Astrophotography Sub-Exposure Calculator
Use a measured test frame or a known electron rate to size sub-exposures that keep camera read noise small without ignoring tracking and full-well limits.
Recommended sub-exposure54.88
- Measured background rate
- 0.4
- Read-noise swamp factor
- 9.76
- Minimum sky-limited exposure
- 54.88
- Background-limited maximum exposure
- 12,500
- Expected background level
- 527.44
- Actual read-noise penalty
- 5%
- Light frames required
- 263
- Exposure-window status
- Sky-limited exposure fits below the background and tracking limits.
Exposure times are seconds, background levels are ADU, and rates are electrons per pixel per second.
The recommendation is a noise-floor target, not a guarantee that bright stars will remain unsaturated.
How to use this calculator
- Choose measured-frame mode unless you already know the detected sky-background rate.
- Enter read noise, conversion gain, and the matching offset for the gain and binning you will use.
- In measured mode, enter the exposure and median background from an unstretched linear light frame.
- Set the full-well headroom, practical exposure cap, and total integration target.
- Check the status row to see whether the sky-limited minimum fits inside the exposure window.
How the sub-exposure is calculated
This calculator starts with the background electrons collected by one pixel. In measured mode it subtracts the camera offset from a star-free background median, converts ADU to electrons with the conversion gain, and divides by the test exposure. In known-rate mode it adds the sky-background rate and dark current directly.
The core equations are:
R = (A - O) × g ÷ t0 for measured frames, or R = S + D for a known rate.
C = 1 ÷ ((1 + p)^2 - 1)
t_min = C × n^2 ÷ R
t_well = f × W ÷ R
t_rec = min(t_min, min(t_well, L)), except that the result is flagged when a cap forces the exposure below t_min.
- R is total background rate in electrons per pixel per second.
- n is camera read noise in electrons RMS.
- p is the allowed read-noise penalty as a fraction.
- W is full-well capacity and f is the background fraction allowed to use it.
- L is the practical tracking, guiding, wind, or risk limit.
What moves the recommendation
Read noise and background rate dominate the minimum. A lower-noise camera, brighter sky, wider filter, faster optical system, or larger pixels usually shortens the needed sub-exposure. Narrowband filters, dark skies, slow optics, and very low dark current can push the sky-limited exposure far longer. The allowed penalty also matters: 5% corresponds to a swamp factor of about 9.76, while 10% corresponds to about 4.76.
The maximum side of the window is separate. Full-well headroom limits how much smooth background you allow before leaving too little room for stars and target signal. The practical cap represents mount tracking, guiding quality, wind, satellites, and the cost of losing a single long frame. When either cap is shorter than the sky-limited minimum, the calculator recommends the cap and shows the actual remaining read-noise penalty.
What the calculator leaves out
The result is a background-noise target, not a complete exposure decision. It does not predict saturated stars, clipped target cores, amp glow, calibration defects, focus drift, changing transparency, meridian flips, dithering overhead, or the value of rejecting bad frames. Check a real sub for star saturation and histogram placement before committing a night to the number.
Worked example
Suppose a 30 second linear test frame has a background median of 515 ADU and a matching offset of 500 ADU. With 0.8 e-/ADU gain, the background rate is (515 - 500) × 0.8 ÷ 30 = 0.4 e-/pixel/s.
At 1.5 e- read noise and a 5% allowed penalty, the swamp factor is 9.756. The minimum sky-limited exposure is 9.756 × 1.5^2 ÷ 0.4 = 54.88 seconds. A 20,000 e- full well with 25% background usage allows 12,500 seconds, and the practical cap is 300 seconds, so the recommended sub-exposure is 54.88 seconds.
The expected background is about 527.44 ADU, the actual read-noise penalty is 5%, and a 4 hour integration needs 263 light frames.
Common questions
What does it mean to swamp camera read noise?
It means collecting enough background signal that shot noise from the sky background is much larger than the camera read noise added to each sub. The read noise is still present, but it contributes only the allowed extra noise percentage you selected.
Why is a measured background better than Bortle class?
Bortle class does not know your filter, focal ratio, pixel size, gain, transparency, moonlight, or local light sources. A measured linear light frame includes all of those effects at once, so it gives a more specific rate for the setup you are actually using.
Should I use mean or median background ADU?
Use a median from a star-free region of the unstretched raw frame. The median is less affected by stars, hot pixels, cosmic rays, and faint target structure than the mean, though strong nebulosity can still make the background estimate too high.
Is a 5% read-noise penalty the same as a 10x swamp factor?
Nearly. The exact factor is 1 divided by the square of the permitted noise increase minus one, so 5% gives 9.756 rather than exactly 10. A 10% penalty gives 4.762, which is less demanding and therefore produces shorter subs.
What should I do when tracking limits prevent a sky-limited exposure?
Use the capped exposure and collect more frames, or reduce the read-noise penalty target only if the extra noise is acceptable. Other practical options are using a higher conversion-gain mode with lower read noise, improving guiding, or accepting that narrowband data under dark skies may remain read-noise constrained.