I have to admit I'm more than a bit jealous of the things he's been able to do (also BreakingTaps) with optics at home. We've sort of hit this point in optics where people can build stuff in their garage that does very cool things for "not that much money". It usually combines high quality imagers (modern mirrorless cameras are amazing for the price), inexpensive optical trains (like the video says, you can't really use super-cheap objectives, but high quality objectives are available for not obscene prices), inexpensive stages (mainly thanks to modern 3d printers), illumination, microcontrollers, high end computers with lots of cores, GPUs, and storage, and clever algorithms (which often resemble machine learning, or are machine learning).<p>The real challenge is in learning how to put all these together to make a system that does something interesting. For example, I made a goal of designing a self-built microscope using 3d-printer style motion, inexpensive widely available LEDs and cameras, all within a modest budget (for me that's about $1-2K per year) compared to what you would spend if you worked in a well-funded lab. I'd say it took me a good decade to get proficient with all the bits and pieces required to build my scope (which uses an object tracking algorithm to move the stage around to follow individual tardigrades). There is a fractal-like complexity to each of the parts, and while some amount of mechanical or electrical engineering knowledge can greatly reduce the cost of building these things, reaching the accuracy and precision that is seen in well-funded labs is still challenging.<p>For those who spent the time reading this comment, here's a reward: a tardigrade hatching from its egg and exiting the moult of its mother (the empty tardigrade-shaped husk around the eggs): <a href="https://www.youtube.com/watch?v=snUQTOCHito" rel="nofollow">https://www.youtube.com/watch?v=snUQTOCHito</a> or jump ahead to <a href="https://youtu.be/snUQTOCHito?t=690" rel="nofollow">https://youtu.be/snUQTOCHito?t=690</a> to see the actual exit.