I have been processing images for years now, and have explored various tools for image processing including JS9 & JS9-4L, FITS Liberator, SAOImage DS9, and MaxIM DL.

I recently decided to try PixInsight to find out if I could improve my images.

The images are derived from 10 Slooh.com missions on T2 totaling approximately 30 minutes exposure time on Messier 51, taken over a number of months.

Result

This is the image processed using PixInsight:

From the image you can see good detail in the galaxy, the center is not saturated, and the nebulous regions are visible around the galaxy. The image is a little noisy and has some artifacts. I worked on this image for about an hour.

This is the image I produced using MaxIM DL and Photoshop:

This image has less detail in the galaxy, however, the center is not saturated, there’s good color information, however, the nebulous regions around the galaxy are not visible. I worked on this image for about 20 minutes.

Details

The images speak for themselves in terms of the detail that both MaxIM-DL and PixInsight resolved. While MaxIM-DL was able to recover a lot of color information, it missed out on the fainter details of the galaxy and fainter nebulous regions. The MaxIM-DL version is less noisy as compared to the PixInsight version since it has few artifacts and a dark background.

The histograms tell a different story – the histograms of each of the images follow:

The histogram of the MaxIM-DL image is much more compressed to the left as compared to the PixInsight histogram, indicating that the PixInsight version has more tonal variation. This variation is apparent in the nebulous regions around the galaxy as well as within the structure of the galaxy itself.

In addition, the histogram for the PixInsight version is longer and trails off slower to white as compared to the MaxIM-DL version.

One of the main differences between the MaxIM-DL version and PixInsight versions is the amount of time spent working on the image.  I worked on the MaxIM-DL version of the image for a total of about 20 minutes whereas I worked on the PixInsight version for about an hour – three times longer.

Process Details – MaxIM-DL

The MaxIM-DL version of the image was created by:

  • Stacking and aligning
  • Adjust screen stretch
  • Digital Development
  • Adjust screen stretch
  • Save As TIFF

From this point I used Photoshop:

  • Levels adjustment
  • Reduce noise using Gaussian blur
  • Curves adjustment to make the background darker and improve the appearance of the galaxy
  • Various adjustments and processes to improve contrast and reduce noise

Process Details – PixInsight

Having experimented with PixInsight, I decided on the following workflow:

  • StarAlignment
  • ImageIntegration
  • ChannelCombination
  • AutomaticBackgroundExtractor
  • DynamicBackgroundExtraction
  • ColorCalibration
  • ATrousWaveletTransform
  • AdaptiveStretch
  • HistogramTransformation
  • LocalHistogramEqualization
  • MultiScaleMedianTransform
  • Adaptive Contrast-Driven Noise Reduction (ACDNR)
  • TGVDenoise
  • HDRMultiScaleTransform

The image was linear until the HistogramTransformation process. I imported the image into Photoshop to resize it for this article, so I exported it as a 16-bit TIFF and resized it.

The Final Result

So, the question is, which image is better? The answer depends on what you’re looking for:

  • In terms of processing time, the clear winner is MaxIM DL because it took only 20 minutes to produce a reasonable looking result.
  • In terms of overall image quality, PixInsight wins because there’s much more tonal variation in the image even though there are artifacts and noise present.

Conclusion

In this article, I demonstrated image processing the same set of data both in MaxIM-DL and PixInsight. Each process provides its own benefit and which you choose depends on what you’re trying to achieve in terms of processing time and image quality.