"What we saw was this pretty striking pattern that software engineers got about 20% more feedback if they were sitting near their colleagues than if they were distant from them... It was [young workers] who had the most to learn that really saw this deficit in feedback." www.npr.org/2026/06/01/n...

Using a wavefront sensor and bisection to place the focal plane of a microscope tube lens in the back focal plane of an objective lens. I used beam diameter as the readout because the radius of curvature did not change significantly over the range of lens positions that I tested. #microscopy

I'm still trying to work out whether this is truly useful, but you can use a coverslip to pick off the beam from an alignment laser that is back reflected from a mirror at the end of a setup (not shown). Deviations of the beam upon insertion of a lens indicate its misalignment. #optics

I added a galvo and scan lens example to my Cherry ray tracer to help one of our PhD students understand scan systems. Click and drag on the theta rotational offset value for the first mirror surface to see what happens to the ray fan in real time. #optics #microscopy kmdouglass.github.io/cherry

After 9 years, I am finally finishing the story about normalized cross-correlation (and reviving my blog). The new blog post about deep technical details on the most used image registration technique. I know y'all have been waiting. katpyxa.info/feedbacks/?p...

"[I]n any science less complete than physics, the more general mind still has its uses, though everyday the chances grow less." C. P. Snow, lamenting the loss of synthesis within a science as it matures.

When a single line of code you wrote without thought three years ago leads to three weeks of nightly debugging after the kid goes down. Ray-Surface Intersections with the Newton-Raphson Algorithm kylemdouglass.com/posts/ray-su...

An apologia for theoretical biology. From "Energy Flow in Biology" by Harold J. Morowitz.

"Optimal lens design and use in laser-scanning microscopy" doi.org/10.1364/BOE.... #optics #journalclub

I implemented folds into the Cherry Ray Tracer today. Rather than freely placing components in space, I took inspiration from video games and used a RUF coordinate system to build up the system surface by surface. kmdouglass.github.io/cherry kylemdouglass.com/posts/3d-seq... #optics

Does anyone have any tips for putting a galvo in the center of a 4f system in the scan/descan arm of a laser point scanning fluorescence microscope? I normally align everything to the objective, but here I wonder whether the galvo should be the primary datum to align to.

I can never remember the electrostatic conventions for the mitochondrial membrane potential, so I wrote them down: kylemdouglass.com/posts/electr...

I finally managed to rewrite the Cherry Ray Tracer GUI using egui, an immediate mode GUI for Rust with WASM support. In addition, I figured out threading in WebAssembly via SharedArrayBuffers, so this thing is now lightning fast. kmdouglass.github.io/cherry #optics #wasm #raytracing

Last week I learned that I can use the internal counter on a NI-DAQ board as a triggerable pulse sequence that serves as the clock for an analog output waveform. Pretty cool! #microscopy kylemdouglass.com/posts/trigge...

I finally got around to writing the discussion section of our lab's basic training course on live cell #microscopy. I explain the tradeoffs involved in designing an imaging experiment and the art of thinking of the experiment as an optimization problem. leb-epfl.github.io/basic_traini...

I had the happy realization the other day that conservation of radiance / étendue is purely geometrical. Radiance is just proportional to the size of the set of all lines through two area elements, and the product of area and solid angle is an invariant over this set. #optics #radiometry

A holiday demonstration of the different penetration depths of red and blue light into tissue. #optics

When X-rays are incident upon a crystal, some of their energy is reflected from the lattice planes. Certain angles of incidence will lead to the observation of diffraction peaks due to constructive interference between the reflected wavefronts. #optics #microscopy youtu.be/FZ4gilAe7x0

The construction of the support of the 3D aperture (a.k.a. the optical transfer function or OTF) using the Ewald sphere in a brightfield microscope under partially coherent illumination. #optics #microscopy

How to use the Ewald Sphere to determine the support of the 3D aperture of a brightfield microscope under coherent illumination. #optics #microscopy

Object models for image formation theory in brightfield microscopes. kylemdouglass.com/posts/image-...

I spent the morning playing with GeoGebra for the first time and, along the way, published an app about aligning a laser beam to an axis in space defined by two irises using two rotating mirrors. There's a learning curve, but it's not too bad. #optics www.geogebra.org/m/tfrykp7y

Check out my dope new fixed focal length lens.

It's kind of nuts that most of the cost of kinematic #optics mounts usually covers the alignment phase of a setup, after which the mount serves the same role as an unmovable piece of stock metal.

Trying to get a sense of how much current I can pass through salt water and happened to observe electrolysis by accident.

Do any of you fellow #optics nerds know whether First Contact solution will remove residue from alcohols left behind on glass?

The more I learn about floating point arithmetic, the more I am impressed that scientists can get anything done at all on a computer.

Pro-tip: you can use the hole in a Thorlabs FC/PC fiber mount to self-center a domed LED when gluing it to a cage plate blank. #optics

This week I wrote a small Python library for computing vector EM fields in the focal region of high NA objectives. github.com/LEB-EPFL/jus... I'd be pleased if would try it out and leave any feedback or bugs in the issue tracker. :)

> So Lemkin's attempts to communicate with the AI model... were fundamentally misguides. Maybe the real, lasting value in LLMs is that they have brought fundamental questions about cognition from scientific specialty to the working lives of everyday people. arstechnica.com/information-...

I've resorted to printing out code and checking it by hand just to validate AI slop. #thefutureisnow

Can anyone recommend a service where I upload 3D models of simple machine parts and they mail me the CNC'ed results? Something like Osh Park that does small batches but for machine parts? I feel like I saw something about this recently but can't remember which platform I saw it on.

Here's a short post about converting the surface sag of a class of common optical surfaces into the normal form of a quadric. The trick is to use a method from high school #math that doesn't get as much attention as, say, the quadratic equation or the FOIL method. kylemdouglass.com/posts/comple...

You know you're f***ed when the Wikipedia article on the problem you're working on only exists in one language and it isn't English. de.m.wikipedia.org/wiki/Umrissk...

I'm brushing up on my computational #geometry skills and found this great resource for surface representations: "Representing Smooth Surfaces" by Ken Takusagawa Fun fact: surface sag in #optics is just an explicit surface representation. groups.csail.mit.edu/graphics/cla...

I realized recently that I have always taken for granted the fact that camera read noise is Gaussian, but I never questioned why this is. The answer, I think, is due to the central limit theorem. #optics #imaging kylemdouglass.com/posts/why-is...

This month's Henry Poincare Webinar Series on Optical Polarization and Related Phenomena will be led by Aristide Dogariu! 💡 Join us on 24 June for this presentation on light polarization and the statistical properties of the electric field vector.⚡ Register here: spie-org.zoom.us/web...

I did a deep dive into the mathematics for 3D optical system design software, focusing on ways that we can make layouts more intuitive. I hope you find it as interesting as I did in writing it. #optics kylemdouglass.com/posts/3d-seq...

Sometimes you come across a site and you just have to instantly bookmark it. Mathjax examples for easy copying and pasting: easy-copy-mathjax.nakaken88.com/en/matrix/

Claude, you cheeky little troublemaker.

One piece of advice I have for people living in Switzerland is to always stay grounded.

I just added reflective surfaces to my ray tracer project and a concave mirror example. I had to choose a slightly unconventional sign notation to get the paraxial ray trace algos to work, but all the examples match what I calculated by hand. kmdouglass.github.io/cherry

The 2025 Scientific Computing in Rust Annual Workshop will take place June 4 - 6. scientificcomputing.rs/2025/ I'm interested to see where the community is headed 🤓

I'm finally getting the hang of CSS. I fixed a long standing issue for rendering the specs table for my ray tracer on mobile just now, and added preliminary spot diagram support. Now I just need to make the spot diagram look good on desktop... kmdouglass.github.io/cherry/

I decided I'd go for a bit of light reading for my next book. 💣

Can anyone recommend an online database for tracking plasmids in a small lab? It should support search, custom fields and ideally visualization of SnapGene files. I'm aware of Benchling, but am wondering if there are others.

A bit of coding fun today: the spot diagrams for a 10 mm diameter, f=+50 mm N-BK7 planoconvex lens with 0 and 5 degree field angles for each of the Fraunhofer F, d, and C lines. Ray trace results computed by the Cherry Ray Tracer. github.com/kmdouglass/p... #optics

The older I get, the further back in time I go in my literature searches. I'm currently in the 1950's and realizing that H. H. Hopkins contributed to nearly all aspects of modern #optics theory.

I added the ability to model point source fields to my ray tracer just now. I also added a biconvex lens example to demonstrate it. Check it out and let me know whether you find it useful. kmdouglass.github.io/cherry/ #optics

I wrote a rather technical blog post yesterday about data type optimization for improved cache locality and reduced memory footprint in ray tracing with Rust. Long story short: keep the memory footprint small but only if you actually need the performance boost. kylemdouglass.com/posts/data-t...