ads-medium

High-Resolution Planetary Imaging Guide, Part 4: Imaging Software and Processing

Data processed using “lucky imaging”, plus other enhancements. Credit: Rouzbeh Bidshahri

With the advent of new sensor technology, powerful computers, and advanced software to capture and process data, we can now squeeze the most resolution out of our telescopes by “cheating” the atmospheric seeing. 

We looked at hardware selection and setup in the previous articles in this series. Here we discuss data capture and processing with planetary-specific software.

Lucky Imaging Technique


The actual theoretical resolution of a telescope is a lot higher than we can usually record through the atmosphere, especially with larger reflectors. To achieve the maximum possible resolution, we have to try to reduce the effects of the atmosphere, or at least identify times when it is least degrading to the image.

Not having access to adaptive optics as professional observatories do, the main weapon in our arsenal is “lucky imaging”, a technique that allows us to take advantage of the moments when the atmosphere happens to be relatively still. Atmospheric turbulence varies, so using data only when it is lowest gives us the best possible results.

Raw data of Jupiter on a good night. Not much detail is visible without processing to bring out the details we’ve captured. Credit: Rouzbeh Bidshahri

Computer

The computer is needed for two stages. A laptop might be used for imaging in the field, with a more powerful desktop used for data processing later.

  • To Capture: 

You will need a computer with a USB 3.0 port. Planetary capture is extremely data and speed intensive. I would recommend against using a mechanical hard disk (spinner), which can be too slow. Currently, M.2 NVME drives are the fastest. I use the Samsung Evo 970 or 980 EVO. Nightly imaging session can gather hundreds of gigabytes of data so aim for a 512GB or larger storage device.

  • To Process: 

Planetary image processing involves sifting through tens or hundreds of thousands of frames, selecting the best ones, and stacking them into a single image. This process is very CPU intensive. A slower CPU will work but could take a very long time. Aim for an Intel I7 or equivalent AMD Ryzen and up.

Software

Planetary imaging can be broken up into five main phases, each with its own software.

1-Capture: 

The aim is to use the shortest exposure time possible; this will increase our chances of catching a still moment in the atmosphere as it moves very rapidly. With such short exposures, there isn’t going to be much light or “signal”, but that is fine as we then take thousands of them that can then be added up.

Data capture requires controlling the shutter speed and gain on the fly, while also controlling the motorized focusing mechanism and filter wheel.

For this step, my recommendation is FireCapture, which is designed specifically for high resolution planetary imaging and has many useful features. There are tutorials on the FireCapture website and, best of all, it’s completely free. 

Certain features such as auto align and autoguide are lifesavers, as the planet is bound to wander out of your tiny field of view. FireCapture also labels files in Winjupos format, a very important feature that is discussed below.

Sharpcap can be used as well, though even the paid Pro version isn’t as powerful as FireCapture for planetary data capture.

FireCapture live data capture screen. Credit: Rouzbeh Bidshahri

2-Stacking:

Stacking is the process of adding images together. The planet image remains constant (signal), so the signal adds up as each image is added (stacked). Meanwhile, the noise (like the camera’s read noise) diminishes because sensors exhibit noise randomly, so it is averaged out with each iteration (each frame). This way is reduced dramatically.

We also pick only the sharpest images with the least amount of blur caused by turbulence. Using only the frames when the seeing was best is what we refer to as lucky imaging. The best software for this is Autostakkert!, which is fairly simple software that is also free.

To stack images, each frame is graded for quality, with only those meeting a user-defined criterion. For example, the top 40% sharpest images may be chosen, and those are then added (stacked) into a single final image. 

On nights with poor seeing, you may decide only the top 25% or so of the frames are good enough to use. On nights with good seeing you may be able to push that to 50%.

Frames analyzed and sorted for quality, with the best 25% then stacked into a single image. Credit Rouzbeh Bidshahri

Picking only the sharpest frames will result in a sharp, contrasty image but with poor signal to noise ratio. Allowing too many frames, including ones with lower quality and contrast, will increase the signal to noise ratio but will produce a soft, blurred image.

The art of processing is knowing where the fine balance is. With practice you will master this skill and your stacked images will look a lot cleaner than any single frame.

Single frame compared to a stacked image of Jupiter from the same imaging session. Credit Rouzbeh Bidshahri

3-Sharpening:

The stacked image can then be dramatically sharpened, revealing incredible detail. This is an essential step in bringing out the detail you’ve captured.

There are few options here, the most popular one is the free software Registax, which uses wavelet sharpening techniques. I prefer Astra Image, which is paid software ($42) that can also deconvolute images in batches.

The same stacked image as above after sharpening. Credit Rouzbeh Bidshahri

4-Derotation: 

Unlike static deep sky objects, planets rotate. You can see the rotation of Jupiter in just a few minutes. We need to rotate all the images to a single reference point (de-rotate) by shifting the pixels and then stack them.

Winjupos is unrivaled for this. Used by professionals as well, it too is free but can be quite complicated to use. Derotation with Winjupos is optional for color cameras, but you will get much better results with it.

With mono cameras, the individual RGB frames require Winjupos to stack a composite color image because the images through each filter are taken serially, and thus at slightly different times. The larger number of stacked images will further increase the signal to noise ratio, allowing for even more sharpening and deconvolution to be done as well.

Left:  Stacked and sharpened image. Right: Derotated images of several stacks. Credit Rouzbeh Bidshahri

5-Post processing: 

Post processing includes things like touch ups, color adjustment, saturation, levels, etc. on the image that results from the processes above. There are many options available for this more generic type of processing, such as Photoshop, GIMP(free), and so on.

Images resulting from steps in the process, from data capture to the final image. Credit: Rouzbeh Bidshahri

Data Capture

Unlike deep sky objects, planets are very bright, allowing very short exposures in the millisecond range. Planetary data capture is thus a numbers game – the more frames you capture, the higher the probability of finding sharp ones with minimal atmospheric blur (being “lucky”).

So, in planetary imaging, we capture short bursts of what are essentially video clips. Each clip is limited to a minute or two, though, as the planet’s rotation will start to blur the image.

Here are some tips for data capture:

  • Aim to increase the frame rate.
  • Don’t be afraid to increase the gain to 60% or even a bit more. The raw data might look terrible but that is normal.
  • I use 60 seconds per clip for Jupiter and 90 seconds for Saturn. Smaller telescopes can go a bit longer because they have less resolution and will not resolve or capture the rotational movement until the planet has rotated farther.
  • My typical shutter speeds are about 10ms for Jupiter, 25ms for Saturn, and 5ms for the Moon.
  • Make sure your histogram is not clipped. Try to not exceed 2/3 of the range. Dimmer Saturn will usually be below 50% of maximum.
  • Crop the image to eliminate the unused background data. This will save disk space and increase your frame rate.
  • Try to avoid USB hubs, and long or low-quality USB 3.0 cables.
  • Remember that you are limited in time. You only have about a 30-minute window to capture all your data. Beyond that you will see artifacts from the derotation process.

Here is a sample of one of my captures of Jupiter with the mono ASI290 and a C14 with a 2x Barlow:

Duration = 60.001s

Frames captured = 8156

File type = SER

ROI = 744x740FPS (avg.)=135

Shutter = 7.256ms

Gain = 385 (64%)

Histogram = 51%

Sensor temperature = 40.0°C

The fast rotation of Jupiter evident over a few hours. Credit: Rouzbeh Bidshahri

In Summary

To discuss everything related to planetary work in detail would require several volumes of books, but this broad, practical outline of what is required should get you started.

Planetary imaging may seem simple, and it can be, but when attempting high resolution imaging at 10 times the scale of deep sky imaging, you need a combination of optimized equipment, patience, intuition, and, most of all, skill.

With practice and meticulous attention to detail, high-resolution planetary imaging can be very rewarding as your skill level improves. It can be really exciting to be able to capture the Cassini Division in Saturn’s rings or details inside Jupiter’s Great Red Spot, all from your backyard. I still remember the first time Saturn’s polar hexagon showed up on my screen. The journey was well worth it for me!

The author’s Imaging technique improvements over time. Credit: Rouzbeh Bidshahri

 

More from the High-Resolution Planetary Imaging Series

Part 1: Telescope and Mount Selection

Part 2: Designing the Imaging Train

Part 3: Imaging Session Preparation

 

Useful Links:

Author’s gallery with more example: https://www.astrobin.com/users/Rouzbeh/

FireCapture: http://www.firecapture.de/

Autostakkert: https://www.autostakkert.com/

Winjupos: http://jupos.org/gh/download.htm

Registax: https://www.astronomie.be/registax/

GIMP (Post processing): https://www.gimp.org/

 

About Rouzbeh Bidshahri

Rouzbeh Bidshahri is a mechanical engineer with a lifelong passion for astrophotography. He has tested dozens of telescopes ranging from 3 to 20 inches in aperture and has spent several years optimizing systems for very high-resolution planetary imaging in the sub 0.1 arcsecond/pixel range. He has contributed to several institutions such as ALPO (The Association of Lunar and Planetary Observers). His main area of interest has been designing and operating larger setups, and he is currently focusing on high resolution, long exposure photography for both broadband and narrowband deep sky imaging.

Related posts