Right now at EGCC with 40+ aircraft on ground, i am loosing 20fps from the model matching (AIG,FSLTL)
Probably not surprising, especially depending on your computerās specs. āEfficiencyā of the addon scenery and the AI models affects frame rates, and your computerās specs have a lot to do with it too. You can try reducing the radius that AI aircraft are drawn (say to 4 or 5 miles) and see if that helps, though 40 on the ground, especially with an older/less-capable machine, can definitely have an affect.
I wonder if we could make vPilot show a dummy (or no model) model for traffic far away (say 10/15 miles) and only good quality models for close traffic say <5nm. Notice this is just an example but in general the idea is show better models for close traffic around the airplane and basic models to no models for farther traffic (but still have the traffic so it is shown o TCAS).
No, there is no way to do that currently. Iām not sure it would help much anyway, since you should be using low quality models for model matching anyway. The difference in performance hit between a low quality āAIā model like FSLTL/AIG and no model at all should be fairly minimal on average-spec PCs, though I havenāt tried to collect any empirical data on this. I doubt it would be worth the work it would take. It would require that the vPilot UI has a way of marking models as ālow detailā versus ānormal detailā and then all the code necessary to delete and recreate these models as they cross the distance threshold.
I think there is definitely value in showing no model (otherwise the filters by distance or airplanes no. would have no purpose). In my current setup I use no aircraft shown >5nm and only show 10. This to make FPS good enough. I have a very high end PC, btw.
Maybe Iām not making my point clear. In my current scenario it would be great to see more than 10 airplanes on TCAS or keep seeing that aircraft that crossed the 10nm threshold, even though no model is being rendered in those cases. I guess vpilot is already doing some of this in order to only show those models/airplanes closer to the pilot location (and removing the rest no meeting the filters). The only question here is whether you can sort of ātrickā the simulator by having an āinvisibleā (no model) aircraft (so it is still captured by TCAS), the logic to detect whether the aircraft is to be āmasked-outā or not is there already.
The low quality model thing is just a byproduct not focus point here for me.
I understand what you are saying, but the thing is that the distance and count limits are mainly used for delaying when you see a big pile of aircraft appear as you are approaching a busy airport during an event. (And for making those aircraft disappear sooner when you depart a busy airport.) This is why I added those limits originally. People with slower computers were experiencing a sudden performance drop as they were on approach to a busy airport.
The vast majority of these aircraft are on the ground and donāt need to appear on TCAS anyway. I would further argue that when you are airborne, very few aircraft that are beyond 10 NM are going to be relevant to you on TCAS, especially when you are in the terminal area at much lower speeds. Aircraft beyond 10 NM become more relevant on TCAS when youāre at cruise speeds, but in that case, there arenāt likely to be more than a few total aircraft in range anyway.
Lastly, in order for an aircraft to appear on TCAS, there must be a model created. So I canāt use āno modelā ā¦ I would have to use an invisible model, or a really tiny model with just a few polygons. This would be a special aircraft that would have to be installed on the userās machine. There would have to be one for FSX/P3D, and another one for MSFS2020/2024. This approach would also require much of the same development effort that I listed earlier for the approach of swapping in low-quality models beyond 10 NM.
In short, I donāt see this as being worth the effort just to keep the distant aircraft appearing on TCAS.