top of page
LogoMainScreen.png

From Darks to Flats: Essential Calibration frames in astrophotography

  • Writer: Thomas McCrorie
    Thomas McCrorie
  • Jun 13
  • 4 min read

Updated: Jun 24

Understanding your calibration frames in astrophotography

When you first get into astrophotography, it can feel like half the battle is just figuring out what gear you need, how to focus your telescope, or which deep sky object to go after. But as you gain in experience, you take your first stack of images and realise: something’s off. Dust spots, uneven light, weird little glowing pixels scattered about. That’s when you hear the phrase "calibration frames" and wonder if this hobby is ever going to get simpler. I put them off for so long and I kind of got away with it... to an extent.


Calibration frames are the unsung heroes of astro imaging. They don’t look pretty. In fact, they’re mostly black, grey, or just plain boring. But what they do is often magical. They help clean up your final images by correcting for imperfections in your camera sensor and optical path. Think of them as the deep clean your data deserves before you start stretching and tweaking.


What Are Calibration Frames?

Calibration frames are additional sets of images taken specifically to remove or correct for known defects or inconsistencies in your imaging setup. There are three primary types:


Dark frames

Flat frames

Bias frames

Dark Flats


Each calibration frame plays a specific role, and together they dramatically improve the quality of your final stacked image.


Let’s break them down.


Dark Frames: Battling Sensor Noise

Every camera sensor produces heat, and with heat comes noise. Long exposures, like the ones we use in astrophotography, can introduce a range of noise types — especially thermal noise and hot pixels. That’s where dark frames come in.


A dark frame is simply a photo taken with the same exposure time, ISO/gain, and temperature as your light frames (the actual images of the night sky), but with the lens cap or dust cover on. No light gets in — you’re just capturing the noise.


Dark frames allow you to identify and subtract this unwanted data. Once stacked into a "master dark," they’re applied to your light frames to eliminate:


Hot pixels (those annoying tiny bright dots)

Thermal noise patterns

Amp glow (depending on your camera)


Most astrophotographers build up a library of darks for different exposure times and reuse them — particularly if you have a cooled camera that keeps temperature consistent. Some people recommedn redoing your darks every 6 months or so, but again this might depend on your camera conditions etc.


Flat Frames: Correcting Uneven Illumination


Flats are taken to correct for vignetting (those darker corners), dust motes on your sensor or filters, and any uneven illumination across the frame. You’ve probably seen the effects — those weird dark donuts or subtle light fall-off that seems impossible to remove in post-processing.


To take a flat, you aim your setup at a uniformly illuminated surface. Common techniques include:


Using a white T-shirt stretched over the telescope during twilight

Pointing at the dawn sky

Using an LED light panel or dedicated flat field generator


The key things:

Keep the camera and telescope in the exact same setup as your light frames

Don’t change focus or image train at all

Keep exposure short but avoid clipping (histogram peak usually around 1/3 to 1/2)


When you stack these flats into a master flat, you can correct:

Vignetting

Dust bunnies or shadows

Uneven light due to reducers or filters


It’s one of the easiest ways to boost your image’s quality — and yet one of the most overlooked by beginners.


Bias Frames: Capturing Read Noise

Bias frames record the read noise — the electronic noise introduced when the sensor reads data. These are the fastest exposures your camera can take (often 1/8000s or faster), with the cap on.


While bias frames are less talked about these days — especially with CMOS sensors — they’re still sometimes useful, especially if you’re using software that relies on them or working with scaled darks.


If you stack a set of these into a master bias, it can help:

Correct for fixed-pattern noise in some workflows

Improve calibration with certain stacking tools

That said, many modern workflows can skip bias if you’re already using well-matched darks.


Where Do They Go in the Workflow?

Most stacking software (PixInsight, DeepSkyStacker, AstroPixelProcessor) will have a calibration stage. You’ll input your light frames along with your master dark, master flat, and optionally, master bias.


The software applies these corrections before it does any alignment or integration. It’s essentially cleaning your data before doing any heavy lifting.


A typical calibration flow might look like this:


Light frames — your actual night sky images


Dark calibration — remove hot pixels and thermal noise, although programmes like Pixinsihg tare very good at removing hot pixels. below is a typical dark frame from a cooled monochrome camera like the 533MMPro.


Flat correction — even out brightness and remove dust


Bias (optional) — fine-tune sensor readout corrections


Registration — align the stars across frames, if using Pixinsight the software does this for you if you are using WBPP for example.


Stacking — combine the data to reduce noise and improve detail


The Case for Skipping None of Them

You might hear people say, "I didn’t bother with flats" or "I’ll skip darks because it’s a cooled camera." But even with modern astronomy camera tech, calibration frames matter. They are the potential difference between an image that looks clean and polished, and one that has distracting artefacts.


Skipping calibration is a bit like building a house and not bothering to level the foundation.


Tips for Better Calibration Frames

Take enough frames: A good rule of thumb is 20–30 of each filter type. The more you stack, the cleaner the master calibration frame will be. I'm not sure if there is a cut off point but I don't imagine that more the more calibration frames will help your final image.


Keep them separate: Store calibration libraries in a well-labelled folder structure so you can reuse them.


A really important point to note:

Matching settings: For darks and bias, settings must match your lights exactly. For flats, just keep the same focus and optical train.


Hope this helps someone just starting out. I used to cut corners but I now consistently use calibration frames.

 
 
 

Comments


bottom of page