Print from first iteration test

 

 

 

 

 

 

Iteration 1, 2 & 3. With the initial concept for a machine design and ideas and experiences from exploring the parts we conducted a first ad hoc experiment in three iterations to test if the machine could produce a valuable result for us.

Iteration 1, render out 500 images, we then select ten images of value (hard to know on this stage if an image actually can produce a landscape the neural network does a parallel search and selection. We then chose one of the selected images and used it in Modo 3d  as a displacement map creating a basic 3d model.

 

Iteration 2, Put the 3d generated model on the disc and let a basic animated shape be used as a 3d dimensional mask (artistic interaction) to restrict where the shadows can be cast. Render 500 images then select 10 images of value (a bit easier in this stage) Choose one of the selected images and run it again in Modo 3d  as a displacement map creating a new more complex 3d model.

 

Iteration 3, Put the newly  3d generated model on the disc and let a new basic animated shape be used as a 3d dimensional mask (artistic interaction) to restrict where the shadows can be cast. Render 500 images then select 10 images of value (a bit easier in this stage) Choose one of the selected images and run it again in Modo 3d  as a displacement map creating a more complex 3d model.

 

Print from full 3 iteration completed into baked and freezed geometry