← Academy · Specs
Academy · SpecsAspect ratio correction: the dial that makes the doubling checkbox obsolete.
Some cameras offer a checkbox that “doubles the scan rate” — and their own manual admits the sensor actually slows down. We built the honest version: a dial that goes both ways.
Somewhere on a spec sheet you may meet line rate doubling: tick the box and the scan rate on screen doubles — 3,000 becomes 6,000. The feature isn’t fake, and the problem it solves is real. But it’s the crayon version of something much better.
The real problem: proportions are a moving target
A competitor’s width in the image depends on the ratio of two speeds — how fast the camera slices, and how fast the competitor crosses. Scan slower than the race and a horse comes out squeezed; faster and it comes out stretched. Neither distorts the evidence — every slice keeps its timestamp — but a squeezed image is hard to read. And the correct correction factor is a property of each race, not of the camera. Which makes a checkbox the wrong shape for the solution.
What the doubling checkbox actually does
It inserts a computed line between every two real lines — interpolation — and writes those computed lines into the recording. Per the offering vendor’s own manual, its real engine is light, not speed: with doubling active, the camera scans at half the displayed rate, giving each real line twice the exposure. The display says 6,000; the sensor saw 3,000. One frozen factor, baked into the evidence, and the doubled figure leaks into marketing.
The dial: aspect ratio correction from RAW
Our system keeps only real optical lines in the recording; the viewer corrects the aspect ratio at display time — a continuous dial, any factor, either direction, per race, reversible forever. Right factor instead of a fixed guess, screen instead of evidence. And the dial has a second direction the checkbox can’t go: when the light is good, deliberately scan faster than the race, then squeeze the image back. In that squeeze, several real optical samples merge into each displayed column — genuinely finer detail, built entirely from lines the sensor saw. Stretching adds comfort; squeezing adds information.
The workflow this unlocks: never touch the iris again
Set the aperture once for the depth of field your track needs — it’s now a quality control, not a brightness control. Answer light with the rate: sun comes out, add lines (each is another real sample); cloud rolls in, ease back toward the matched floor. Surplus light becomes resolution instead of being thrown away by a smaller aperture. And a quiet bonus: since nobody rides the iris from the control room, a simple fixed lens does the job.
1. Doubling is active and the display shows 6,000 lines/s. What is the sensor doing?
Scanning 3,000 times per second with double exposure per line — the other half is interpolation. Useful in the dark; nothing to do with speed.
2. Why is a continuous factor better than a fixed ×2?
The right factor depends on how fast each competitor crossed — a property of the race, unknowable at the moment a checkbox commits to 2. A display-time dial measures and readjusts per race, forever.
3. Squeezing an oversampled image and stretching an undersampled one — same operation?
No. Stretching spreads existing samples (cosmetic); squeezing merges multiple real samples into each column (a genuine quality gain).
The explanation worth remembering
Doubling is a checkbox that guesses ×2 at capture and writes the guess into your evidence. Aspect ratio correction is a dial that knows the right answer after the race — stretch to rescue a dark scan, squeeze to turn surplus speed into real resolution — while the file underneath stays pure optical truth.
Next in the Academy: Line rate: the spec that doesn’t work like megapixels →