Rendering for Book Illustration

I need to create book illustrations from my Rhino files. In some cases the files contain “solids” and other cases they are just surfaces. Here is example of such a file containing surfaces.

I don’t care about fancy lighting or textures. I need to get clear outlines. I am curious what solution might work for me here.

It’s likely just me, but I don’t understand what problem you might be having. How to redraw a ship hull?

Before going forward, I might check to see that the “surfaces” are really 2d surfaces, and not 3D solids/polysurfaces that aren’t being rendered well because of their repetitive thinness and rendering distance–resulting in surfaces that would have to be rendered in less than a single pixel.

It might be good to check with what was exported, and which settings were used for export/import, before proceeding because…

If some things are only drawn as simple planes, they could each be extruded, perhaps with all being subtracted with the out hull plate, though if the resultant edges are quite close to the hull skin, you might have tolerance issues, because the surfaces may be almost co-planar.

That image may be of the same material, but it appears to be rendered, at least in phong shading, with a directional light source of some kind. As the longitudinal stringers are lighter than the bulkheads.

If that were rendered, it would seem to only need a single gray material, and just a one or two directional or sun-like lighting sources. Rhino’s Default lighting used on some of it’s display modes actually has a few light sources.

To keep nice detail with fine/delicate geometry, as opposed to rendering things with an even aspect ratio, you are likely going to need to render it in high resolution for the final render, with a fair bit of oversampling, that will allow the finer parts to be seen better.

What about simple _make2d command?

1 Like

Make2D does not give me very good results. I get many disappearing lines. I have been hoping to see it get better.

Plus, in some cases, i’d like to get perspective views.

Make2D has gotten better, but for ‘publication-quality’ illustration you’re still going to need some cleanup.

There are a ton of options available to modify the OpenGL settings to get a linework-like image. You can start with ‘technical’ or ‘artistic’ etc and tweak a lot of things, I produced dozens of product assembly manuals using such images that looked way the hell better than whatever they had before. Of course that can’t produce vectors, so you have to be aware of resolution.

If you have V-Ray for example, you can use Toon shaders.

I’d also recommend you to use Make2D. What I usually do for quick visualisations is render an ambient occlusion (i.e. artic mode in Rhino), and overlay the lines from the Make2D. This can be done in an image processing app (i.e. Photoshop, Affinity Designer, Gimp [free], etc.), or even better a vector program (i.e. Illustrator, Inkscape [free], etc.).
The vector based workflow has the benefit that you can rescale the lines to any resolution without losing definition or getting a pixelated image. Only the ambient occlusion needs to get re-rendered from time to time, which is pretty quick, but rescaling and adjusting the line weights, styles and line colours usually isn’t.
In order to make it easier to define line styles, simply keep your Make2D lines on separate layers. This way you can select whole objects by layer, which oftentimes comes in handy.

If at all possible and depending on the scale of the publication, I’d get rid of some of the geometries beforehand to clear the drawing up. Usually model fidelity is less important than showing a clear diagram. You should find a better balance here.
You could highlight more detailed parts in separate drawings.

A stronger object silhouette and lighter inside lines also oftentimes does wonders for a drawing.

just in case you really did not stumble over it, did you try the pen display mode (right click viewport - pen) ? you can also replace the fancy crumbly paper background it contains with anything else. if you render that big enough it might be enough? you can also set the open gl antialiasing to the highest to achieve better results, also the outcome may depend on how your display mesh is set.

also in version 6 which you can try as a trial you can use ViewCaptureToFile with option transparent background and save it as a png.

To follow up:

I would like to do a color illustration in which I have used different colors to represent different types of structural members. The problem is that members of the same color merge into a single blob. I would probably like some kind of render that gives subtle differences in color to differentiate the parts.

Is there some way to accomplish this directly in Rhino or through an export to some 3d party tool?