I wish I had answers for you. I never installed AIG, and I only tried FSLTL 2x, and I didn’t like the performance penalty I was getting with it, so I removed it completely. I am on SU12 like you. Live Traffic + Simple Traffic. This solution has worked for me since SU9/10.
1 Like
Just to close up the apples-to-apples comparison - I launched out of KATL to JAX, here’s what it looked like by the time I fired up the jet and made my way to the Hold Short at 26R.
FlightAware Map open in separate window at same time:
My perspective at the hold short matches:
At climbout:
What’s behind me:
1 Like
Now see, you’re showing the same issue I am. You’re having aircraft only pile on the approach. You’re not getting any surrounding approaching traffic. Those two aircraft from the south should be there. You’re also zooming or cropping the area just immediately at the airport. What about a full zoomed out photo where you’re looking beyond the approach. You’re only looking at about 20-25NM. You should be getting traffic at 40NM+ just like my FSLTL image. The issue isn’t the airport. It’s the areas surrounding the airport which should have plenty of traffic as well.
So here’s the final deal on my part. I let the live traffic injection run for 10 minutes. This is using live traffic with Simple Traffic same as you and same settings.
Zoomed out:
This is what FlightAware shows:
If this is what I am seeing then that is what you’re also seeing. Why you keep posting cropped and zoomed photos just where the traffic is at and not the entire LNM I have no idea. I know departures take awhile to populate but arrivals should spawn almost immediately. To have 158 spawned AI and only 5 of those are approaching aircraft only on the approach path when per FlightAware there are 10 arrivals outside of the 20-25NM range that aren’t being displayed shows and proves there’s an issue. Seeing as how I can remember something similar from back with SU10 it’s likely this is how the traffic has been for quite some time. I doubt anyone until now has really realized that the injection stops about 25NM from around their aircraft and have just been thinking that the traffic there is the way it should be. For most large airports that have a final approach path of about 15-20 miles, this makes sense why there’s only traffic populating the approach and not on the arrival. Once departures are injected and takeoff, they’re free to fly on. I also imagine they disappear about the 25NM too. That I haven’t tested. But what applies to arrivals would also apply to departures in this instance. Going off of how FSLTL models their injector, it sounds like MSFS needs to look at what their injection radius is actually set to, because it’s not 80NM. FSLTL proves this and so does the testing and these results which I have been able to duplicate both at KDFW and KATL, both of which are default airports. Performance wise that makes sense, especially since it wants to populate the entire airport closest to the user. But that’s not what is advertised.
I wanted to show up close that it matched. I did zoom out in the other shot to show what was behind me on Departure. And you saw DFW. I’m not covering up anything if that’s what you’re implying. You wanted an apples-to-apples, I gave my best effort.
Why it works for me and not you, not sure.
Behind you would have been the airport, which of course would show traffic. I’m talking about arrival traffic outside of the immediate area of the approach pr traffic in general outside of the immediate area of the airport. The images and testing you’ve done are the same results I am getting. The difference is that you’re not posting your entire LNM and I am which shows the issue. The issue being there’s no traffic outside of 25NM even though supposedly the injection stops at 80NM. I can guarantee if you zoomed out from the airport you’d see the same thing I am. You’re not going to get any traffic outside of being right on the approach or within 25NM from the airport. Just like the other person who tested had.
Using live traffic, can you produce an image like this that shows stacked arriving traffic still on the STAR and not on the approach and is farther out than 20-25NM and also shows traffic that is probably 60-80NM away from the airport. You can see my full LNM so you can use where traffic is and the airports on the map for reference in your own test. FSLTL with 101 injected AI
Live traffic with 158 injected AI:
From my testing I can’t and from what you’re showing me, it sounds like you can’t either. The evidence doesn’t get clearer.
No I cannot. I gave you comparisons to what was happening in the area and compared it to FlightAware. I’ll let the posts stand on heir own. Good luck.
I’ve talked with people who have come across their IRL flight in sim with live traffic, flying along their same flight path, at the same flight level. Completely by accident too.
Live Traffic will usually have some delay because the information has to get from flightaware, through the API, into the game servers and projected into your sim. That’s just two different systems trying to talk to each other and show you the output there’s gonna be a delay especially if you’re connection is slow, spotty or otherwise unreliable.
I would not expect 100% pos for pos unless you pay for it. The immersion should be no less ruined.
That makes no sense. How’s the data getting from FlightAware into the sim in your scenario? Letter mail?
OK, we’ll see if that theory holds for there being a delay. Perhaps 10 minutes between the start of the injection and looking at the map was too long. I’ll time it for 30 minutes and report back. Since ATL is getting late I’ll use LAX. There should be no shortage of aircraft arriving from the east as it’s about rush hour and those flights have to get in so they can be used for the redeye rush. I’ll post my finding of Live traffic after 30 minutes and then immediately after with FSLTL injection.
Ran the injection at LAX for 30 minutes with live traffic. Here are the results:
189 injected AI with all the AI either being on the ground at LAX or stacked on the approach at 20-25NM again. And again only showing traffic for LAX.
This is FlightAware:
Notice I only have traffic showing to/from LAX so that’s not even with the surrounding airports. Which would make things a lot more busier.
Here is FSLTL with 70 injected AI after only letting injection run for about 1/2 minutes.
Here you can see a more accurate picture that represents FlightAware. Even the approaching aircraft from the north are present and aircraft from the east are further out. Something is clearly broken or has been changed. This test, yet again proves this issue beyond a doubt. Live traffic is clearly being restricted or limited to within a certain distance from the user. The 80NM so advertised is not present. Also, since live traffic insists on trying to fill every single parking spot, my FPS have tanked.
Summary:
30 minutes or 5 minutes, doesn’t matter. Live traffic is only around the immediate area of the user no more than 20-25 NM. Every parking position is filled so FPS tank. Live traffic injected 189 AI aircraft within 30 minutes while FSLTL injected 70 in less than 2 minutes, yet there’s virtually no air traffic in live traffic. Things can’t get more clear cut than this and has been reproduced 3 other times today at KDFW, KATL, and KLAX earlier. Anyone else, feel free to confirm my findings. Remember to screenshot your FULL Littlenavmap within about a 80NM radius from your position… Zooming in doesn’t count.
1 Like
WOW!..careful what you ask for I was always told. Really guys those were some amazing posts, thank you. I only wish I could appreciate them but it was information over kill for me. So let me please ask for a simple answer on those 2 questions.
-
I have AI planes set to generic and live planes set to high. So do I assume maybe some part of my live interaction is near real…with no consistency? I shouldn’t wait on the runway for my gf’s flight to leave and try following it…got it!
-
The airport information on flight arrivals and departures…how accurate is that?
Thanks again for all your responses.
-
Generic just more or less means using Asobo native aircraft models to represent the traffic. Asobo models have lower LOD (level of detail) so the FPS impact is less than other AI models such as TFS, FAIB, AIG, etc… By installing other AI models/liveries, if you have generic models for AI set to ON then MSFS will utilize those liveries from those outside planes on those lower quality Asobo planes which will help keep your FPS impact lower. What do you mean by live planes set to high? If you’re using live traffic from MSFS there’s no changing the amount of traffic that gets injected. If you’re talking about the variety setting between low, medium, high, ultra that just means that instead of repeating the same models it’ll try to randomize it more. So instead of 50 A330’s you might see 30 A330’s and 10 Airbus’s and 10 737’s. It’s more spread out. The data is real and the flights in the sim that are injected are real. So you could sit at an airport and see your GF’s flight. It wouldn’t be 1:1 but it would be generally there. You could follow it.
-
Accuracy of the traffic is again accurate in that if you were to see an aircraft in the sim and type it’s flight number you could see it’s a real world flight. The status of that flight in the sim from what the real world says per FlightAware might be different such as planned to depart, scheduled, taxiing, enroute, landing shortly, arrived.
The whole test, argument, discussion that was going on wasn’t about the accuracy of the flights (we know it’s real data) but more the depiction of those flights. Some of the busiest airspaces in the U.S. while having upwards of over 100-150 injected aircraft mostly injected into the airport seemed to have no more than 5-10 airborne planes that all seemed to be centralized on the approach to the airport and the bubble in which those aircraft were being displayed seemed to be limited to no more than around 25NM from the user. Which is definitely different than the supposed 80NM limit Because of this, AI depiction seems limited only to aircraft approaching/on final to an airport and not arriving or farther out like what you would see IRL. This can be seen by the differences between FSLTL where AI are injected to the max limit of 80NM and you can see strings of AI arriving into the approach area or AI traversing around the airspace of the airport. With MSFS you don’t see these strings of arrivals except for immediately on the approach path once on a general final to the airport and you also don’t see other AI traversing around the airspace which in places like SOCAL is extremely common. Other places similar would be KDEN, KSFO, KSTL where you often get a lot of traffic flying overhead enroute. So it seems you’ll only see traffic based on your stage of flight in a limited geographical region.
1 Like
Using Live traffic, aircraft within the red don’t disappear. Outside of the red boundary are getting injected but then disappear repeatedly. From my location and marking the distance to various waypoints around where aircraft don’t disappear and do is an average between 26-31/33 NM from my location.
Same shot but with FSLTL with new perimeter ring:
Top right aircraft is around 60 miles. Top left aircraft north of SFO is 80NM. Lower left aircraft are around 80NM. Towards SAC to the north there’s not any active traffic there yet so you can probably move that line up more. But yeah, generally around 80NM in all directions.
So from about 5 minutes of testing with the new WU13 here are the AI findings. Seeing that there was an AI_Traffic file being downloaded for a considerable amount of time I had hoped for some small basic changes. Planes still land halfway down the runway. Cruising traffic is stuck at 306 knots. AI reality bubble is still no more than 30 NM. So aircraft are constantly popping in and out until they get within that limit. Departing AI don’t call for clearance or taxi (supposedly fixed a few updates back) they mostly just pop either taxiing to the runway, at the hold short of the runway, or after having departed. Same for arrivals. While some planes successfully fly the approach and land, there are some that just pop taxiing to the gate and are radio silent.
In my copy of MSFS I do hear departing AI planes asking for departure clearance, taxi clearance, pushback and they usually (but not always) are aware of your presence when taxiing, so they stop or change taxiways when possible.
I do agree that arriving planes don’t seem to communicate at all with ground control and just silently taxi to their gates or disappear after landing.
I’ve seen this both with Asobo’s stock injector and FBW’s add-on too.
Can you give an example airport that you remember hearing this at? With FBW/FSLTL clearances has never been an issue, they usually start rolling after about 2-3 minutes from injection once the aircraft settle and trigger their departure state. Using live traffic I sat at LAX tuned in clearance for about 10-20 minutes and nothing but could see aircraft moving around me and heading to the runways. Today I sat at SMF and tuned clearance and ground and again I could see aircraft heading to the runway but none called for clearance or taxi relating back to the popping issue.
Is there an argument for AI generated aircraft, to at least be able to optionally display a label with their Callsign.
Because of limited screen size & pixel resolution, it is often far more difficult to see MSFS AI traffic, than one would see in RL… so some sort of compensation in the way of a user controlled Callsign label would seem appropriate.
The information is obviously available to do this, as depicted by LNM being able to identify AI Traffic data, including Callsign.
Let me paste a screenshot here of such ATC chatter. I’m ending a transatlantic flight but once I land I’ll start a new one at a busy airport.