(1) The phase space of illumination optics

We investigate rays in an optical system by means of screens we can place anywhere we want (but smart as we are we put them to adequate places). A ray crossing a screen is described by FOUR variables (two spatial and two angular ones) which form the phase space of illumination optics. We can plot the distribution of a meridional section in a phase space diagram (PSD), See below : The effect of an optical element between two screen can be described as a transformation in the phase space. In the above example, we have four screens connected by the thre4e phase space transformation of free space propagation, lensing, and again free space propagation. The result is a rotation in the phase The phase space concept is prehaps not (yet) a design tool but a means for a deeper understanding of the effect of optics to light distributions.

(2) Do aberrations increase etendue ?

No., but they increase the effective etendue. This is the etendue we need to collect a light distribution using top hat angle and area acceptance definitions. The below phase space transformation describes the effect of distortion. A larger etendue compared to the original distribution (left) would be required to collect all light. Distortion can be, in principle, reversed by a suitable optics.

(3) Capacity utilization

As I‘m just organizing and planning my freelancing life, I am wondering how much time should be assigned to active projects ? I learned from a LInkedIn post by Gunter Dueck (German mathematician and author) that capacity utilization C and the number of projects on the waiting list P are roughly connected like this : Hence, at 95% utilization you‘re dead. Where should I aim to ? 50% ? 70% ?

(4) Office space

I rebuilt the office in our flat in Rorschach for the new job. Do you like it ? Below is a tiny 32core workstation. Years ago it was ten times the price and size ! And not to forget the 3D printers.

(5) Breakthrough starshot communication and diffraction

For a few years, I‘ve been fascinated by breakthrough starshot , a project planning interstellar travel by laser propelled sailcrafts. Let‘s assume the devices can be accelerated, brake around Alpha Centauri (sic!), reach a planet of Proxima Centauri and take pictures. How can these be sent home ? There were some considerations published (arXiv:2005.08940). But even if the sail can be used as a parabolic mirror - diffraction at a (educated guess) D=10m mirror would spread the signal to which would after 4 lightyears spread to a disk with a radius of So it‘s hard to receive a reasonable share of the beam by a large telescope, even if a whole fleet of sailcrafts fires synchronously. Any suggestions ? OK, breakthrough starshot is dead now as a project. Anyway.

(6) Bauhaus, Dessau

In summer 2025, my lovely wife and I spent a visit to Dessau. There is a beatiful „After modern brightness“ exhibition that show the beginning of interior lighting on an industrial base, Very inspiring. Until March 2026. A lot of OSRAM history. And in the main museum is a LDC (light distribution curve) tuning apparatus with a live plot of the curve.

(7) The Phoebus cartel

In 1925, a number of lighting manufacturers founded the Phœbus S.A. Compagnie Industrielle pour le Développement de l'Éclairage . The cartel introduced the common 1000 hrs. lifetime standard for incan - descent bulbs. This standardization (or cartel agreement ?) is frequently interpreted as an example for planned obsolescence - see the documentrary „ The light bulb conspiracy “ (2010). My humble opinion on that is fundamentally different and is based on lamp technology. When designing a filament, you have a few degrees of freedom such as the wire diameter and the length. This way, you select the operating temperature. Higher temperature means higher luminance and efficacy but less lifetime. „Medical“ halogen lamps have a much higher luminance but live only for 50 hours. Simply speaking, efficacy times lifetime is a constant. It‘s simply lamp design, not conspiracy.

(8) LEDs are better, aren‘t they ? A lamp story.

My lovely wife (an ObGyn doctor) needed a new lamp for her wonderful Carl Zeiss colposcope. So the responsible MedTech assistant bought a contemporary one of the correct voltage, power, socket, and size : The result (amount of light at the target) was close to zero . What‘s wrong ? I had to mansplain a lot. In contrast to the Phoebus standard , there are different lamps of the MR 16 size on the market: • color temperature 2700K and 3300K (correlates with the luminance of the filament) • smooth or faceted reflectors (MR16 means „multifaceted“ reflector with a front diameter of 16 / 8 inches) • halogen or LED retrofit There are lamps for general lighting such as this one : with a rather low color temperature and a faceted reflector. For a use in optical devices such as a micros - cope, a surgical light, a fiber light or a colposcope, you want to concentrate the light in a - say - 5mm diameter lightguide. Let‘s check this with the above lamp. I quickly designed a faceted reflector and a filament source of proper size and 5 Mnits luminance. (If you ever need a rule of thumb how to set up an incandescent filament, let me The result is just 22 lm into the lightguide - an efficiency of around 5%. But there‘s nothing wrong with the lamp. We just judge a fish by its capability to climb a wall. The medical or specialty halogen lamp features a smooth reflector and a 3300K filament. What can we expect ? The simulation shows 118 lm at 2700K and 360 lm at 3300K (23 Mnits) in the same lightguide. We pay for this with a much shorter lifetime. Now for the LED . In the acquired sample seem to be four chips of, say, 160 lm each, arranged behind a microlens array. The simulations shows 5 lm in the lightguide. No chance for anything. There‘s basically no secondary focus . It seems there are no „ LED medical MR16 “ lamps on the market. Prove me wrong. I may come up with a suggestion - stay tuned. And then big business lurks.

(9) More colposcope illumination

At her secondary job site, my lovely wife encountered another colposcope. Nice optics but the poor illumination left me speachless. Keep in mind that the target is a coin sized spot at a distance of 30cm.

(10) Pre-distorted targets

Optical designs (especially in the microscopic domain) often suffer from more or less heavy distortion. For endoscopes, this is sometimes mitigated by an aspheric front lens. In other (specifically wide angle) cases, distortion is simply accepted or electronically compensated. But how do you test such a lens ? I learned it is useful to have a pre-distorted target produced by an optical designer. I want to share some helpful tools. (A) Start by generating e.g. a regular checkerboard pattern, for example using this online tool. (B) Open the file in Gimp. If necessary, you can add a white or black collar around it. Go to --> Filter --> Distorts --> Lens Distortion. The parameters available in Gimp do not perfectly match the optical ones but with some try-and-error one may come close to the desired solution. Save the result to a bmp file. (C) Verify the result with some optics software. Export the optical design to some non-sequential software (Zemax or LightTools). If you are lucky enough to have access to LightTools, go to Tools --> Utility Library and there Miscelleaneous --> Image processor. Open your distorted file, set the actual size of your object as the zone width and height and „Make a spatial apodization file“ which can then be used to set up the object as a source with just that pattern. Set the image plane as a receiver and you are ready to go. Not too bad for the first shot. (D) My final suggestion for those who need actual hardware is to approach a Gobo maker. Those folks can replicate your data file as a Chrome-on-glass version at moderate cost. (E) Take care to illuminate it in a way that the rays from the source hit the entrance pupil of the lens after passing the object. Sounds like Köhler illumination.