Case Study: Combining CGI and Photography
16 Jun 2015
Our CGI artists worked with photographer Paul Harvey to create this complete image.
Creating the image - on location
The concept was started from an initial open discussion held at Junction Eleven’s Banbury studio. The photographer and our CG artists discussed suitable cars and angles and then began the search for locations that would fit the bill. The location chosen in the end was in Dartmoor National Park.
As they were shooting in the summer, and had to be up before first light in order to be in place for the sunrise, the road was pretty much deserted (other than the sleeping sheep!)
The photographer shot the high res back plates on his Hasselblad medium format camera, one as a clean plate and then one with some reference objects in to aid camera matching in 3D later.
A series of images at different exposures were then shot to create a 360 dome, known as a High Dynamic Range Image, which was then used in the CG artist's 3D software to provide “on-location” lighting and environment reflections information, and help to make the car appear if it was there.
The team worked together over 3 days to capture both sunrise and sundown versions of 3 locations in order to find the best light for the shot.
Creating the image - in the post-production studio
Once the team returned from location, the RAW shots were processed and the selected shots for the backplates were composed into the shot we wanted.
The CGI car model chosen was a rare Alfa Romeo 8C. The CG artists created CGI based materials to replicate the various finishes found on the car such as car body paint, tyre rubber, windscreen glass and alloy wheels. These are then applied to the model parts, a virtual sun was positioned to match the sun from the location and the HDRI was wrapped around the model to apply reflections and light tones.
A virtual camera was created and placed within the 3D scene to match the position of the photographer relative to the car. This was used to frame the shot as what the CG artists see through the virtual lens is what gets processed by the software.
Finally, the image was rendered; where all the elements gets finalised and the results are processed and saved.
With a little final creative retouching the image was complete.
We hope you like it. Let us know if you have any comments, we'd love to hear from you.