 |
|
|
|
|
|
| |
| |
|
|
|
|
| |
| |
|
|
So, I played around with this a bit more.
Got some tile face normals, used the array indices to generate tile labels.
Reworked the cube-to-sphere transform.
Need to write an animation to step through a cubic map and generate image maps
for all the tiles.
I'm not exactly sure how to generate a tile, map it, then translate/squish it to
fit in its proper space.
This is at a "face resolution" of 12, which give an 11x11 tiled side of the base
cube (so the labels are legible)
The idea here is that at full resolution, each tile will have an associated
image map or include file for rendering whatever goes there.
This only took 2 seconds to render.
Post a reply to this message
Attachments:
Download 'proceduralmeshplanet.png' (111 KB)
Preview of image 'proceduralmeshplanet.png'
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Got the 2009 Chris Froeschlin starfield added, Lat and Long for reference, and
an overlay I grabbed from
https://en.wikipedia.org/wiki/Geographic_coordinate_system#/media/File:Earthmap720x360_grid.jpg
Outlined the base cube faces to better separate the different sets of grid
labels.
So, that should be a good start. :)
Post a reply to this message
Attachments:
Download 'proceduralmeshplanet.png' (1102 KB)
Preview of image 'proceduralmeshplanet.png'
|
|
| |
| |
|
|
|
|
| |
| |
|
|
You could also try generating polygons with more points.
http://experilous.com/1/blog/post/procedural-planet-generation
Mike
On 5/25/2020 7:42 PM, Bald Eagle wrote:
> So, I played around with this a bit more.
>
> Got some tile face normals, used the array indices to generate tile labels.
> Reworked the cube-to-sphere transform.
>
> Need to write an animation to step through a cubic map and generate image maps
> for all the tiles.
>
>
> I'm not exactly sure how to generate a tile, map it, then translate/squish it to
> fit in its proper space.
>
>
> This is at a "face resolution" of 12, which give an 11x11 tiled side of the base
> cube (so the labels are legible)
>
> The idea here is that at full resolution, each tile will have an associated
> image map or include file for rendering whatever goes there.
>
> This only took 2 seconds to render.
>
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
So, I'm trying to take the face subdivisions from a QuadSphere and uv-unwrap
them so I can overlay it on an image map.
I tried generating cartesian corner vertices, converting to spherical
coordinates, and then just plotting those theta and phi angles on the x and y
with faux-sphere-sweeps for the horizontal and vertical coordinates, and got
some odd results.
Apparently Blender has some sort of add-on / plugin to do this with a mercator,
but I couldn't heads or tails of the code --- yet.
https://blender.stackexchange.com/questions/3315/how-to-get-perfect-uv-sphere-mercator-projection
// Draw Horizontal Lines
union {
#for (Y, 0, faceResolution)
#for (X, 0, faceResolution, Smoothness)
#local Percent = <X/(faceResolution), Y/(faceResolution)>;
#local pointOnUnitCube = upVector + (Percent.x-0.5)*2*AxisA +
(Percent.y-0.5)*2*AxisB;
#local pointOnUnitSphere = Remap (pointOnUnitCube);
#local Vertex = pointOnUnitSphere*PlanetRadius;
#local LL = C2S (Vertex);
#local CurrentNS = LL.x/90;
#local CurrentEW = LL.y/180;
#local Current = <CurrentEW*S, CurrentNS*S-HalfY, 0>;
#debug concat ( "Current = ", vstr(3, Current, ", ", 3, 3), " \n")
sphere {Current, Line}
sphere {Vertex, Line}
#if (X > 0)
cylinder {Last, Current Line}
#end
#local Last = Current;
#end
#end
pigment {rgb 0}
}
Post a reply to this message
Attachments:
Download 'cubesphere_uvgridmaker.png' (45 KB)
Preview of image 'cubesphere_uvgridmaker.png'
|
|
| |
| |
|
|
|
|
| |
| |
|
|
"Bald Eagle" <cre### [at] netscape net> wrote:
> So, I'm trying to take the face subdivisions from a QuadSphere and uv-unwrap
> them so I can overlay it on an image map.
>
> I tried generating cartesian corner vertices, converting to spherical
> coordinates, and then just plotting those theta and phi angles on the x and y
> with faux-sphere-sweeps for the horizontal and vertical coordinates, and got
> some odd results.
reworked the code a bit, and that didn't help any. Then I switched over to f_th
and f_ph instead of whatever I was using to convert from Cartesian to
latitude/longitude.
That fixed things nicely - almost.
A have some straight horizontal lines now that create "extra" latitudinal
lines/circles. So I'll have to puzzle out the origin of those.
Post a reply to this message
Attachments:
Download 'cubesphere_uvgridmaker.png' (186 KB)
Preview of image 'cubesphere_uvgridmaker.png'
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Update:
My man jr and I have discussed the many aspects of this at great length, and
I've formulated a suite of files to generate the framework of the planet and
create tiles as separate include files.
Some more work needs to be done in order to create some DEM-style terrain, but
at the present time, all of the "parts" successfully work to render the planet.
The idea here is that although a huge number of tiles may be needed to render
all of Earth, the likelyhood that anyone will need to accurately and
realistically render all of the areas of Earth covered in water, ice, sand, or
forest is --- low.
So the "engine" looks for existing tile definitions, and when nothing is there
(zero data storage), a default texture is rendered.
The include files are "autonomous" in the sense that they are included to
provide content for the region they describe, and the engine need not worry
about it.
jr has worked out and written a lot of QuadTree code, so once some more
development has occurred, we'll start experimenting with tile subdivision, and
distance-based Level Of Detail (LOD).
So, a .pov file takes the number of subdivisions of each face of the QuadSphere
and uses that metric to create NxNx6 include files. They are currently randomly
textured based on a randomly generated biome type that is defined in it include
file. The plan for later is to scan a global image map and make a rough
assignment of tile biome type based on that.
That will essentially take care of 89% of the planet, and leave 11% of the
planet where something of interest might actually be.
Individual tiles can be created/copied, subdivided, and edited to add things
like oil platforms in the middle of the ocean, TdG on an arctic ship, or
anything else, wherever one might like.
The attached proof-of-concept image is 8*8*6 = 384 individual include files,
automatically generated, and without any user-added content.
Post a reply to this message
Attachments:
Download 'myplanet.png' (355 KB)
Preview of image 'myplanet.png'
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Op 11/07/2020 om 23:51 schreef Bald Eagle:
> Update:
>
[snip]
As you are aware, this is a view of the planet Quilt in the Fomalhaut
system ;-)
I am following this work with interest. You are making good progress indeed.
--
Thomas
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Thomas de Groot <tho### [at] degrootorg> wrote:
> As you are aware, this is a view of the planet Quilt in the Fomalhaut
> system ;-)
You certainly do keep abreast of topics astronomical, don't you..
IIRC there was a sci fi novel about a "planet" made of stolen shards of other
planets all cobbled together... can't recall or find it...
> I am following this work with interest. You are making good progress indeed.
Thanks - the key idea is that it's going to be very much like a particle system,
where the engine doesn't control the particles - the particles direct
themselves.
So one can add, delete, or modify an include file (within the given
constraints(?)) and you can still render an entire planet - straight "out of the
box" or even if the only thing you've modeled is the Eiffel tower - or a sphere.
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Op 12/07/2020 om 13:12 schreef Bald Eagle:
> Thomas de Groot <tho### [at] degrootorg> wrote:
>
>> As you are aware, this is a view of the planet Quilt in the Fomalhaut
>> system ;-)
>
> You certainly do keep abreast of topics astronomical, don't you..
>
> IIRC there was a sci fi novel about a "planet" made of stolen shards of other
> planets all cobbled together... can't recall or find it...
>
Hmm... I don't recall that one for the moment.
>> I am following this work with interest. You are making good progress indeed.
>
> Thanks - the key idea is that it's going to be very much like a particle system,
> where the engine doesn't control the particles - the particles direct
> themselves.
> So one can add, delete, or modify an include file (within the given
> constraints(?)) and you can still render an entire planet - straight "out of the
> box" or even if the only thing you've modeled is the Eiffel tower - or a sphere.
>
Yeah! That opens up really interesting perspectives. My (long-dormant)
Gancaloon world would perfectly fit the picture.
--
Thomas
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
A bit more progress:
Got the cartesian to spherical to uv conversions worked out, used the mapping to
look up altitude at any given point, and have bezier arrays for each of the
tiles. Using the mapping to sample the center rgb value of a tile from a color
world map gave me a rough value which hopefully may be used to automate terrain
designations and texture definitions.
I overrode the random format assigned to the tiles, and made them all bezier
patches - about 1500. Elevation is greatly exaggerated (x300). Can render as a
wireframe of bezier splines, and still have to work out how to generate the
appropriate heightfields from the world map for the individual tiles.
Post a reply to this message
Attachments:
Download 'myplanet.png' (137 KB)
Preview of image 'myplanet.png'
|
|
| |
| |
|
|
|
|
| |
|
|
