NPF rule: Only one formula is needed to shoot the starry sky, and the longest exposure time can be measured with high accuracy.

When we observe the stars in the night sky, we will regard each star as a sharp point of light. So when we shoot, we usually record them as stars. But because of the rotation of the earth, the stars seem to move in the night sky. Often in a long exposure, we will see the trajectories of these stars appear in our images, just like tadpoles.

If you want to shoot a sharp starry sky through long exposure, one of the most common rules of thumb is to divide the focal length by 500 to calculate the longest exposure time available, which is the so-called 500 rule. Sometimes, depending on the size of your sensor, you will also use parameters such as 600 rule or 400 rule. But we found that this is not a very accurate rule.

In fact, the so-called 500 rule was designed for the film photography period, and 35mm film was used under the higher ISO. But the resolution of digital sensor is much higher than that of film. This rule does not consider pixels per inch, aperture and diffraction. The reason why this rule is vigorously promoted is only because it is relatively easy to remember and operate.

In order to greatly improve the shutter speed control accuracy of starry sky photography, through the continuous efforts of some professional photographers in recent years, we finally ushered in a high-precision calculation mode, which is the NPF calculation rule. Although this mode is relatively complicated, it is indeed a good choice for some photographers who pursue high-quality shooting. In this paper, we will deeply understand its principle through some cases and master its application in actual shooting.

In order to get the best or more accurate results with modern digital cameras, you need more accurate exposure time. You must consider the focal length and aperture, as well as the sensor size and resolution.

So, in the NPF rule:

The actual calculation formula is:

Note: p (pixel pitch) = physical width of camera sensor (in millimeters) ÷ pixel width x 1000 in microns (microns).

So, for example, for Nikon D750 camera.

So the pixel pitch p = (35.9÷ 6016) ×1000 micron =5.967≈5.97 micron.

Assuming that a D750 camera uses a 24mm F/2.8 lens, according to NPF rules, the shutter speed is approximately equal to {(35× 2.8)+(30× 5.97)}-24 {98+179}-24 seconds.

If we use the rule of 500 to calculate the exposure time, it is 500÷24=20.8≈2 1 sec. Images printed with this setting or enlarged on a mobile device will show some star tracks instead of point stars.

With the improvement of sensor resolution, this will lead to the increase of pixels per inch, and the appearance of star orbit will be more obvious at the pixel level. Therefore, with the improvement of resolution, it is necessary to reduce the exposure time to eliminate the star orbit.

Here we also need to understand a concept, which is pixel tolerance. This is any number of moving pixels that we choose to tolerate in the image. A pixel tolerance of 7 pixels means that the stars in the picture will move a distance of up to 7 pixels to obtain the recommended shutter speed. A larger pixel tolerance will produce a longer star orbit, while a smaller pixel tolerance will produce a shorter star orbit.

The theory is very good. But the actual shooting is where we really use our knowledge.

This time, I chose to shoot on the sand dunes in the suburbs (the main purpose is to avoid urban light pollution). We also chose Nikon Z6 with large sensor and high pixel, and IRIX 15mm F/2.4 lens.

Rule 500: 500/15mm = 33 seconds.

I shot this with the 500 rule, which gave me a shutter speed of 33 seconds (the actual shooting used a shutter speed of 30 seconds). This forced me to use F/2.4 aperture and 3200 ISO. It looks good in Lightroom's standard view, and there is no problem if it is posted on social media. However, after 100% magnification, there will be obvious tailing, and the pixel tolerance is about 15.

NPF law = 17.26 seconds

Under the NPF law, the given shutter speed is 17.26 seconds. In actual shooting, we use a shutter speed of 20 seconds. Finally, our ISO is 6400, and the aperture is still the maximum aperture F/2.4. At the magnification of 100%, the star is still round, just a little rectangular. Further magnification, I think this result is completely acceptable, and the actual pixel tolerance is around 6.

More advanced calculation provides us with relatively accurate shutter time.

Usually, you can use the "500 Rule" on a full-frame camera, the "400 Rule" on an APS-C camera and the "250 Rule" on an M4/3 camera. Although this will lead to a larger pixel capacity, it will also lead to a longer shutter time and provide a better exposure environment. When you want to minimize the star tail to a very specific number of pixels, the advantages of using NPF rule for calculation are particularly obvious.

For the final choice of different calculation methods, you can establish it according to your acceptable pixel tolerance range in the later shooting experiment. If you have any other thoughts and questions, please leave a message in the comments section below.

This is the "Harvard" in the hearts of giant photographers and photographers. Thank you for your attention.