Great Wall POSS2 glass plate colorised

[gorann]

I am currently entertaining myself with data from the monochrome (red or blue filter) glass plate images from the Palomar Observatory Sky Survey II (1990ties) by adding RGB data from my own images. I download the data from the Mikulski Archives site (I am sure many of you know about it but I only discovered it yesterday). They used the Oschin 48-inch (1.2-m) Schmidt camera on the Mt Palomar Obsy and I used my humble 5" ES apo refractor and Canon 60Da. Here is the download (red filter image) from the Mikulski Archives site and the final colourised version made in PS. I added the glass plate data to my RGB in PS using both Soft Light and Luminosity mixing, and then stretched, improved contrast, and worked on the colours. Much of the job went into the initial the scaling, rotating and aligning (manually in PS).

Comments and suggestions most welcome.

[wimvb]

Looks great. What does the RGB image look like without the added Palomar plate?

The weather is so bad, I've been looking into doing something similar. But first I want to get my guiding software set up. I miss those cold -15 C nights with crisp skies. Even when stars are visible, high clouds spoil imaging.

Thanks for sharing

[gorann]

7 hours ago, wimvb said:

Looks great. What does the RGB image look like without the added Palomar plate?

The weather is so bad, I've been looking into doing something similar. But first I want to get my guiding software set up. I miss those cold -15 C nights with crisp skies. Even when stars are visible, high clouds spoil imaging.

Thanks for sharing

Tack Wim!
In this case, the RGB data is from a single 10 min exposure that I took when I just started astrophotography in 2015. So it is quite noisy and rather embarrassing to show. On the other hand, it reveals that you do not need much RGB data to colourise good Lum data (and the worst noise in your colour data can be supressed with some gaussian blur before you merge it with the lum). So you could start digging out old data that you put away in the darkest corners of your hard drive, and have a go.

One little problem with merging these images is to match them up. It is almost impossible to make the stars fit on top of each other in all the corners. It may have to do with the projection used for the digitalization of the POSS2 images. I expect that they have been projected by a computer to fit the sphere that makes up the sky (like making maps of the earth).

In any case, here is the RGB image:

[RichLD]

Fantastic job there sir - I'm particularly impressed with your manual frame registration!

Very pleasing final image, you must be well chuffed! 

[gorann]

3 minutes ago, RichLD said:

Fantastic job there sir - I'm particularly impressed with your manual frame registration!

Very pleasing final image, you must be well chuffed! 

Thanks Rich!

Yes, well chuffed I am! (had to look up that word - apparently an informal word in UK English and not very common in Swenglish)

The registration is mostly about being patient. First see to it that both images are on the same dpi (I use 72 dpi). Then use the ruler tool and measure the distance between two distant stars in each of the images. That gives you the difference in scale (dividing one distance with the other) and use this to adjust the image size so they become equal (I expand the smaller of the images not to lose any resolution). Then you just have to start rotating one of them and paste it as a layer with 50 % fill on top of the other to see how well they line up. Delete the layer, rotate again and try again. This you have to do many times in smaller and smaller adjustments and you may at the end also slightly adjust the width and height of one image to get a perfect line up.

 

 

[RichLD]

Apologies for the slang and thanks very much for the detailed explanation!