Bug description:
Our older aircraft rely on simple exterior lights (no emissive mesh) using existing Asobo effects: LIGHT_ASOBO_NavigationGreen, LIGHT_ASOBO_NavigationRed, LIGHT_ASOBO_StrobeSimple.
Those effects are not displayed correctly in FS2024, their intensity and size are too low, we can barely see them in the middle of the night.
I understand that FS2024 brings a new logic for effects brightness, but FS2020 effects should be considered as legacy and displayed the same way as in FS2020.
Repro steps:
Load the AzurPoly BD-5J (FS2020 package) inside FS2024 Community folder.
Launch a free flight.
Turn navigation lights and strobes on (switches on main panel).
SU3 beta introduces a big change for this.
Light flares have been reworked. They will not render as they did in MSFS 2020, as their lack of visibility was a recurrent complaint.
Please have a look and let us know what you think.
On the H145 specifically the problem is the sprite/flare visibility from afar. If you move the camera away from the aircraft you might notice how in MSFS 2020 the lights will remain visible while in MSFS 2024 / SU3 they will fade out fairly quickly with distance, which is very unrealistic.
I think there is something fundamentally different with how the legacy .fx parameters are interpreted by the sim. I’m mainly suspicious of the stuff in the sections [ParticleAttributes.0] and [LightAttributes.0] - like MinProjSize, Size and Range parameters… they seem to have radically different meaning compared to MSFS 2020…
Disclaimer: I’m not involved with the development of the HPG H145 - just noticing big differences between 2020 and 2024 by comparing the rendering of the identical package and also based on tweaking parameters via the “User Lights” tool as well as modifying the *.fx files for nav, beacon and strobe lights as per their references in systems.cfg.
There is something seriously broken with legacy .fx lights as interpreted by 2024. I am still working out the kinks and differences, but it might also have something to do with the circuit assigned to them even, which makes no sense to me (why would a WING circuit light be interpreted differently than a GLARESHIELD circuit light in terms of its projection, intensity, angles, etc, since those are all defined in the .fx file - and if there is some reason for it, at least document it thoroughly otherwise we spend WAY too many hours trying to figure out what is what).
I’ll post images of the differences we notice later on today, but if someone has more details or info on that, it’d be appreciated.
Hello,
I can tell it is way better because we can see the effects now.
However I confirm that the effect size and intensity is not the same as in FS 2020.
I’ve had the case of the H145 reviewed by our engine team and this is the expected result on 2024.
Back in 2020, light sprites were essentially a 2x2 px opaque quads, which means they were visible at extremely high distance, regardless of the actual light configuration.
In 2024, with the new sprite system, the light will look closer to what it should, based on its configuration.
In your case, for example, the value in [LightAttributes.0] Intensity in your LIGHT_H145_Strobe.fx file is too low for a strobe light.
If you increase it, you should preserve the look of the light at long distance in MSFS 2020, while improving its visibility in MSFS 2024.
Can you try this and let us know if you are happy with the result?
Increasing the intensity helps a bit in 2024 but IMO the effect still behaves weird overall compared to 2020. To achieve a realistic far visibility it seems I need to crank up the intensity to such high values that the passive illumination effect of nearby surfaces will be greatly exaggerated and it will just not look right from up close anymore. Frankly, I think the new lighting model needs more tweaking to achieve a realistic/balanced visibility across a wide range of distances.
But what I find even worse is the fact that this forces content creators to publish and maintain two versions of their aircraft if they only want to make sure the exterior lights behave similarly in 2020 and 2024.
That’s something we are considering but it will require more feedback in a consistent direction to motivate this.
We need dust to settle down a bit and see where the feedback is pointing to.
If you see any related discussions on forums or other devs platforms that are consistent with your feedback, feel free to share it.
That’s something we avoid as much as possible, and we are able to keep a legacy code path for other features like systems behaviors, etc… but when it comes to rendering, this is not possible.
And what can someone tell me about the display of the flares after SU3 release. What can this be?
If someone can point out my issue, the light effect zip is attached to my reply.
Sorry to bring this topic back up once again, but having an issue with my light textures still not displaying the same as they should be from FS2020. The intensity, etc. has all been fixed, but the textures are still not correct.
Here is an image of my aircraft in FS2020 with custom light textures:
