[Csantucci]

Here is what I see with a legacy GP38-2 with headlights on with the actual testing release

and here is what I see with the actual unstable release

The difference is due to the gitf2 pull request.
IMHO the unstable release does not provide a bright enough result.
Note also the difference in FPS.
Do others experience the same?

[pschlik]

The light cone effect used by the gltf has the potential to look better than the previous effect but I agree it definitely has problems.

• As you see here, the new light effect gets dimmer the more focused the beam is, which is honestly the opposite of what you'd probably want
• The transparency component of LightColour doesn't seem to change anything about the light brightness in gltf, it would be useful to restore the ability to control brightness with that
• The position the light emits from will 'lag' the position of the locomotive its attached to; moving forward quickly causes the light to move backwards into the locomotive body even if the light has been defined to be in front of the locomotive body
• It also just looks particularly bad on some old MSTS stock where developers placed the light emitter in the middle of the locomotive. The previous lighting model didn't look right in this case but this new one makes it even more obvious. Maybe we need a 'correct questionable lighting parameters' option

[Genma Saotome]

Part of the problem here is most MSTS values for the cone of light are wrong for MSTS and the redefinition of that parameter for OR has made the problem much worse. That will goof up a lot of feature testing.

On real locomotives the cone of light is about 5 degrees. I don't recall right now if that is 5 degrees on both sides of the center line or 5 degrees in total. I'm pretty sure it is 5d in total, just not positive.

Most MSTS cones are 1) grossly too wide and 2) grossly too short. I've seen cones of light illuminating the roadbed 3m in front of the train, about 60d either side of the centerline of light, and fading out 250m down the track. The light needs to illuminate what's on the track well beyond the stopping distance of the train. In the US that's often 1.5km.

There are the OR values I use:
Light (
comment( Cone of Bright Light )
Type ( 1 )
Conditions (
Headlight ( 3 )
Unit ( 2 )
Penalty ( 1 )
TimeOfDay ( 1 )
)
Cycle ( 0 )
FadeIn ( 0.5 )
FadeOut ( 0.5 )
States ( 1
State (
Duration ( 0.0 )
LightColour ( eff9ec95 )
Position ( 0.0 3.3 7.22 )
Transition ( 0 )
Radius ( 1000.0 )
Angle ( 5.0 )
)
)
)

[Jonatan]

I've noticed over time that depending on what unstable I used (I update very rarely) the lighting system could vary between the new or old. I write this down to there being misses in keeping track what version of code is worked on or implemented in the build.

On my end, it's made a harrowing mess out of my carefully configured lighting settings.

[ATSF3751]

I have the same issue! I recently have had to change a ton of my locomotives to make it look even somewhat decent. I was actually messing around with it yesterday a little bit and I came up with these results. This is the recent unstable release. U2014.3.29-1638 I tested out quite a few locomotives that I have and it seemed to differ locomotive to locomotive. Some looked good while others looked horrible. I feel the light code still needs some work but its getting there!

Brandon

Here are some of my results.

Train Simulations locomotives

Bright







Dim






Bright





Steam locomotives

[Genma Saotome]

BRandon, WRT the cone of light, change the value for Radius(?) to Radius(1000) and Angle(?) to Angle (5). What does that look like?

[ATSF3751]

So I did a little testing and I figured out that 1.0-5.0 does absolutely nothing. Here is what I had to change the cone lighting effects too to even make it look somewhat decent.

The radius is at 1000 and the Angle is set to 45.0

[ATSF3751]

Here are some screen shots in increments of 10 starting at 20.0 for the angle and the radius set at 1000 for all of them!

[Genma Saotome]

Set the angle at 4 or 5. The crew does not need to see the track right in front of the locomotive nor off to the sides. What does it look like from the cab view?

Radius() the the radius, in meters for a sphere that represents the limit the light will be shown. On road diesel locomotives it should be at least 1km.
Angle() is the breadth of the cone along a centerline with the peak at the locomotive. It's been 4-5 degrees for about 100 years.
Position() is the XYZ values on the model where the light emanates from.


You want to see something like this (from above), shot in version 905; current is 925:

[Jonatan]

I tend to put the origin point for the light at the height of the physical headmap of the locomotive, to get a realistic light casting across the terrain.

The new system seems to suffer some issue where the origin points drifts back with speed so I had to set it a few meters ahead to prevent it from illuminating the model.