Lets get to the final result first, and then talk tech.
I believe this is the worlds first example of an orthographic photo of a firearm.

As a reminder, these are the advantages of telecentricity:
With a telecentric lens, there is zero distortion, perspective, or warping. Every single pixel is as if you were looking directly down on it.
It also means everything is dimensionally accurate, and nothing changes size even if the object gets closer/farther.

Image above is not my own. Normal image in the center, telecentric on the right.

To create this, I took roughly 20 images, and used Photoshops “Align Images” and “Blend Images” tools religiously.
Each image had to be cropped to the center 50%, roughly, as the lens is for a 2/3’s sensor, but my camera has an APS-C.
The composited image before downscaling and compressing was ~250 million pixels. It took roughly two hours to piece, blend, and edit the image for the final result.

It’s not perfect, but I am very happy that the concept works at all.
What’s wrong with it?

• Need to dial in focus to increase detail and quality. There may be many pixels, but what worth are they if the pixels are blurry? I know I can do better.
• Correct chromatic aberration better
• Lighting quirks (holy shit this thing is so blue on the edges it’s impossible to use any other kind of background, generated or not)
• Balancing angle correction between the various images as they get merged
• Verify color bitness? (set to 8 in PS but maybe I go higher?)

Anyways…
What’s next? Larger items. More items. Automated stitching. A motion gantry. Experiments with alternate software’s to assist in the process.

BTS

Thank you for your time and interest.