Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[2eyed]

To be more precise, this is actually what cameni wrote:
Colors - yes it's using compiled Blue Marble data, the problem is that the colors there do not represent raw color values, but were manipulated to make them fit into a single color map. It's non-linear, and we need to apply a reverse transformation - this too is subject to further evolution.

[Heyworth]

I think tweaking certain colours will have a very profound effect on the overall graphics...
A 'simple' coding tweak to make each 2D tree vary between 10 different colour shades as opposed to a blanket 'stock dark green', could vastly improve the appearance of the landscape. Likewise with blades if grass etc.

[HiFlyer]

Its not even a subtle wrongness; its very obvious.

As a reality check I took several flights through X-plane 10, Microsoft flight, FSX, DCS, Aerofly, War Thunder etc, and looked carefully at ground color and tree distribution.

Nothing else I looked at showed anything like the disparity between ground and vegetation colors shown in Outerra.

Microsoft flight (RV6 repaint by me)




DCS


Proland


Outerra

[ZeosPantera]

Well if the blue marble data says the grass is green. It is. Maybe a tool to tone it down. Try adjusting your Ground Reflectance in your atmospheric settings.

[PytonPago]

Hmm ... would be interesting to see DCS and OT footage side by side whyte an identical view on or from the Elbrus ... (proland and MFS too if they have the regions so the engine terrain-specializes could be nicely visible ...)

[M7]

If you dont like the color of the grass, you can always colorise the whole view  by clicking the 'color' in sun color & intensity in environment setting. It's a bit like wearing sunglasses

[M7]

I think tweaking certain colours will have a very profound effect on the overall graphics...
A 'simple' coding tweak to make each 2D tree vary between 10 different colour shades as opposed to a blanket 'stock dark green', could vastly improve the appearance of the landscape. Likewise with blades if grass etc.

For trees there's quite a bit of variations in color already.

[HiFlyer]

The problem with modifying those settings is that they can affect other things as well and not necessarily for the better.

I think they need an artist on the team, but for now I'm just gonna drop it.

Maybe it will be looked at, one day.

[2eyed]

I think tree shading is exceptionally well done, given that they are only 2d.
Biome colors are still off, but cameni is aware of needed refinements.
Changing the atmospheric settings is no solution, since it affects  sky and water accordingly.
Let's hope for the next update!

[Heyworth]

I think tweaking certain colours will have a very profound effect on the overall graphics...
A 'simple' coding tweak to make each 2D tree vary between 10 different colour shades as opposed to a blanket 'stock dark green', could vastly improve the appearance of the landscape. Likewise with blades if grass etc.

For trees there's quite a bit of variations in color already.

Not much and from height, even less.

[M7]

I think this is an example of  some of the odd texture colors combination of the biome. Saturated green next to reddish earth ground. In painting to harmonise things, you would want to dirty one color with a bit of the other. Like here  by adding a bit of green to the reddish brown and add a bit of orange brown to the green. 

Now how to apply that to the biome ...???i have no clue,  but i'm also confident that Angry and Cameni will find a solution.

[M7]

Its not even a subtle wrongness; its very obvious.

As a reality check I took several flights through X-plane 10, Microsoft flight, FSX, DCS, Aerofly, War Thunder etc, and looked carefully at ground color and tree distribution.

Nothing else I looked at showed anything like the disparity between ground and vegetation colors shown in Outerra.

Looks like i completly missed your point. Yeah trees could be lighter and ground a bit darker. I think when we get leafy trees, it should make for lighter patch of forest. I played once with the tree texture and i think i made it lighter, will check again.

Also i tried to play with ground texture few days ago but ended up with pixelate patches. Dont know how these are supposed to be saved.

[cameni]

The problem is how to get it consistently good over the whole Earth, complicated by the fact that the vegetation density dataset isn't exactly in sync with the color data. That's what HiFlyer asked about - how come the Proland, that apparently uses the same color data, doesn't have this problem? It's because they are using just the color data, and not mixing vegetation and bare land at all. If we used just the colors, it would be the same. But then we would not have the ground mixed by fractal patterns.
The color splitter is driven by the vegetation density value, which is often inconsistent with the colors. One option (apart from getting better veg data) is to extract the "greenness" from the color data instead, or to implement a heuristics that will correct the veg data if they are off for any reason.

I think they need an artist on the team, but for now I'm just gonna drop it.

Artists can't code procedural stuff
Don't despair, it will be taken care of. There are better datasets out there that we will incorporate, and we said the biomes are an initial version.

[Heyworth]

[Artists can't code procedural stuff ]

I think (for the sake of aesthetics and eye candy), that you shouldn't attempt to be too accurate...
You should allow yourselves a certain amount of 'artistic license'.

[cameni]

I meant that it's an output of a procedural equation applied on global data. Hiflyer suggested that we needed an artist, but the wild green colors in some locations are not because we are color blind and can't see the problem ourselves

We can tell that some colors are wrong and mostly what they should be like too. The problem is, we aren't flying around the world in OT and manually repainting it - the procedural rule that processes the available data needs to be able to handle it, while not making other parts worse at the same time. We *could* have an artist painstakingly fixing the dataset instead, but that would not be very productive, especially when the formats and datasets may and will change.
Other things, like the dark tree colors are a known issue of the existing code that will be replaced anyway, so we aren't hurrying to fix the old one at all costs.