Visual Arts Center at Dartmouth College, for Machado & Silvetti Associates

View from the South West

View from the South West

This is one from a pair of images I did for Machado and Silvetti Associates of Boston.  Both were digital pieces, modeled and lit in Cinema 4D, then worked over by hand in Photoshop.

The Model in a Cinema 4D Screenshot

The Model in a Cinema 4D Screenshot

I use the modeling program not only to set up the camera station and determine the view in perspective, but also to light the scene in a way that picks up ambient color and lighting from the materials and atmosphere.  Rather than a flat gray road, I find that the color of the sky and nearby materials will affect the color of the light itself, and thereby modify the final colors of surfaces as the light literally reflects off one surface and then hits and finally “colors” another.  With complex models having many surfaces, lots of glass, and many objects, the process can take a while if not managed correctly.  About a year ago I had a Boxx workstation put together for me specifically in order to maximize speed and render time, and now these solutions which formerly took 8 to 10 hours can be computed in less than an hour at most, often just a few minutes.  This allows for an iterative progress, and numerous trials to achieve light, shade, shadow and color in a palette that works best for the intent.

Cinema 4D permits rendering of the model to a Photoshop-format file, with a number of channels saved as layers or selection masks in addition to the RGB image.  I typically will render the image with some basic colors, even textures, for materials, in order to help bring color into shade and shadows.  But I’ll almost always render it out with additional channels that contain the shadow information alone, major materials, depth information (to later introduce atmospheric perspective with paint), reflections, and even shadows from separate lights.

    The Model Rendered Bare-Bones (top) with Separate Channels for (from top to Bottom) Shadows, Transparency/Glass, and Material Selection (Slate, in this case)

The Model Rendered Bare-Bones (top) with Separate Channels for (from top to Bottom) Shadows, Transparency/Glass, and Material Selection (Slate, in this case)

This allows me to throw a simple brick mock-up texture or surface color on to the model, in order to establish a base color for the ambient light.  The channels will permit me to later go back in to the image in Photoshop and entirely remove the generic material and then re-paint in something more varied and interesting.  This way, shadows can be worked independently of each other in 2D, colors shifted, atmospheric depth painted in, etc.  It takes far less time to paint in a bluish haze than it does to, say, add a physical (modeled) light blue fog to the scene and have it solve for the depth effect by itself.  It’s also entirely independent of the image (on a separate layer), and I can omit the depth if I want at the last minute.  If the haze were part of the physical model solution, it’s in there, and not very “undo-able”.

As a first-pass toward establishing the composition and checking things over with the Architect and their client, I will sketch out a very rough composition test, introducing people, trees, cars, a little roughness and texture where it needs it.  I’ll suppress detail from the model I find extraneous, and which fights the center of interest, while adding in detail where needed by hand.  Colors are tweaked, either cranked up a bit, or dialed back in saturation, shifted one way or another with regard to hue, adjusted for color shifts, etc.  The selection masks made while rendered in Cinema 4D are an immense help in this part of the process.  The composition test is sent off usually via email, and because I’ll invariably have further questions or comments, I’ll include a marked up version of the composition and color test to get the issues out on paper.

Quick Composition Test Sketch with Entourage, and Markups

Quick Composition Test Sketch with Entourage, and Markups

This process of reviews and iterations takes place a few more times, each time dialing in on the final work.  In this case, the class of students sitting around the scene during an outdoor sketch session was ultimately abandoned.  It was also determined that we should see some of the existing Dartmouth Stadium at the far end of the street, to establish the context a bit more concretely.

While working in Photoshop, I take pains to avoid any use of the stock filters.  They are notoriously limited in range, and look nothing like what they purport to be.  The “watercolor” filter looks nothing like watercolor, for example.  Even if it did, the use of such a filter would merely create an image that looks like every other “Photoshop watercolor” out there.  The intent isn’t so much to replicate watercolor, which this doesn’t pretend to do, but to create an image which is expressive, gestural, less-literal and more conceptual, and which clearly shows evidence of a human hand.

When working in photoshop, I will usually use a simple round brush with a hard edge, pattern stamp brushes, texture overlays, and a system of layer modes that is so interwoven, it can become a house of cards if I’m not careful about it.

Left, a typical Paintbrush; Right, Layers and Layer Effects

Left, a typical Paintbrush; Right, Layers and Layer Effects

Below, the final image, …with details following.

South West View, 13x19, 400dpi

South West View, 13x19, 400dpi

South East View, 13x19, 400dpi

South East View, 13x19, 400dpi




Filed under Commissioned Work

9 responses to “Visual Arts Center at Dartmouth College, for Machado & Silvetti Associates

  1. Pingback: Framing the Visual Arts Center :: Dartmo.

  2. Denver Abaño

    i do the manual perspective, with no software, but i think this is great stuff and i could use the info to make my works come to life even more, the coloring’s great. thanks.

  3. Denver Abaño

    wow, this is great stuff.

  4. Pingback: How the images of the Visual Arts Center were made :: Dartmo.

  5. reza asgari pour

    realy wonderfull works. i didnt believe that software can work so good.

    • Thanks very much. I don’t really have a process, and don’t use the software in any step-by-step sense. I like to think it’s a case of experimenting. And it feels like xperimenting every single time, too. Your point about the software is interesting. I don’t know how many copies of Photoshop that Adobe has sold, but I think you’ll see that it’s less about what the software can do, than it is what the person can do with the software. If I can offer one tip, never simply run the “watercolor” filter in an attempt to get a softer look. I almost never use the filers themselves. If I ever do use a filter, it’s almost always in combination with others, and only as a first step. I might run a subtle pass over a layer, and then run it through a few more filters after that, building it up. But always, the human hand needs to come into play and work the pencil or brush, just as it would in ‘traditional’ media.

      Thanks again.
      Jeff Stikeman

      • I agree. The filters modify the texture, but it would take some artistic guidance to get the best effects from them. I can see the models helped create a structure with realistic perspective, but I really appreciated how you softened the cubes and fit them into the environment.

      • Thanks Lucy, very much.
        The computer can be pretty powerful as a tool, but in the end, I like to break down all that mathematical perfection or quasi-realism and do things by hand.

        …I appreciate your comments.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s