I am trying to model an old roman fortress (castra stativa).
Currently I have limited success: https://osenski.artstation.com/projects/QrkXbx [osenski.artstation.com]

The original wall that I am trying to model looks like this:

So far I was able to approximate the wall with the following approach:
1. use grid of points to mark the stone places.
2. Use point jitter to create some randomness
3. offset every even row of dots on left using group + transformation.
4. Model simple stone using box
5. Use "mountain" node to introduce randomness.
6. Use polybevel to make the stones rounded
7. use "copy to points" to copy the stones to the points of the grid.
The problem is that the above approach does not account for:
1. The different size and shape of the stones.
2. Stones not overlapping each other.
4. (my biggest problem) using differently shaped stones and placing them in the way that the space between them is minimized.

Any help will be highly appreciated !
Thank you in advance.

Hi,

Noob here, but I think this will help, especially the beginning part:

https://www.youtube.com/watch?v=4VS4b3YcIFk [www.youtube.com]

The short answer is that instead of creating bricks and stacking them you should probably experiment with the concept of creating a rough geometry of the wall and fracturing it in a brick-like pattern (e.g. some variations of the Voronoi Fracture SOP).

This way you won't have to worry about the placement and overlapping of bricks at all, even if they'll be of irregular shape.

Hello
Here is my take on this. Sorry ran out of time but next step would be to automate copying in the Y direction. And need to then swap out the boxes for a selection of 'bricks'

Probably a simpler solution, but a starter for ten.

I'd imagine there is a way to do this with a for loop... but I quite like vex so went that route.