even -w 1 shouldn’t be too bad)
(yeah -w 1 is still a bad idea, just received a TIS-B message … unlikely in Europe)
Edit: still needs -w 3, but it should remove the last bogus aircraft.
I kind of agree that the graphics should not show a sum of two quantities by stacking.
I am used with graphics starting all from zero and then you can read on the axis the values - one for good messages with green and the other for bad with red.
But that’s more of a presentation preference, from my usual work related experience.
Here it can be whatever the boss wants to be. 
I’d probably just go with separate lines.
But i’m trying to keep the original style at least somewhat.
Random thoughts of additional capabilities for tar1090
A visual view of message received count visible per aircraft
Display a circle/shape on top or to the side of the aircraft (colored by scale e.g. white [1-1000], green [1001 - 5000] etc) based on the volume of messages received.
The sorting by message count in the table is one of my “views” on whats happening via my receiver.
Trails only within 1 range ring
Show all trails where all messages were only received within 1 range ring. A use of this for example would be to help isolate all trails from traffic originating and terminating local to the receiver or within a specific range ring of interest
My random thought for the web interface both on the table and map of plane size/color…
Could it change or highlight an emergency squawk of the 7700, 7500 ?
Already displays something, just like the default dump1090-fa interface.
my forum search found Squawk 7500
But I have never seen one, so I don’t have experiences/simulations to see the messages
I like to load Skyview on my TV so I can watch the planes while doing stuff around the house on the big screen.
dump1090-fa displays on my TV web browser like normal,
tar1090 displays, but the planes don’t move, and the list doesn’t update.
I’m guessing it is cause the built in “TV Browser” doesn’t support some aspect of the backend.
Its too bad, as I prefer tar1090.
I have a HiSense UHD 4k tv.
Really can’t debug that properly without having such a TV …
So the initial load of the map works?
Are there never any planes displayed or do they just not move?
initial load works slowly, then everything shows, planes, and planes in list.
Unless you know how to open the javascript/browser console and can give me some error i can work with, can’t help you, sorry.
(yeah … seems you need to do some fancy Android remote debugging, not sure that’s feasible for you to do)
I imagine it doesn’t support some aspect of javascript or the like.
I found what user agent it advertises as hidden in the settings somewhere, but now I can’t find it.
Pretty sure it’s this:
2763 const search = new URLSearchParams(window.location.search);
I’ve put it into a try catch block.
Update and try again.
That was fast. 
no change though.
Of course I have no idea how to flush the browser cache on a stupid TV app.
This should apply: Clear data and cache on your Android TV or Google TV when apps aren't working properly | Sony USA
Anyway, install the new version, i’ve made an popup alert box that should display any errors at the end of loading.
I have a Sony TV as well which is Android 8.0 driven. The first thing i did was installing Firefox.
Tar1090 works in Fullscreen view without issues.
Have you tried casting the video from your PC to TV?