JavaScript support is required to run this website properly.
Please enable it in your browser's settings, or turn off the script blocker.

Documentation / Astrophotography and the HDR image processing techniques


Photography of astronomical objects brings many problems as compared to the photography as most people know. The exposure times can be very long (even minutes) and the lenses, or telescopes used, often characterize with big focal lengths (hundreds to thousands of millimeters), while because of Earth's rotation (and not only), the celestial objects have to be tracked during the exposure. The quality of tracking, the optics, the imaging sensor, the ambient temperature, wind, athmosphere's turbulences, thin clouds, light pollution, etc., that all makes astrophotography difficult.

What is more, there are the same problems as in the "daylight" photography. One of them is the high dynamic range of the photographed objects - comets, nebulae, galaxies, crescent Moon, partial lunar or total solar eclipses. All of these are hard to be photographed in a single shot. Because of long exposure times that are often used, it is worth to take many photos and stack (average) them. That is one of the ways to increase the dynamic range - by making the SNR (signal to noise ratio) higher. Further dynamic range boost requires merging photographs taken at different exposure times. Having an image with elevated dynamic range, you still need to compress it (apply tone mapping) in order to make all the details visible on a screen, or a print.

HDR Moon
A tone mapped HDR photo of a crescent Moon. Visible are both the earthshine and the sunlit lunar surface. This kind of photo cannot be taken without using HDR techniques because of very high luminance ratio. The photo has been processed with easyHDR of a series of 18 photos taken at exposure times ranging from 1/100 sec (at ISO 200) up to 2 sec (at ISO 400). Some of the shortest exposure photos were taken at the same settings to further reduce the noise.

HDR Moon, Lunar HDR Eclipse
HDR photo of a partial lunar eclipse - September 28th 2015, about 15 minutes before totality. Seven photos processed with easyHDR - all taken at ISO 400, exp. times: 1/100 sec, 1/30 sec, 1/15 sec, 0.4 sec, 3 x 1 sec.

One of the examples is the Orion Nebula (M42). It is very difficult (top shelf class equipment is needed) to take a photo with properly exposed core as well as the outer fine detail.

The sample M42 (Orion Nebula) photos, shown below, were taken by Luke Bellani with Meade DSI Pro 2 astrophotography camera attached to Skywatcher ED80 telescope. The photos were taken at 4, 15, 60 and 340 seconds of exposure. EasyHDR directly imported the FITS image format (that is used by the camera to store image data) along with the exposure parameters, which were used to calculate the exposure values, necessary for true HDR radiance map generation.

M42, 4 sec
4 seconds (EV: -3.91)
M42, 15 sec
15 seconds (EV: -2.00)
M42, 60 sec
60 seconds (EV: 0.00)
M42, 300 sec
340 seconds (EV: 2.50)

FITS - generate HDR
The HDR generation window with the loaded M42 sequence (True-HDR method is used)


Below are the results achieved with various HDR generation methods and different FITS contrast correction options.

True-HDR method, selectivity: normal Smart-Merge method
True-HDR method, selectivity: "normal". Smart-Merge method.


If you're aiming for the best realism, you should ensure the following:

High Dynamic Range astrophotography with easyHDR


Another example is the Horsehead nebula (IC434), also in the constellation of Orion. Vincent Steinmetz, the author of the photograph, used a professional astrophotography camera to take twenty 10 minute exposures that were stacked together. The photos were taken under dark skies through a H-alpha narrow-band filter. That kind of filter blocks all light wavelengths, passing just a narrow window around Hydrogen-alpha emission line. The resulting stack, stored in FITS image format was directly loaded into easyHDR and tone mapped. The stack has a very wide dynamic range due to high quality of the single exposures taken by the camera, that was further increased by averaging the noise. There was no need to vary the exposure time with this setup and this particular object. EasyHDR's job was to boost fine details in the hydrogen clouds of the nebula, while not overexposing the brightest details.

Comparison of underexposed, overexposed and a HDR tone mapped with easyHDR - Horsehead nebula.
The stacked, high dynamic range image has been "exposed" twice, by stretching the histogram to preserve the most of the highlights (image to the left) and to bring some details from the darkness (image to the right). In the middle, the result achieved with easyHDR is shown.

Comparison of underexposed, overexposed and a HDR tone mapped with easyHDR - detail close to the Horsehead nebula.
Zoomed in to show that stretching the histogram is not enough. It either causes details in the lights, or darks to be lost. The local contrast tone mapping applied by easyHDR can bring the fine details, while keeping the lights.


Astrophotography, Horsehead nebula before HDR processing Astrophotography, Horsehead nebula after tone mapping

High Dynamic Range astrophotography with easyHDR. HDR Horsehead nebula.
The original photo had a very narrow histogram. To help easyHDR better use it's local contrast algorithm, one pre-processing step was added. The image was first loaded and just the gamma was increased (neutral preset + gamma boost). The result was saved as 16-bit TIFF and re-loaded. Then noise filtering was applied at the LDR enhancement step. Most of the tone mapping presets do clip the histogram a bit (for example 0.2% of brightest pixels are clipped to white), which in case of deep-sky photos may be too much actually. Therefore one of the most important parameters to adjust is lowering the clip margin (in the example above: 0.0318%).


Back to the main documentation page


Current easyHDR version
3.14.1   (March 7th 2020)