Astrophotography Exposure: The 500 Rule, the NPF Rule, and Star Trails

Point stars or trailing stars is a decision the Earth makes for you. It rotates 15 degrees an hour, and past a certain shutter time that motion is wide enough to smear a star across more than one pixel. The 500 rule gives you a rough ceiling; the NPF rule gives you an accurate one. On a modern high-resolution sensor the difference between them is often a factor of two.

Maximum exposure before stars visibly trail, by focal length. The 500 rule is consistently more permissive than NPF on a modern 24MP full-frame sensor.

Why stars trail at all

The sky appears to rotate because the Earth does, at roughly 15 degrees per hour, or 15 arcseconds per second of time. A star near the celestial equator therefore moves about 15 arcseconds in one second of exposure. Whether that shows up as a trail depends entirely on how much of your frame 15 arcseconds covers, which is a question about focal length and pixel size, not about the sky.

A 14mm lens spreads a huge amount of sky across the sensor, so each pixel covers a large angle and a star can drift for many seconds before it crosses into the next one. A 200mm lens magnifies that same drift enormously. This is why the maximum exposure falls off so steeply in the chart above.

The 500 rule and where it comes from

The 500 rule says: maximum exposure in seconds = 500 ÷ focal length. On a 24mm lens that gives you about 21 seconds; on a 50mm, 10 seconds. If you are shooting a crop sensor, divide by the equivalent focal length rather than the marked one, so a 24mm on APS-C behaves like a 36mm and gets about 14 seconds.

The number 500 is not physics. It is a constant reverse-engineered from what looked acceptable on 35mm film, viewed at a normal print size. That is the important caveat: the rule was calibrated against grain, not pixels, and against prints, not 100 per cent zoom on a 27-inch monitor.

Why the 500 rule fails on modern sensors

Film had no fixed sampling grid. A digital sensor does, and its pixel pitch decides how far a star can move before the smear becomes measurable. A 12MP full-frame sensor has pixels around 8.5 microns across; a 61MP full-frame sensor has pixels around 3.8 microns. The same star drift that stays within one pixel on the 12MP body crosses two or three on the 61MP body.

Aperture matters too, though less obviously. A fast lens wide open produces a slightly larger, softer star image because of aberrations and diffraction, and that larger disc tolerates a little more movement before the trail becomes obvious. Stop down and the star point tightens, which paradoxically makes trailing easier to see.

The NPF rule

The NPF rule accounts for all three variables. The common simplified form is:

t = (35 × N + 30 × p) ÷ f

where N is the aperture f-number, p is the pixel pitch in microns, and f is the focal length in millimetres. For a 24MP full-frame body at f/2.8, pixel pitch is about 5.9 microns, so the numerator comes to roughly 275 — giving you a little over half of what the 500 rule allows.

Setup	500 rule	NPF rule
14mm f/2.8, 24MP FF	36 s	20 s
24mm f/2.8, 24MP FF	21 s	11 s
35mm f/1.4, 45MP FF	14 s	5 s
50mm f/2.8, 24MP FF	10 s	5.5 s

The NPF numbers look punishingly short the first time you see them, and they are — if your standard is pixel-level perfection at full magnification. If your output is a web-sized image or a modest print, the 500 rule result will often look perfectly fine. Decide which standard you are actually shooting to before you decide the rule is wrong.

Declination changes the answer

Both rules assume the worst case: a star on the celestial equator, moving at the full 15 arcseconds per second. Stars closer to the celestial pole trace smaller circles and move more slowly. The correction is a cosine of declination — a star at 60 degrees declination moves at half the equatorial rate, so you can double your exposure.

In practice this means a shot framed on Polaris tolerates far longer exposures than one framed on Orion. If you are shooting the Milky Way core, which sits low and near the equator from most northern latitudes, assume the worst case and use the unadjusted number.

When the limit is too short

If NPF hands you a five-second ceiling and five seconds does not gather enough light, you have four ways out, roughly in order of cost:

Open up. A stop of aperture is a stop of light and it barely changes the NPF result. This is why fast wide primes dominate astro work.
Raise ISO. Modern sensors are close to ISO-invariant in the range that matters, so the noise cost of ISO 6400 over ISO 1600 is smaller than it used to be.
Stack frames. Twenty short exposures averaged together give you roughly the noise performance of one much longer exposure, with none of the trailing. This is the standard approach and it costs nothing but time and processing.
Use a tracker. A star tracker rotates the camera to match the sky, which removes the exposure limit entirely for the stars — at the cost of blurring anything on the ground, so foregrounds get shot separately and blended.

Going the other way: deliberate star trails

Everything above is about avoiding trails. If trails are the point, the arithmetic inverts: you want total exposure time measured in tens of minutes to hours, and the arc length is simply 15 degrees per hour of rotation around the celestial pole.

Almost nobody shoots a two-hour single exposure any more. A long single frame accumulates thermal noise, risks one aircraft ruining everything, and gives you no way to adjust later. The modern method is a sequence of short frames — typically 20 to 30 seconds each, shot back to back for an hour or more — then stacked with a lighten blend so the trails join up. That also gives you a star-lapse video from the same data set for free.

Two practical notes: set the interval as short as your camera allows, because gaps between frames become visible dashes in the trails; and disable long-exposure noise reduction, which doubles your effective frame time and puts a gap in every single trail.

Run the numbers

Get your sharp-star exposure limit for a specific lens, aperture and sensor, plan a stacked star-lapse sequence, and see how long the resulting trails will be.

Open the star trail calculator →