Above is one of a few images I was asked to create for client needing city approval for an interim space.  The client requested to see what it would look like on site.  Below is the process for creating/compositing the image…… Please, respect the mobile home.

It started off with a wonderful photo to work with.  This is where the “mobile home” needed to be.  Right next to the wonderful cargo truck.

The existing photo was hacked into in order to place the model of the building in place.

People brought back in for scale reference, and camera points were created based on eyeballing the perspective distances based on knowing the centerlines from parking space to parking space where approx. 9ft.

This image was imported into the viewport background in model space and a camera match utility was used to create camera to line up with model and background.

Next, the camera tweaked a bit….

Once the building/model was were it needed to be, the surrounding “dummy” environment was created to mimic site conditions from photo.  All of which became invisible at render time.

Render settings for environment and lighting.

Hit render and the following elements were produced:

Color (overall render)

Alpha (for cropping building into photo)

Material ID (for color adjustments and easy object matting)

Shadow element was produced separately by hiding the staged environment and applying a shadow enabled material to the ground plane sitting beneath the building.

Once the rendered elements were produced, they all went into the existing photo.

Here’s one way to mask out your objects with your alpha channel from the maxwell think site: http://think.maxwellrender.com/maxwell_render___photoshop_alpha_action-110.html

The technique I have always used varies:

1. Open up both your basic color render and your alpha render

2. Select the entire alpha render

3. With your color render open, create a layer mask

4. After creating the layer mask, open the channels tab (next to the layers tab) and turn on the little eyeball for the alpha channel at the bottom (enable it)

5. Paste the alpha render into the alpha channel you just set up.

6. Go back to the layers tab and select the layer image….. Now you gotcher self just a buildin’

This keeps all original image integrity and (in my experience) is a cleaner and more effective way of adjusting output all within a single layer…. At any point in time if you F up, all you have to do is delete the info in the alpha channel and paste in new.

Finally, the shadow element was placed on top layer with “multiply” blending mode and color adjustments, graphics were added.

Each graphic/addition consists of two layers: bottom layer set to “overlay” at about 75% opacity, with the top layer set to “multiply” at about 30% opacity.  All masked out easily by using the material ID to make selections for certain parts of the model.

Here shows the overall process…..  Hope this was helpful…….


  • patrick anderson says:

    Very nice and useful tutorial. The image blends into the photo beautifully

  • Nuno Braz says:

    Way to go Tom! Very altruistic as usual :) Those are basically the same straightforward principles i normally follow when doing compo’s. I’m just curious, why Maxwell?

  • Tom Rusteberg says:

    Thank you Patrick. Glad to hear you found it useful.

  • Tom Rusteberg says:

    Hey Nuno, I overlooked your comment earlier. In my opinion, Maxwell provides the most linear workflow for image compositing due to the way it translates materials and lighting, which are physically correct when applied appropriately.
    Knowing the lighting and environment conditions that exist in a photo, Maxwell makes it easy to translate those conditions and apply them to the subject(s) being placed/composited into any given environment….. If that makes any sense…..

  • Nuno Braz says:

    Of course it does :) I normally use mental for that (commissioned work that is…) with heavy post but the principles are basically the same.

  • Comments closed


    ©Tom Rusteberg | tom[at]wanderplay[dot]com