This stunning ad was constructed by The Embassy Visual Effects. The entire spot is computer generated. The environment utilizes high-res digital photos and photogrammetry and the foreground elements, including the Citroen, are 3D modeled and rendered. The car itself was scanned by Gentle Giant and then decimated by Viewpoint. The Embassy had the dubious task of cleaning up the scan data, adding details back into the model, extending the detail to include a fully articulated robot, applying motion capture data and then rendering and compositing the elements to be a cohesive photo-realistic spot.

Clearly they reached that goal. Hats off to these FX mavericks. Luxology had a chance to catch up with uber-busy Embassy founder Winston Helgason to ask him about the project and how they were able to utilize modo.

Lux: Why did you bring modo in for this project?

WH: We had been told about the edge tools in modo and how easy it is to add detail to existing meshes. Since we were starting with scan data from Viewpoint, we needed to be able to quickly clean up the high-res meshes and then add detail where appropriate. We had heard a lot about the modo toolset in this regard and we were not disappointed.

Lux: How long did you have modo before you started this project?

WH: Actually we didn’t have modo when we started the project. We had heard a lot about modo and once we got into the project it seemed like modo could potentially save us a lot of time and frustration. So we decided to bring it in and give it a go.

Lux: It is unusual for a studio to bring in a new tool during the production of a project due to learning curve and pipeline integration issues. How long did it take to get up to speed with modo?

WH: We installed modo in the morning and by the end of the day we had already modeled several pieces of the car. We basically broke the car into segments and assigned various components to different artists. Within the first day of using modo we had completed the wheel rim, side mirrors and a ton of clean up on the door.

Lux: How difficult was it to integrate into the pipeline from a technical/data standpoint?

WH: Surprisingly there were no issues with pipeline integration at all. modo supported all of the data we needed natively so we could go back and forth between modo and LightWave 3D seamlessly. This was really quite critical. The only issue we were concerned about was that we used the modo subdivision surfaces extensively and so we often modeled with n-gons (more than four vertices in a polygon). Using n-gons provided a lot of freedom during the modeling process, but we were concerned since LightWave does not support n-gons. However, we were pleasantly surprised to find that we were able to use modo to split the n-gons into quads and then use tools like edge turning to optimize the flow of the mesh before sending it to LightWave for rendering.