Study for a current project, pretentiously entitled “Paradis House” (and it really is spelled that way, without the “e”). I was interested in depicting, using a “chunk” of the roof system cut from the larger project model, a wind vane/sculpture and how it pokes out through the central roof oculus.
16 Likes
very cool stuff!! Thanks for sharing! I’d love to hear more about your process for making these images, they have a really great feel to them.
1 Like
Thanks for your interest, Kyle.
These are produced with an elaboration of a hack to avoid long ray-traced rendering times that I first started using almost twenty years ago, when Rhino and other software would spend hours working out actual renders. It is simpler than it reads.
So I begin with customized viewport display modes that I have cooked up. I capture them as PNG files using the Capture Viewport to File tool, and then I stack the PNG files as individual layers in a layered PSD or PSB file in Photoshop (or in a similar layer-capable painting software package, like GIMP, Fresco, or Krita…but Photoshop has the best capabilities currently). And then I use painting tools (sometimes controlled with a Wacom stylus and tablet) to mask or reveal parts of the individual layers to get the composite effect I want. I tend to apply a minimum of materials and mapped textures in Rhino; most surface material-simulating effects are painted with Photoshop.
The key to having this all work is a custom viewport display mode I made that I call Layer Mask, which has no shading, curves, or edges, and where all visible objects are displayed with the flat layer color. For this particular image, here’s the “Layer Mask” captured viewport:
The Layer Mask will never be visible in the final image. But in Photoshop, I can use various tools (like Photoshop’s Color Range function) to isolate specific portions of the image by color on this Layer Mask layer, save the selection as an Channel (which in Photoshop is a kind of saved selection), and then apply filters or paint textures into the corresponding areas in other layers captured from the Rhino model.
Since my layers (and thus their colors in this custom viewport display mode) in the original model are strictly organized by material and object, I can use the Layer Mask in Photoshop to isolate or treat individual parts of the image based on their assumed materiality or identity in the original 3D model. If I need to affect “closer” or “farther” parts of the scene while painting in Photoshop, I can export a Z-buffer depth mask from the original Rhino model, save it as a Channel, and adjust it to select parts of the image at whatever simulated “depth” I need.
Here’s a screencap of part of the layer stack for this rendering, composed of captured Rhino viewports and texture images from other sources, in Photoshop:
On the left you can see the black-and-white layer masks made using the Layer Mask layer next to thumbnails for each layer.
If I have a really complex lighting issue that I can’t simulate in a “live” Rhino viewport with shadows (or at least “Arctic” ambient occlusion) turned on, I can always do a quick material-less Rhino Render of the same viewport at the same resolution as my viewport captures, and layer that into the composition. (This latter is tricky. Sometime true rendered images don’t exactly correspond to the viewport captures of the same view of the model. The two won’t exactly line up. I think it has to do with different anti-aliasing strategies between Rhino Render and Rhino viewports.)
5 Likes
so very cool… love it. the results are stunning
1 Like
I should have said “on the right” not “on the left” when referring to the layer thumbnails and masks in that screencap.
really appreciate your discussion of this strategy 
1 Like
In the replies to another post, I uploaded copies of the basic customized viewport display modes that I use to create images like this with Rhino and Photoshop.
Hi @LewisW
Beautiful work!
The look of your drawings is very unique. I think you could do some great animations of this subject right in photoshop since you’re all set up.
Thanks for posting and sharing your working methods.
RM
I’ve been experimenting with adapting this technique to animation in Photoshop…it’s a little challenging with the level of detail I typically put into the still images, if I do more than simply panning my Photoshoped Rhino output as background. More elaborate cameral movements seem to work best if I operate on individual animation frames, which gets tedious really fast. Working like that one starts to understand why Ralph Bakshi never completed his rotoscoped and animated version of The Lord of the Rings.
It’s been suggested to me that After Effects might be more useful with Photoshop here than Photoshop alone in animating these painted Rhino items, but AE is probably the only major Adobe product I have never used so there’s going to be a bit of a learning curve.
Hi @LewisW
I can imagine it’s quite difficult. I use PS for scanned animation cells so I know how tedious it can be to deal with hundreds of frames and I’m not doing the level of layers you are dealing with.
I don’t know about AfterFX seems you would need the stills before you can do anything but I guess you could use their 3dcamera on your current stills to do pan shots or zoom ins.
Before you splurge on AfterFX try Davinci it’s free and is the industry standard. Also I would check out Krita it has animation and tons of features and is also free.
I remember seeing Bakshi’s ring at the theaters it was pretty cool, too bad it went unfinished.
RM
1 Like
Really interesting and eye catching style. I had a look at your website too. I like how you explain how and why you did the projects and the results look great. Thanks for sharing.
1 Like
That’s really interesting, @3dsynergy. I’ve used Krita but not for animation, and I’m not familiar with Davinci Resolve. Actually, I have a license to After Effects that I have never used because it made economic sense to get a package from Adobe that included it (along with the five graphics software titles for which I have a legitimate need) as opposed to licensing the other software à la carte. But I’m not wedded to AE, if Davinci is a better standard. Something else to explore! Thanks!