Commendium

Riegl scan to 3D film – The process

Home > Articles > Riegl scan to 3D film – The process
Share this content to social media
Share on Facebook
Facebook
Pin on Pinterest
Pinterest
Tweet about this on Twitter
Twitter
Share on LinkedIn
Linkedin
Email this to someone
email

A glorious film project landed on our desk at the beginning of December. A mixture of photogrammetry, a LIDAR Scan, and drone footage. The task: To take all the above, fix, remodel, and turn the resulting 3D objects into a film.

We had collected data earlier in the year when we visited the cave in winter to gain the advantage of low water levels and sparse vegetation. The first model, created by drone, using thousands of aerial shots was of the entrance dolines, huge depressions in the ground, along the bottom of which the River Reka flows before entering the caves. The second model of the cave, a stupendously huge file, created using a Riegl Vz-400 terrestrial scanner converted to a single 3D point cloud.

Upon opening the resulting point-cloud, it presented itself as perfect, but consisted of four million polygons, which meant slow processing even with our high spec computers. First task, the model had to be decimated, bringing it down to a manageable size.

The model, an exact replica of Škocjanske Jame, one of the largest known caves in the World, started out in magnificent detail. The decimation was handled carefully, a fine balance was required, to reduce the size and maintain the exquisite details captured by the Riegl scanner. Once the model was decimated, it was time for a thorough inspection. As we delved inside the cave, it felt like virtual caving, as the camera twisted and turned through the labyrinth of passages. Working one’s way around a cave system in 3D software can be quite disorientating, even when it is one of the largest in the world.

The second issue presented itself while working our way through the caverns. We found areas of the model that were blank because they had been hidden from the scanner; it is quite impossible to place the scanner in all the positions required to capture everything. We solved this by digitally creating realistic walls, under the guidance of our own expert caver, who has first-hand knowledge of the cave.

Once the erroneous areas had been cleared, the next task was to texture the cave. Due to the sheer size of the cave, it is impossible to capture the textures, as one would with photogrammetry. Using photographs taken from sections of the cave, our art department created bespoke images to be used as textures. This is a tricky task, because all cave photography has artificial light, usually an orange or green hue cast by the lighting used. The cave must look real, with the colours looking as though you were viewing it in perfect daylight.Skojanske 3d model before and after

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Texturing a cave of this size needs care, thoughtful planning, and a flair for environmental imagery. Careful seams were placed throughout the cave model, following the natural contours of the cave system. The textures were then baked onto the model to add height and increase realism.

With the cave textured, we moved on to animating and adding a river, humans, rocks, bridges and via-ferratas. The cave, being naturally dark, required additional lighting to highlight the beauty of the passages.

With everything in place, we began filming the scene using camera paths to capture all the required complex angles and the animations. Once happy with the animations they were rendered to MP4 film files. The resulting rushes were transferred to our film department, who spliced the clips and animations in Adobe After-Effects. With the sound department adding the final bespoke underground cave water sounds.

Share this content to social media
Share on Facebook
Facebook
Pin on Pinterest
Pinterest
Tweet about this on Twitter
Twitter
Share on LinkedIn
Linkedin
Email this to someone
email
Tags: