Image Processing

The telescope, the mount, the acquisition technique, the number of pics you stack and with which method, all that counts and of course counts a lot to define what the final result will be.

But image processing is probably the single most important operation you have to perform in the quest for a nice DSO picture.

Sure, if you are imaging from the Canary Island or the Andes, with a big bore scope mounted on an equatorial mount allowing you to take single exposure of 15 minutes or more, you may even limit yourself to just some histogram stretching and to fiddle a little bit with curves.

But if you are working with small (4 inches, 102 mm maximum aperture) telescopes on an AltAz mount allowing you exposures up to a maximum of 60s and under an urban sky, you darn need a lot of image processing if you want to get something at least acceptable.

I have never been particularly comfortable with the “traditional” path to AP image enhancement, be it through Photoshop or Maxim or Astroart (i.e. histogram stretching, curves, levels, masks or kernel filters, etc.).

Not that I did not try, indeed both in my “first AP life” (back in 2006-2007) and for a good part of the first year of this second one, I just tried to follow what you may find in an endless number of tutorials.

Not that I got nothing, but what I got never let me particularly satisfied: I always thought that somehow there was “more” in those images that was just waiting to be “pulled out” in some way.

I first read about StarTools in one of the many book and articles dedicated to AP I read while I was beginning to rekindle my old interest for stargazing (or indeed “starshooting”), I could not even say which one specifically now.

StarTools

The first time I had a look at their site I was rather baffled: that software looked like no other I had worked with (or even just read of), since it seemed to take a totally different approach to DSO image processing, without the ubiquitous histograms or curves and so on.

I was also fascinated by the incredibly small size of the program (at the time just 1.4 MB: a floppy disk from the 80s!), and it was also rather cheap, but I simply could not figure out how it worked, so much its workflow was unlike those of the other softwares.

Then one day, perhaps by just reading more attentively at the (rather concise) tutorial on the StarTools site itself, I had my epiphany (I do no remember exactly how), understood how it worked and from that day on I have used only StarTools for DSO image processing (moon and planetary imaging is better dealt with with other programs).

My workflow while processing DSO images with StarTools essentially follows the basic one outlined on the StarTools site , with some variation depending on the type of object (galaxy or nebula rather than star clusters, for instance) imaged.

Particularly, the approach substitutive of ordinary stretching provided by the “Autodev” module is something quite unlike all other programs and indeed may even surprise or “upset” the first time it is used because of the “extreme” equivalent stretching it provides.

But it certainly allows to “pull out of noise” a faint nebulosity better than the “Develop” module, which is more akin to a standard Digital Development Process, albeit sometimes I opt for the classic procedure because Autodev may give results which may appear unnatural.

The most vaunted feature of StarTools is, however, Tracking, which basically allows you to go back in time in your image processing flow, reverting the effect of basically anything you have done till then.

Wavelet denoising is also top notch, while wavelet sharpening (“Sharp”) and light pollution removal (“Wipe”) are good, but the latter function is in my opinion somewhat less efficient than its equivalent in APP (Astro Pixel Processor).

There is also an excellent Lens module to correct lens distortion and do field flattening: particularly useful to those who, like me, use lenses and scopes that are a far cry from being “well corrected to the outer border”.

So far I haven’t obtained great result with StarTools’ “Decon” module (for deconvolution): while it helps at times, I reckon it gives rather “mild” results when compared to, say, Iris, albeit “panda-eye” effect prevention looks better in StarTools.

The Life module is a real bonanza and its sub-modules “Less-is-More” and “Isolate” are, in particular, the two I appreciate the most, for their ability to “improve” the separation between the object you are interested in and the background.

The “Bin” module performs a software binning not limited to integer values. I don’t want to enter in the controversy about software vs hardware binning efficiency, but in my experience it really helps to improve an image, trading resolution for noise reduction.

I also find extremely useful and welcome the “Repair” module: it basically repairs your elongated stars making them round again, something particularly useful when your tracking is not optimal and that is something that may happen rather frequently when you are trying to squeeze out the longest possible exposure from your AltAz mount.

While purists (and snobs) may object to the use of “retouching” tools like Repair, I surmise everybody who wants to take a relatively cheap approach to astrophotography will praise them.

As regards purism and different approaches to image processing, but indeed to astrophotography as a whole, I have found this interesting thread on the PixInsight forum, with an exchange of points of view (and of philosophy) between one of the main developers of PixInsight (Juan Conejero) and StarTools’ creator Ivo Jager.

You may find in it a perfect contraposition of two almost opposite philosophies re what an AP image processing program should be, but it is a philosophical difference also rooted in a different vision of astrophotography itself.

A point in particularly directly concerns tools like the Repair module in StarTools:

Conejero:

Solutions of the ‘correct the stars’ kind fall in the painting/retouching category. We are not interested in these. This is astrophotography, not artistic/general photography, and hence retouching tricks are excluded from the game.

[…]

So the only solution to tracking problems is fixing the hardware that is causing them: throw away all wrongly tracked frames, look for the causes of bad tracking in your equipment, analyze and understand them, and fix them. Then acquire the data again. Astrophotography requires the pursuit of excellence, both technically and aesthetically. No software ‘magical box’ can help you to achieve that.

Jager:

this stance effectively shuts out a large number of less well equipped, less tech-savvy or less-affluent users. How do I know this is the case? I was one of them! But instead of putting our wonderful hobby/profession in the ‘too hard, too expensive’ basket and give up, I did something about it.

Cheers, Ivo!

Indeed Jager’s point of view makes me appreciate his software even more.

It is not the first time I perceive a sort of “elitist vs proletarian” confrontation among astrophotographer, specifically DSO imagers.

More then once I got the impression that to some AP is an expensive, exclusive country club where you are admitted only if you have the right stuff, that is the right and sufficiently expensive equipment.

That the “right stuff” is made more of equipment than of skill can be seen when you read in some “milieu” the comments about substantially equivalent images produced by a guy with an expensive equipment and by one with a cheap setup.

More often than not the guy with the expensive equipment gets comforting comments of the “not bad and you can surely do better” type, while the “cheap” guy gets something more along the line of “you cannot do serious astrophotography with that equipment”.

Curiously this approach is almost the reverse of the one I saw in high resolution (i.e. lunar and planetary imaging) circles: often you get pats on the shoulder when you present your first photos with a basic equipment, but sometimes, if you come back sporting an equipment on par with that of the local “kings of the castle”, you are in for acrimonious comments finding defects and/or artifacts everywhere.

I have been on a fair number of forums in the past decade, but I now prefer to take care of my own little garden (this site), mostly because of the behaviors/stances described above.

 

Back to Contents

 

2 thoughts on “Image Processing

  1. I can only agree with you 100%.
    With my Meade 8″ (Alt/Az) and an Atik Infinity (b/w) I have a lot of fun ( and success ) observing galaxies.
    (0.63 Reducer, most time 40×15 (20) sec. only b/w , like in eyepiece )
    With Startools I usually get very nice astrophotos from the Infinity stack and I don’t give a damn if so-called experts think that this is no astrophoto.
    We don’t let our fun at the hobby spoil after the motto “if you are too poor or too stupid to use PI, you can’t take serious astrophotos …”. Neither is true 🙂

    Best regards

    Klaus

    Liked by 1 person

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s