Reflectance Transformation Imaging (RTI) Dome


Image Scanning Da Vinci drawings in Windsor Castle (UK)

For fun, I made a RTI (Reflectance Transformation Imaging) Dome (geodesic semi-hemisphere). What it does, is very simple. It creates an image which resemblance is computed a function of the direction of the incoming light. So other than a normal picture, you can change the direction of the light after the fact. This is actually super interesting if you care about subtle surface topology, as you can compute material properties and surface normals, which can then be used to calculate depth (by integration).

The files it creates can be stored in PTMs (Polynomial Texture Maps).

Such domes are usually super expensive, tens of thousands of dollars, and they are clunky and usually of poor hand-made quality. Because I have knowledge of both software and hardware, I decided to make some software and printed circuit boards, and come up with an awesome and simple design.

The dome costs less than $100 dollars total to make. Disassembled, it fits in your pocket, and can be mounted on any DSLR camera. I’ve used it to scan Aztec Codices and works by Michelangelo, Rafael and Da Vinci.

Recto: Sketches for the Last Supper, and other studies. Verso: Calculations with architectural, engineering, and geometric sketches c.1492-4


As with any design, it took a few iterations to get it right. I started with a transparent plastic half-dome that i spray painted black with tar. I put some LEDs on the inside, and a massive 10x10” (30x30cm) printed circuit board.

Above was my first design. Took ages too construct.


I made two versions of the ‘mini rti dome’. One that used some 3D printed components, and one without. The one with the 3D printed components wasn’t stiff anough, while the other (2nd design) was rock solid. The following pictures I intermixed the two, but the idea is the same.

Image The dome consist entirely out of printed circuit boards

Image Stitch them all up in a pattern

Image Finally put them together as a dome

Image Daisy chain the connections

Image Add (3D printed) camera mount and camera.

Image Scan works by Da Vinci


I’ve made a small video to introduce how this thing works. The quality of the video is terrible, and I’m even speaking Dutch, apologies. 736


Image Windsor Castle head of collection Alan Donnithorne introducing a drawing Da Vinci

RTI of a portion of item 912616. (c) Windsor Castle, Paper Conservation This is an image of a drawing by Leonardo Da Vinci, where he has studied hands. You can see several fingers in the image. It was drawn with a silver-point pensil. The silver is gone, but the indentation, as you can see, remains.

TODO(tzaman): add some RTI images here


Component A

Download Gerbers: (zip)

Component B

Download Gerbers: (zip)