# Exposure Desk > EVERY STOP IS THE SAME STOP. Exposure is one quantity written in four currencies - aperture, > shutter, sensitivity and filter - and they trade EXACTLY, so a look is not something you choose > but something you pay for. The consequence nobody budgets for: THE FRAME RATE IS AN EXPOSURE > DECISION. Shutter time is angle/(360 x fps), so holding the 180-degree shutter and doubling the > rate is exactly one stop, every time. Five lanes over one shoot, plus a free browser-side engine > doing exact stop arithmetic. URL: https://exposure-desk.skillsafe.ai/ API: https://exposure-desk.skillsafe.ai/api.html Derived from nexu-io/open-design (https://github.com/nexu-io/open-design), Apache-2.0, whose video-hyperframes skill composes video - animations, title cards, overlays, captions and audio-reactive visuals. A composition is built around footage, the footage was exposed by a real camera under a real light budget, and motion blur in an animation and motion blur in a camera are governed by the same number: the shutter angle. Not affiliated with or endorsed by nexu-io. ## The arithmetic ```text EV = log2(N^2 / t) what the SETTINGS admit, at ISO 100 scene = EV100 + log2(ISO/100) what the LOCATION provides error = settings - scene + ND positive is short, negative wants a filter ``` ## The frame rate is an exposure decision Shutter time is `angle / (360 x fps)`. THE FRAME RATE IS IN THAT EQUATION, so holding the 180-degree shutter and doubling the rate halves the time - EXACTLY ONE STOP, every time. 120 fps is 2.3 stops darker than 24 with the same angle, the same aperture, the same sensitivity and nothing about the lighting different. HOLDING THE ANGLE IS WHAT KEEPS THE MOTION CUTTING TOGETHER, and it is also what makes the rate cost light. They are the same decision from either end: an angle is a fraction of a frame, so it buys the same-looking motion at every rate and a different amount of light at every rate. ## The three ways to pay, and the one that runs out ```text aperture N -> N / 2^(stops/2) costs DEPTH OF FIELD sensitivity S -> S * 2^stops costs NOISE shutter a -> a * 2^stops costs MOTION, and runs out at 360 degrees ``` GIVE ALL THREE AND RECOMMEND NONE. They buy identical stops and spend different things. THE SHUTTER IS THE ONE THAT ENDS. A frame has 360 degrees in it and no more, so a shortfall over one stop from the standard 180 cannot be paid there at all. AND A FILTER ONLY GOES ONE WAY. Neutral density takes light away, so it answers the bright problem and never the dark one. A setup that is short is short. ## One filter, three names ```text density D -> stops = D / log10(2) because transmission is 10^-D fraction f -> stops = log2(f) because transmission is 1/f stops -> itself ``` THE SAME PIECE OF GLASS IS CALLED ND 0.9, ND8 AND THREE STOPS. Three scales are in simultaneous use and reading one as another is a several-stop mistake: somebody who reads ND8 as eight stops is FIVE STOPS OUT, in the direction that ruins the take rather than the direction that grades back. And ND1000 is ten stops - the thousand is a fraction of the light, not a count of anything. WRITE THE UNIT. `0.9` is ambiguous and `3 stops` is not. PAST ABOUT SIX STOPS THE FILTER STOPS BEING NEUTRAL. Plain ND is neutral to visible light; deep filters pass proportionally much more near-infrared, a sensor sees it, and blacks go brown. That is where the glass has to be an IRND. ## The shutter, the mains, and 172.8 degrees A SHUTTER ANGLE MEANS NOTHING UNTIL A FRAME RATE IS NAMED. 1/50 second is 180 degrees at 25 fps and 90 degrees at 50 fps - the same time and a different look. 172.8 DEGREES IS NOT AN AFFECTATION. At 24 fps it is exactly 1/50 second, which is a whole number of half-cycles of 50 Hz mains, so the lighting cannot beat against the shutter. 144 degrees is the same trick for 60 Hz. And at 25 fps a plain 180 is already 1/50, which is why 25p and 50 Hz go together and why 24p needs the dodge. FLICKER IS INVISIBLE ON SET AND OBVIOUS IN THE EDIT, and it does not grade out - but it is a question about interiors and night exteriors, not about every shot. The sun does not pulse. ## What this page cannot do Everything here is arithmetic over numbers you supply. THE SCENE TABLE IS A SET OF CONVENTIONS, NOT MEASUREMENTS. `overcast` is EV 12 because that is the number everybody quotes. A meter measures; this remembers. IT HAS NOT SEEN THE FOOTAGE AND DOES NOT KNOW THE CAMERA. Gain is reported as stops over ISO 100 and the noise judgement is left to somebody who has looked. EXPOSURE IS NOT DYNAMIC RANGE. Every figure is about where middle grey lands; whether the window in the same shot holds is a different question. ## The sheet ```text SHOOT scene | overcast mains | 50 SETUPS s1 | 24 | 180 | 13 | 400 | 0.3 | the wide s2 | 48 | 180 | 13 | 400 | 0.3 | the slow-motion ``` A setup row is `id | frame rate | shutter | stop | sensitivity | filter | description`. The shutter may be an angle or a time - `180` or `1/50` - converted through that row's frame rate. A bare number over 360 is a time that lost its `1/`. The filter may be blank, or a density, a fraction or a count. `scene` is required and has no default. `mains` is optional and turns on the flicker check. A MISSING NUMBER STAYS MISSING. A setup without all of rate, shutter, stop and sensitivity is in no calculation and the page says so. ## The lanes - `plan` - works out the light budget before anything is ordered - `check` - says what every setup actually exposes at - `shutter` - says what the frame rate is costing, and whether it can flicker - `filter` - turns the sheet into a filter order, in a named scale - `deliver` - sorts findings into as written, change a setting, change the kit ## What it does not do It meters nothing and lights nothing. Every figure is arithmetic over numbers you supply, so it is a starting point a meter and a monitor will correct.