DIY Star Charts

[Ags]

I am focussing my visual observing on double stars and needed a book about double stars to help me. I have got form for buying astro books and leaving them to collect dust, so I decided to write the book myself - I guess I should learn more that way. And at least I can put my skills as a technical writer to the service of astronomy and not in the service of soulless Moloch.

As a background, I am using the Henry Draper catalog with stellar positions updated to epoch 2000 by the Vizier service of CDS. I am focusing on stars referenced by these online resources:

http://users.compaqnet.be/doublestars/

http://www.jouscout.com/astro/belmont/belmontd.htm

https://www.stelledoppie.it/index2.php?section=1

And then various resources to research the stars.

The books is assembled using some Java scripting to produce XML Formatting Objects and SVG diagrams, and then using Apache FOP to produce the PDF file. I am printing the book through Lulu  as a ring-bound A4 color book. I chose Lulu because I know from other projects their print quality is better than most and they offer ring binding, which I really wanted.

The book is organized around two kinds of diagrams - 50 degree diagrams showing bright stars visible from a city, and a zoomed in 10 degree diagram showing the 5 degree circle of a finder scope. The wider diagram is for rough navigation with a red dot finder. In the diagrams I map spectral classes of stars to colors. The diagrams have a deep grey background for field use.

Here is the finder view of 55 Psc in Stellarium:

And here is my draft generated diagram:

 

Edited May 6, 2021 by Ags

[wookie1965]

Brilliant will be following with interest. 

[Ags]

Started adding in constellation outlines. Not sure about the green color! 

Need to label the double stars and add page references too.

[wookie1965]

I like the green its how its going to print out. 

[Ags]

Yes, my only concern with green (aside from aesthetics) is that the line might disappear under a red light.

[Spile]

This screenshot gives you an idea of how SkEye presents double stars loaded as a custom catalog. You can change the opacity and size of the labels but the rating (5* being the 'best') is what I use to identify suitable targets in an area of the sky.

 

[Ags]

@Spile The 500 list is really useful! May I borrow from it to fill out the regions where my maps are a bit bare?

I am trying to keep to less than 200 stars as I can do two stars per page and color printing isn't the cheapest. Also, I want to keep the book relatively light.

Edited May 10, 2021 by Ags

[Spile]

16 hours ago, Ags said:

@Spile The 500 list is really useful! May I borrow from it to fill out the regions where my maps are a bit bare?

I am trying to keep to less than 200 stars as I can do two stars per page and color printing isn't the cheapest. Also, I want to keep the book relatively light.

Of course @Ags  All I did was combine two sources (both cited), put them into a spreadsheet with a few additional fields. Like you, I needed just the right number of targets and 500 was right for me.

[Ags]

Added grid lines and constellation labels. Changed the green to have a lot more red in it for night use.

[wookie1965]

Looks brilliant 

[Ags]

I have been struggling with labels. Finding a way to include the names of prominent stars and doubles on the same small diagram is quite tricky! My latest ploy is to move the double labels to the left.

[wookie1965]

Yes brilliant idea.

[Len1257]

Any reason you don't use red for your constellation outlines?

[Ags]

Mostly aesthetics... But note the green lines are actually almost as red as they are green, it's just our eyes are biased to seeing green. I think red lines look jarring:

But final colors are dependent on red light torch testing 😀

Edited May 16, 2021 by Ags

[Martin Meredith]

I really appreciate what you're doing here. Label placement is a tricky problem and quite a lot of the code (Matlab, ugh!) I wrote to generate my charts was devoted to it. I ended up computing bounding rectangles of all labels and defining a set of possible locations for each label (steps around a circle mainly, but for constellation names etc the bounds were quite flexible). I then used an iterative approach (simulated annealing) to find the label placements with the least overall overlap (weighted by importance). Often it would take 1000s of iterations but of course that compute cost doesn't really matter for printed final copy (or pdfs). 

Good luck!

Martin

[Ags]

Labels can be a rabbit hole, but I don't want to get drawn in. My focus is on simple diagrams that give an overview, so minimising labels and structuring the diagram are the priorities, rather than trying to include a higher number of labels. I hope I don't have to do too much custom logic for the labels!

[Ags]

I am representing position angles as radiating yellow lines. The angle of the widest separation is the longest line, the angle of the second widest separation is the second longest line, and so on.

Edited May 18, 2021 by Ags

[Ags]

I got a draft copy of the book today and tested it tonight under the stars. I need to print on less glossy paper and increase contrast generally.

[Ags]

I'd appreciate feedback on which diagram feels better- left or right? I suspect the white background might work better with a red light torch.

[Swoop1]

I think white background is far better for viewing under a red torch.

I'm a bit late to this party but, could you not give each double on a page a number and have a table in the margin to give the identity?

Brilliant project by the way!

[wulfrun]

You won't see the yellow, orange and red ones very well on a white background, using a red torch. You also may find the blue ones hard on a black background. Can you not use (say) all blue ones on white?

Edited May 23, 2021 by wulfrun
added yellow

[Ags]

Faking a red torch with gimp... All the star colors are mixed with grey so they don't disappear under a red torch (they do fade a little). @wulfrun I don't want to make all the stars blue - the current colors indicate the rough spectral class.. While I like the idea of reflecting the spectral class, for fainter stars it may make more sense to opt for high contrast colors, namely the inverse of the background. From the below image I think the white background is likely to perform better. Another consideration is that Lulu Standard Color print does not cope with large areas of flat color, so that points to white  background too.

Edited May 23, 2021 by Ags

[Ags]

50 minutes ago, Swoop1 said:

Brilliant project by the way!

Thanks. This is what a page looks like at the moment (need to add a border around the two diagrams to visually separate them):

[Ags]

Page layout enhancements continued...

 

[Ags]

I have been adding in the essential book features that my book generator wasn't handling up till now.

Table of contents:

And the index:

And overflow plates for the more crowded regions of the sky (a second plate is generated when there are too many doubles on the chart):

And brutal label pruning (overlapping labels not generated, as for the middle star in Orion's belt):

I'm losing sleep over whether labels go over or under the stars...