Viper exhausts render issue

[AnotherCommander]

While testing Norby's fix for the Viper materials, I noticed a little render problem. I tested also with a Cobra MKI and the problem did not occur, so I guess it's something local. It is minor, but it's there.
See the two pics below. There is extremely small difference in distance between the two. However, the exhaust plumes appear much shorter on the first one. The jump to the length you see on the second pic occurs from one frame to the next. It looks like there are specific angles from which this happens, or at least, is more evident.


If you approach very slowly while this happens, you can see the plumes getting progressively shorter and look "cut" and at a certain distance they jump back to their normal length and appear as in the second pic.

[AnotherCommander]

Two more observations: 1.80 does not seem to have this issue and it looks like the distance where the change happens is exactly 0.180km (using the standard CobraIII as player ship). It is 0.200km for the Asp, which means that the problem could be related to the size / mesh of the model. Which reminds me, I should really test it also without Kevin's recent fix.
Edit: Nope, does not seem to be affected by recent fix.

[kanthoney]

I believe it entered master with the two pass rendering merge - the bug shows in 1a45adb but not in 7f90b5d.

[kanthoney]

Bug introduced with 03e4cce - I guess when the plume crosses the two pass boundary at 100m, the part on this side of the boundary is truncated until you get close enough for the whole plume to be inside the boundary, at which point it springs back into view.

[kanthoney]

I've tried three fixes, none of which have worked. Firstly I tried a quite involved fix of keeping track of the plume length in the update function. Then I tried a quick and dirty fix - changing

- (double)findCollisionRadius
{
return 0; // FIXME: something sensible. Where does plume length come from anyway?
}

to a constant value - and that doesn't work either, even if you change it to something OTT like 400. Neither did setting collision_radius to something non-zero in init.

[AnotherCommander]

The issue happens also with other entities, like flashers. I have the Docking Lights OXP installed and that puts a line of flashers from nav buoy to station door. Just launching and approaching these flashers demonstrates the bug. The flashers disappear much earlier than they should as they get approached.

Putting the INTERMEDIATE_CLEAR_DEPTH back to 50000 in MyOpenGLView.h fixes the problem, but I very much doubt that it is the solution we are looking for and am not sure what side effects any arbitrary value might have.

[cim--]

50000 - which is the sort of value I originally wanted for it - caused very weird effects as you approach the planet surface. They might have been fixed now - I'll check. (EDIT: no, still there - for some reason while it works for everything else, the planet doesn't join seamlessly when it crosses the pass boundary. If that's fixable 50000 would be a much better value to use, but I suspect that's post-1.82 for someone who knows OpenGL much better than I do)

The problem is that the ShipEntity::frustumRadius method returns _profileRadius, which doesn't include the exhaust plumes or flashers in it. The flashers because addSubentityToCollisionRadius isn't called for them (and shouldn't be - they only need the _profileRadius update) - the exhaust plumes not at all.

@kanthoney - if you try the complicated fix you had before, but apply it back to the _profileRadius of the main entity, it should work. (If you don't have that code around any more, let me know and I'll sort it out myself)

[kanthoney]

OK, I'll have a look at doing that. In the meantime I've uploaded the code as-is to branch issue_150.

[kanthoney]

I think that's the exhaust plumes fixed in the issue_150 branch.

[AnotherCommander]

Fix confirmed, also with flashers. Awesome job guys.

[kanthoney]

I've just tried INTERMEDIATE_CLEAR_DEPTH at 50,000 and the misjoin effect shows up on space stations as well, but not for very long at full speed. There's a partial fix in branch two_pass_planet_slice, but there's still an atmospheric effect I haven't tracked down.