Life-science instruments and observatories often ask a camera to sit open for tens of seconds to many minutes. Chemiluminescence western blots, faint fluorescence and deep-sky photometry all share that problem: the signal arrives slowly, so the exposure has to be long enough to collect it.
Over that time, the camera's own dark current can become the noise floor. This post walks through what a published dark-current figure means in practice, how to take dark frames on a shutterless cooled CMOS camera, and when a CCD is still the better fit. We use Finger Lakes Instrumentation (FLI) Aurora cameras as the worked example. For the preclinical case, see our post on cooled cameras for in vivo bioluminescence imaging.
Why long exposures care about dark current
Dark current is the thermal signal a sensor produces with no light on the detector. It is usually quoted in electrons per pixel per second at a stated sensor temperature. It adds a pedestal that grows with exposure time, and the shot noise on that pedestal grows with the square root of the dark signal.
In a short, bright exposure, read noise and photon shot noise dominate. In a long, dim one, dark current can take over. That is why cooled scientific cameras exist: thermoelectric cooling cuts dark current steeply, so a multi-minute exposure stays useful.
FLI's applications page puts it simply for the Aurora line: the cameras are built for both short and very long exposures. Chemiluminescence imagers, fluorescence instruments and astronomy programs all use that range.
The arithmetic: dark signal vs read noise
Aurora model pages and datasheets list a dark current of 0.002 electrons per pixel per second at −20 °C, with a read noise of 1 electron. Those two numbers let you compare the dark contribution to a single readout:
Using those two numbers:
- 60 s: dark signal 0.12 e− — well below the 1 e− read noise
- 10 min: dark signal 1.2 e− — about equal to one readout
- 1 h: dark signal 7.2 e− — several times the read noise
These figures are simple arithmetic on the published dark-current figure. It is not a finished noise budget. Real images also carry shot noise from the signal, residual pattern and enclosure light leaks. Still, the crossover is clear: at a minute, dark current is almost free; at an hour, it is something you plan for.
Full well capacity on the same pages is 50,000 electrons. That sets the bright end of a single frame. For a long, dim exposure, the low end usually matters more: how little dark and read noise you add while the signal accumulates. Our post on acceptance testing for gain, noise and linearity covers how to measure those terms on a camera you already have.
Cooling is the lever
Dark current falls steeply as the sensor gets colder, so cooling depth is the main control. Aurora and Revelation cameras cool to about 50 °C below ambient with air. Liquid circulation can go further: with liquid near 20 °C, expect roughly 10 °C deeper cooling than with air, as long as the liquid stays above the dew point so the back of the camera does not ice.
Aurora options listed on the model pages include liquid cooling adapters and anti-dew technology. Inside a closed instrument cabinet, removing the heat the cooler dumps is as important as the sensor temperature itself. Our post on air versus liquid cooling for OEM cameras walks through those tradeoffs.
Taking dark frames without a mechanical shutter
Aurora cameras are not offered with a mechanical shutter. Dark frames still matter: you need a matching exposure with no light so you can subtract the offset and dark signal.
The usual approach is a filter wheel with a blocked position. FLI's CFW8-8RND, for example, holds seven 50 mm round filters plus a blank. Park the blank, take the dark, then return to the imaging filter. A light-tight enclosure with the illumination off can serve the same role for instruments that do not use a wheel. See the accessories page for filter-wheel options.
Match the dark's exposure time and sensor temperature to the light frame. A library of darks at a few common exposures is often enough for a production instrument; refresh it when the operating temperature changes.
Where long exposures show up in instruments
A few instrument classes lean on this regime every day:
- Chemiluminescence and blot imagers. The glow from a western blot can be faint enough that exposures run from tens of seconds into minutes. Low dark current keeps the background from drowning weak bands. Our post on replacing CCDs in gel and blot imagers covers the OEM redesign angle.
- Fluorescence and protein analysis. Timed exposures, multi-channel sequences and dim secondary labels all benefit from a camera that stays quiet while the light source is off.
- Astronomy and photometry. Time-series work and deep imaging stretch exposures as far as the sky background allows. Stacking shorter frames is often better than one very long frame, but only if each frame's dark contribution stays small.
For compact instruments, Revelation uses the same Sony sensor family as Aurora in a thin profile suited to OEM housings, and every camera includes our software development kit (SDK). Our post on OEM camera partners for life-science instruments covers how that integration usually starts.
When a CCD still wins
Cooled CMOS is not the only answer for long exposures. FLI's ProLine PL23042 CCD has lower dark current and is easier to calibrate for multi-minute exposures, though it reads out slowly. For programs that live at many minutes per frame and can wait on readout, a ProLine CCD can still be the cleaner choice.
Pick from the exposure length you actually need. If most frames are under a few minutes and you want fast readout between them, cooled CMOS fits. If you routinely sit open for many minutes and calibrate carefully, ask about the CCD option.
Questions to settle before you choose
- What is your longest useful exposure, and is dark current or sky/background the real limit?
- At that exposure, does the published dark current stay below your read noise, or above it?
- How will you take dark frames if the camera has no mechanical shutter?
- Will the camera run in open air, or inside a closed cabinet that needs liquid cooling?
- Do you need a thin OEM form factor, or a benchtop body with room for options?
- Would a slower CCD with still lower dark current fit the workflow better?
Talk to our sales team
FLI cameras are Made in USA, and our sensor chambers are assembled in our 1,000+ sq ft ISO Class 7 clean room. If you are specifying a camera for chemiluminescence, fluorescence or long-exposure astronomy, tell our sales team your longest exposure, target sensor temperature and how you plan to take darks. We will suggest a model and cooling setup that fits. Request a quote to get started.
