building a 3d world globe

[jumpjack]

R
Below are examples of this process on a basic Earth map to get polar caps.

Cool, looks like it's exactly what I need!
But I can't install Wilbur on my office PC, as the installer requires administrator privileges?!?
Isn't any ZIP/portable version available?

edit:
...and I can't download Fractal Terrains as it appears as a "Game" site! :-(

[waldronate]

Try installing directly from the msi file rather than the exe in the WIlbur setup - it might work better.

[jumpjack]

Try installing directly from the msi file rather than the exe in the WIlbur setup - it might work better.

I found two better solutions:
- I waited till I went home and I installed Wilbur on my home PC
- I installed a virtual PC on office PC

But now I have another difficulty: how do I know how large the saved image must be (proposed size is always 1024x1024) to "match" the sinusoidal projection which I used to obtain the petals?
Let's suppose the starting image is 640x320 size.

[waldronate]

Pick any size that you want. It's the inches / degree at constant dpi when printing that's important, not the image size.

For an arbitrary ball, the length of a pole-to-equator sinusoidal gore will be (ball diameter * pi / 4). A common 10-inch playball give a 7.85 inch gore (10 * 3.1415927 / 4).

With a 10-inch ball, you need a sinusoidal gore that's 7.85 inches long pole to equator. If you are using an image that's 320 pixels high, then the pole-to-equator distance is half of that or 160 pixels. 160 pixels / 7.85 inches = 20.4 pixels/inch (very low resolution). To get an endcap that's 15 degrees radius and printed at the same ppi then you'll need an image that's 2/6 of the 160 pixels (15 degrees / 90 degrees = 1/6 radius * 2 = diameter) or 53.33 pixels) or 53 pixels.

For a more reasonable 2048 height image then you're looking at 130 ppi and 341 pixel endcaps.

[jumpjack]

Pick any size that you want. It's the inches / degree at constant dpi when printing that's important, not the image size.

For an arbitrary ball, the length of a pole-to-equator sinusoidal gore will be (ball diameter * pi / 4)[...].

Thanks, I think now I got the point.
Given ball diameter D, I need a W x H map, where:
W=2*H
H=D*pi/2 (H/2 = C/4 = 2*pi*r/4 = D*pi/4)

Given this map, I need, for a Lambert projection 30° wide, an image WL wide, where:
WL=D*pi/12 (WL=C*30/360 = C/12 = 2*pi*r/12 = D*pi/12)

I hope it's correct.

[RobA]

Here is an inspirational video for you:

http://videos.howstuffworks.com/scie...lobe-video.htm

-Rob A>

[jumpjack]

Here is an inspirational video for you:

http://videos.howstuffworks.com/scie...lobe-video.htm

-Rob A>

thanks, very useful.
I need a rod.

[winski]

If somebody is still looking for solution, I coded sinusoidal gores projection in javascript, and it is available here http://www.winski.net/?page_id=12 under BSD license (you can do whatever you want with the program and its source). It works in a web browser so you don't have to download anything.

[jumpjack]

If somebody is still looking for solution, I coded sinusoidal gores projection in javascript, and it is available here http://www.winski.net/?page_id=12 under BSD license (you can do whatever you want with the program and its source). It works in a web browser so you don't have to download anything.

Looks promising from tablet, but I can't access it from Office PC as the site is "uncategorized"... Maybe you can add some keywords like "planets", "globes" or "cartography" in page header?

[johnvanvliet]

if all you need is gores
netpbm already has this built into it and it has been for MANY years
this is a 6 year old thread after all

netpbm is a collection of image tools that is almost ALWAYS installed on every linux Operation system and most windows as part of the printing code

ppmglobe 6 map.ppm

from the manual page
" man ppmglobe "

Ppmglobe User Manual(0)



NAME
ppmglobe - generate strips to glue onto a sphere


SYNOPSIS
ppmglobe [-background=colorname] [-closeok] stripcount [filename]

Minimum unique abbreviation of option is acceptable. You may use double hyphens instead of single
hyphen to denote options. You may use white space in place of the equals sign to separate an option
name from its value.



DESCRIPTION
This program is part of Netpbm(1)

ppmglobe does the inverse of a cylindrical projection of a sphere. Starting with a cylindrical projec-
tion, it produces an image you can cut up and glue onto a sphere to obtain the spherical image of which
it is the cylindrical projection.

What is a cylindrical projection? Imagine a map of the Earth on flat paper. There are lots of differ-
ent ways cartographers show the three dimensional information in such a two dimensional map. The cylin-
drical projection is one. You could make a cylindrical projection by tracing as folows: wrap a rectan-
gular sheet of paper around the globe, touching the globe at the Equator. For each point of color on
the globe, run a horizontal line from the axis of the globe through that point and out to the paper.
Mark the same color on the paper there. Lay the paper out flat and you have a cylindrical projection.

Here's where ppmglobe comes in: Pass the image on that paper through ppmglobe and what comes out the
other side looks something like this:

Example of map of the earth run through ppmglobe

You could cut out the strips and glue it onto a sphere and you'd have a copy of the original globe.

Note that cylindrical projections are not what you normally see as maps of the Earth. You're more
likely to see a Mercator projection. In the Mercator projection, the Earth gets stretched North-South
as well as East-West as you move away from the Equator. It was invented for use in navigation, because
you can draw straight compass courses on it, but is used today because it is pretty.

You can find maps of planets at maps.jpl.nasa.gov ⟨http://maps.jpl.nasa.gov⟩ .


PARAMETERS
stripcount is the number of strips ppmglobe is to generate in the output. More strips makes it easier
to fit onto a sphere (less stretching, tearing, and crumpling of paper), but makes you do more cutting
out of the strips.

The strips are all the same width. If the number of columns of pixels in the image doesn't evenly
divide by the number of strips, ppmglobe truncates the image on the right to create nothing but whole
strips. In the pathological case that there are fewer columns of pixels than the number of strips you
asked for, ppmglobe fails.

Before Netpbm 10.32 (February 2006), instead of truncating the image on the right, ppmglobe produces a
fractional strip on the right.

filename is the name of the input file. If you don't specify this, ppmglobe reads the image from Stan-
dard Input.



OPTIONS
-background=colorname
This specifies the color that goes between the strips.

Specify the color (color) as described for the argument of the ppm_parsecolor() library routine
⟨libppm.html#colorname⟩ .

The default is black.

This option was new in Netpbm 10.31 (December 2005). Before that, the background is always
black.


-closeok
This means it is OK if the background isn't exactly the color you specify. Sometimes, it is
impossible to represent a named color exactly because of the precision (i.e. maxval) of the
image's color space. If you specify -closeok and ppmglobe can't represent the color you name
exactly, it will use instead the closest color to it that is possible. If you don't specify
closeok, ppmglobe fails in that situation.

This option was new in Netpbm 10.31 (December 2005).