The axises shown are as they would look in 2D.
The angles written are showing merely how those axises would look on a 2D plane -paper-.
With the given scale factors it is not an orthogonal axonometry per se.
There are some similar dimetries with scale factors making it for an orthogonal view,
but this uses qx=0,5, xy=qz=1.
Like if it was all done on paper, those values could be constructed much easier.
While it IS close to an orthogonal view in look.
It is a bit illogical to give scale factors for the axises in orthogonal axonometry by the way.
Here is an example of an orthogonal projected dice:
http://www.math.bme.hu/~zlangi/descriptivegeometry/orthogonal_axonometry.pdfIf that was the goal, I would have asked for rotation angle values to write into the render 3D polyhedrons extension.
Which would be a too complicated way of drawing things in inkscape.
I know I can set a camera to be ortographic in blender, but these are different axonometric projections.
At that blenderartist topic I described an ugly solution for doing it.
For example, creating a military perspective would be
of setting the camera like in (0;-15;15), with a rotation value of (45°;0°;0°),
and scaling the model in x dimension with 0,707 factor.
But I'm not sure if the shadows would turn out right in cycles.
Also scaling the model and setting the camera for this dimetric look is a bit problematic that way.
At the same post I quoted the wikipedia's projection example's povray source file
Code: Select all
#case (2) // dimetric #local cam_area = 2;
#local cam_loca = -z * cam_dist;
#local cam_dirc = +z;
#local cam_rgvc = +x * cam_area;
#local cam_upvc = +y * cam_area/cam_aspc;
#local cam_tran = transform
{
rotate +x * asind(tand(15))
rotate +y * 045
translate +y * cam_move }I am not sure though if it doesn't describe an orthogonal projection.
With the "oblique" -military perspective- projected models the different scale factors are more evident.
Which are not displayed right here in the forum,the image is way too big for that:
http://upload.wikimedia.org/wikipedia/commons/archive/4/41/20091120003734%21Graphical_projection_comparison.pngSticking here with the original inkscape purpose: how to construct drawings in this dimetric axonometry.
By drawing only in 2D.
Let's say I have the side and top views drawn before, that I would like to transform
as they were on the actual xz and xy planes.
That means I have to "map" those somehow to these planes.
So figured out these transformation values, to be used on each of the views.
Actually this won't affect the "up" to be tilted.
I cannot recall which description of transformation matrix I saw,
but all those values can be checked on the xy, xz, yz plane's paralelograms.
They all have a value of x and y sizes of the path's right segments, with right senses.
Not mentioning again that those senses need to be changed if typed through the xml editor, as if not it produces a mirrored result.
And the fact that unfortunately even with those values the result is far from being accurate.
Will get together what transformations worked instead of the transformation matrix.
It is a bit more complicated as with the scale and skew values length dimensions were added.
The reason why an extension would be a boost for mapping many different sized objects without grouping them together, one by one, accurately.
This is a read-only archive of the inkscapeforum.com site. You can search for info here or post new questions and comments at 
