TINY WORLD / EVERY EXPOSURE CHANGES THE STORY
Collect signal, keep a view, then count the tiny bright spots inside.
Start with a short exposure. Save a view when the small spots look clearer.
Invented beam-sensitive specimen · fixed display scale · relative exposure units
More counts can reduce grain. They cannot restore information already lost from the specimen.
YOUR FIRST LOOK
The beam also weakens fine detail in this invented specimen. Pause whenever you want: the beam is blanked while paused. A fresh copy gives you another try and keeps your notebook.
There is no countdown or penalty. The third challenge asks you to compare two views from the same copy, at least 8 exposure steps apart.
ABOUT THIS SIMULATION
This TEM-inspired game models an invented beam-sensitive specimen. It is not a real micrograph, a calibrated dose experiment or an instrument procedure. All exposure steps are relative game units. The small spots do not identify atoms or molecules.
Each step adds independent Poisson counts from the specimen at that time. The expected fine-feature signal declines exponentially with accumulated exposure while a broad particle remains. The displayed image equally averages all recorded steps, using a fixed brightness scale. There is no dose weighting, motion correction, contrast-transfer calculation or detector-noise model.
Saved counts are never erased by later damage. A later cumulative image can have lower fine-feature contrast because it includes more signal from the changed specimen. Real frame selection and dose weighting can retain useful earlier information. This game does not establish a universal best exposure.
In the game, pausing also blanks the beam: no extra signal or damage accumulates. Pausing does not repair existing damage. Stopping recording on a real microscope does not necessarily blank its beam. A fresh copy resets this fictional specimen while preserving your saved images.
Real damage rates depend on material and imaging conditions. Damage can happen within an exposure and may weaken fine information before large changes look obvious. Real low-dose methods may focus and track away from the area of interest; that workflow is not modelled here.
Checks consider recorded feature contrast and background grain, together with your count. They do not seek an exact hidden exposure. The final challenge needs two views from one copy at least 8 steps apart, an arbitrary game rule. Text descriptions provide the same visible evidence without a penalty. Progress lasts while the page is open. This is an exploratory puzzle, not a secure test of microscopy skill.
The live exposure will start again on a new identical model specimen. Your kept images stay in the notebook. This does not repair the old specimen.
This clears all kept images and progress and creates new patterns.
MVLscope