PixInsight a revelation

[jkmorse]

Pixinsight World

Hello from a recent convert to PixInsight.  I have been imaging for several years but was stuck in the CCDStack/Photoshop routine which always felt so unintuituve and clumsy.  The tools, while usage, felt like trying to carve a block of stone with a pick and shedgehammer wearing ski mittens.  You can do it, but the frustrations are paramount (which makes me marvel all the more when looking at the work of true Photoshop aficionados).  And it's all hit or miss, no science, at least that I could really control (chainsaws to do ice sculptures is another metaphor that jumps to mind).

And then I read an article mentioning PixInsight.  I looked at it early on but it seemed so complicated for someone just getting their feet wet that I shied away.  But now (with the extremely helpful work of Harry's tutorials as a starting point, huge cudos there!!) it is like walking into a giant toystore for astroimagers!!  There is so much designed just for us and, while often complicated (sometimes VERY complicated) the processes just make sense.  And the experimentation that is possible when tweaking an image has, really for the first time, helped me realize all the fun of the processing side of imaging.  I happen to be working in the Middle East (imaging from a mountain in Oman on my weekends) and since I am on single status (my wife had enough of foreign assignments after our first 10 years stint) I have lots of time on my hands.  I literally spent the entire past weekend redoing several old images just to get a solid grounding in what Pixinsight has to offer.  I have a couple of new image sets from early September that I haven't even touched yet because I was shooting new Darks and Bias frames to work exclusively in the Pixinsight environment.

Anyway, I have blathered on long enough, but do have a couple of questions.  First, the mask building routine in ACDNR is so easy, is it available as a stand alone process to create masks for use with other processes?

Second, is there an easy way to crop images to a specific size?  I have my office walls lined with US sized 8.5 x 11 inch prints.  Cropping to size is easy in Photoshop but I am hoping there is an easy way to do that in Pixinsight as well.

Anyway, thansk again for an awesome product.

Clears Skies. . .

Jim Morse 

[Juan Conejero]

Hi Jim,

Welcome to PixInsight Forum and thank you so much for your words. Reading the opinions of new users who understand PixInsight's philosophy and benefits is a refreshing experience.

First, the mask building routine in ACDNR is so easy, is it available as a stand alone process to create masks for use with other processes?

Not currently, but it will be available soon. A BrightnessMask tool (the final name may change) is in our to-do list with high priority. It is really a very simple and easy-to-write, yet extremely useful tool. We normally generate these masks with the HistogramTransformation, CurvesTransformation and Convolution tools (sometimes also with ACDNR), but automating the initial steps would be great.

... is there an easy way to crop images to a specific size?

Sure. Use the DynamicCrop and Resample tools. Normally, one first uses DynamicCrop to define the part of the image to be printed. Note that if you don't want to rotate your crop, you can also define a preview and extract it as a new image (just drag the preview's view selector to the workspace, or use Preview > Make Image) instead of DynamicCrop.

Let's assume that you want to resize your image to fit into an 8.5' x 11' sheet, printed at 300 dpi resolution. Once you have cropped the image, these are the next steps:

1. Open the Resample tool (Process Explorer, Geometry category).

2. Select your image from the ListView control at the top of the tool. Alternatively, select your image and check the "Track View" button at the bottom of the tool's window.

3. Open the Resolution section and enter Horizontal = Vertical = 300. Make sure that the selected resolution units are inches, and check the Force Resolution check box.

4. At this point, on the Dimensions section you can see the final print size in pixels, centimeters and inches. If required, change the width or the height in inches, as appropriate, to cover the 8.5'x11' page. Press Enter when you change a number, or click on a different input field, to commit changes in the tool. Note that if the Preserve aspect ratio option is enabled, the tool calculates the resulting dimensions automatically when you modify either the width or the height, to keep the aspect ratio of the image unchanged.

5. The Resample tool will select an interpolation algorithm automatically, based on the amount of resampling required, and on whether the image has to be resampled down or up. The default choices usually work very well, but you may want to experiment. See this document for more information on the subject:

http://pixinsight.com/doc/docs/InterpolationAlgorithms/InterpolationAlgorithms.html

6. Apply the process to your image (blue triangle, etc.).

This process can also be performed in two steps: first crop the image vertically or horizontally, as best fits your needs. Then use Resample to achieve the desired width or height. If the other dimension (the one that you didn't forced with Resample) is too large, crop the resampled image with the Crop tool.

[jkmorse]

Thanks for the quick and helpful response.  I will play with the resampling tool tonight.  As to waiting on the new Brightness Mask tool, well, just like Christmas, some things are worth waiting for.  Always a good day when I fire up PixInsight and it starts by loading a new routine. 

[jkmorse]

Juan,

Another quick way to produce a Brightness Mask in PixInsight currently is to create an image clone, then simply do an Invert and adjust the inverted clone with HistogramTransform and apply the mask to the image when happy.  You can also blur through Convolution if required

At least that what I am using and it seems to work.

Regards,

Jim