Book recommendation?

[johnking101]

PixInsight is impressive. I've followed a number of workflows that have been generously shared by PixInsight users, and seen the very helpful tutorials provided by Harry. I feel pretty comfortable navigating the PixInsight application, stretching histograms, creating masks, etc. I really like the flexibility PI gives me to have multiple windows and tools open when I want, and to allow me to work the way *I* want. It's like an IDE for image processing.

Where I find things difficult is in understanding *why* any particular workflow is designed the way it is. E.g. I know Luminance is perceived more sharply than RGB by the human eye, but why would I separate them for processing? How to pick from a set of available sharpening or blurring algorithms? When an experienced user looks at an image, what flaws point them in one direction or another?

Is there any particular book that you'd recommend? Something that covers the general concepts? I was thinking about "Lessons from the Masters" by Robert Gendler, but then I saw a comment from a PixInsight user that the book is too heavily focused on Photoshop.

Thanks.

[JoLo]

Gendler's book is really good, and has several workflows for AP processing.  Yes, some of them are geared toward PS, but there are sections that deal with PI and other software (CCDSoft, Maxim, etc.).  I would not let this deter you from getting the book, it really is excellent.

I know of no book soley dedicted to PI.  What helped me more that anything was Warren Keller's DVDs on PixInsight.  Not heavy on theory but some real nuts and bolts advice, and a good sense of the PI workflow. 

[johnking101]

Thanks Jolo. I'll add that book to my (growing) wishlist. I think I'll hold off on PixInsight-specific DVDs until I've read the general concepts though.

[Juan Conejero]

Unless you want to learn how to create nice hand painted images, I definitely don't recommend that book. There's almost nothing about PixInsight in that book, and actually nothing that you can't learn better from some old tutorials (see for example this one that Vicent Peris wrote in 2005) and other written and video tutorials available online.

With the exception of a few chapters having some interesting information (the first one about theory of astronomical imaging by Stan Moore, Rogelio Bernal's chapter with some examples of multiscale processing, the chapter on high resolution planetary imaging by Damian Peach, and the chapter about images of the solar corona by Fred Espenak), the rest is at best irrelevant for a PixInsight user.

[johnking101]

I haven't seen that tutorial. Even if I did, I might have skipped it, because it refers to a very old version of the software. But it gives a very nice explanation of why the various operations are being performed. Thanks!

[JoLo]

Well, I found the book very helpful for context and flow and there are sections that deal specifically with PixInsight.  I am not devoting my time, money and energy to "create nice hand painted images" and find the information conveyed by these accomplished astroimagers to be much more that such a characterization, even though some do not - dare I say it - use PixInsight exclusively.

I reveiwed many of the tutorials Juan refers to and frankly, some were more useful than others; some were sorely outdated as you pointed out, others were excellent.  I would also disagree regarding Mr. Peach's contribution on planetary imaging.  I found it to be simplistic and very general, nothing he wrote that I did not already know or could not find in more detail with a google search. 

John, I stand by my recommendation on the Gendler book, I and other PI users have found it useful even though it is not all PI all the time.

Joe

[jkmorse]

Joe,

If you are looking to get into the nuts and bolts and the math, I would strongly recommend you look at the Handbook of Astronomical Image Processing by Berry (and another guy who's  name I can never remember).  Its a great resources and goes into great detail on the science and mechanics of noise, wavelets, etc., etc.  Its a little dated but a really solid grounding in the science behind the processes.  Note, however, it does not reference PI but the math translates well.

Best,

Jim

[Juan Conejero]

The problem is not with the fact that PixInsight is or is not used. The main problem with this book is in its general lack of quality and rigor as a text on image processing. With some exceptions, this book provides descriptions of key image processing topics such as deconvolution, dynamic range compression, noise, color and noise reduction---to name some---that can vary from inaccurate to wrong. In addition, arbitrary manual manipulations are presented and justified as perfectly valid practices through step-by-step recipes and the use of hand painting tools. Tools such as magic wand, lasso, marquee and clone stamp are regarded as "advanced tools" offering "tremendous flexibility".

I would also disagree regarding Mr. Peach's contribution on planetary imaging. I found it to be simplistic and very general, nothing he wrote that I did not already know or could not find in more detail with a google search.

I basically agree with your opinion, but at least this chapter provides quite accurate information on some important topics.

[Juan Conejero]

Hi Jim,

It is The Handbook of Astronomical Image Processing, by Richard Berry and James Burnell:

http://www.willbell.com/AIP/Index.htm

I also recommend this book.

[Juan Conejero]

As a good general text on image processing I'd recommend The Image Processing Handbook by John C. Russ. I have the fifth edition (2007) but the sixth edition is available since 2011.

[vicent_peris]

Hi,

Just to make a small contribution, I would say that the description of the HDRMultiscaleTransform is completely wrong. The author never contacted me and he doesn't have an idea on how the algorithm works, giving false information about it. Although I don't speak about the formulation of the algorithm, this small article gives you a way better description of this technique:

http://pixinsight.com/examples/HDRWT/M101/index.html

A similar thing happens with the chapter about multiscale processing. It is based on my own techniques I developed about 10 years ago (some figures are even similar...) but it does not provides any reference. Please read the article on the M33 DSS data Juan posted; although the techniques are outdated, the concepts in that article are still valid IMO. I have to say that my own techniques evolved heavily from this old article.

The final point is that this book has a teaching methodology just opposite to mine. I never teach recipes. I teach people on general methodology and how to think to solve particular problems on every image.


Best regards,
Vicent.

[Geoff]

As a good general text on image processing I'd recommend The Image Processing Handbook by John C. Russ. I have the fifth edition (2007) but the sixth edition is available since 2011.

I spent some time looking at selections from this book using Amazon's "Look Inside" feature.  It certainly looks like a very good book. I had no hesitation in ordering it, even with its $150 price tag.

I also concur with the recommendations for The Handbook of Astronomical Image Processing, by Richard Berry and James Burnell.  It occupies a prominent space on my bookshelf.
Geoff

[pfile]

yes - just to understand how CCDs work and what calibration is all about, HAIP is well worth the money.

rob

[JoLo]

I am not the OP on this thread, and have read Berry's book, most of Vincent's tutorials and papers, and review the information on this forum regularly.  I was merely responding to a specific question regarding a specific book from the OP and, what I think to be, an unfair characterization of what I consider a useful and informative book.  I don't think it ever claimed to be the nuts and bolts of mathematics of image processing that Berry's book clearly is....it did claim to be an agnostic guide from some of today's top imagers on some processes and workflows they use.  For that, I found the book very useful and recommended it to John.

Although I have read the Berry book a couple of times, and reference it frequently, I think there is a pretty good percentage of folks who love to image, but don't have the time and interest to get as deep as he does into the mathematics of it all.  I read those parts, often with glazed over eyes.   I appreciate the work, as well as the more general information presented in a format like Gendler's book does quite well.

Apologies for any misinterpretation or confusion on my part, but there is room in the astroimager's world for A Day In The Live Of Ivan Denisovich and War and Peace...... .  Thanks to the PixInsight team for this great forum.

Joe

[johnking101]

Thank you all for your thoughts. People learn in different ways, and I was hoping to see a few different perspectives. This thread has done that, it gives a number of options for anybody who wants to learn a little more about image processing.