M31 Mosaic From POSS-II Plates

[Knight of Clear Skies]

Thought I'd put this in the Astro Lounge as it's not really my image. Following on from Gorann's thread here, I thought I'd have a quick go at putting together an M31 mosaic using images from the Second Palomar Observatory Sky Survey, taken with the 48" Samuel Oschin telescope. These were photographic plates taken in the 80s and 90s which are now available for download here, they are freely available for non-profit activities.

(Worth a view at full size.)

It was good fun to play with this data, it wouldn't have been possible to manipulate the plates like this when they were first taken - if you wanted a mosaic you'd have had to lay them out on a table.

To download the individual frames I chose the blue channel, put 60 arcminutes in the width and height boxes, selected .gif and clicked the 'Retrieve Image' button. From there I used Stellarium to get a rough idea of the coordinates for each panel and edited the URL directly to grab each frame. I haven't done a great deal to it. I threw the frames into Microsoft ICE to stitch them together, then in PS I adjusted the curves a bit, increased local contrast to bring out the dust lanes more, applied a bit of spot healing and resized.

[happy-kat]

Interesting idea, I haven't the time to divert my own learning off processing of my meagre data. did this exercise take much time?

[John]

Great result

It always amazes me, when I see a really deep image of M31 like that, how apparently close M32 and M110 are to M31, compared to the visual view we get.

 

[Alien 13]

Very nice image, it does make you wonder how big it realy is some estimates say its already touching the milky way.

Alan

[Knight of Clear Skies]

 

13 hours ago, happy-kat said:

Interesting idea, I haven't the time to divert my own learning off processing of my meagre data. did this exercise take much time?

Didn't take long at all, I put my workflow in the OP. Fiddling around downloading the 15 or so frames took half an hour or so, ICE stitched them without complaint. I downloaded the GIFs for speed rather than trying to mess around with the FITS files. The source images are free of distortion - the film was shaped into a curve before loading it into the camera.

There is lot of scope to use these images, for example Gorann also had a go at colourizing one using his own data. This took a lot of work in PS but Registar would probably handle it well.

13 hours ago, John said:

Great result

It always amazes me, when I see a really deep image of M31 like that, how apparently close M32 and M110 are to M31, compared to the visual view we get.

 

The human eye is very sensitive, some experiments suggest we can consciously perceive individual photons. The camera's best trick is to sit there soaking up light until dawn (or more commonly, clouds) intervene.

It's an impressively deep image for film but it also shows how spoiled we are with amateur equipment these days, I got a similar depth with a cheap DSLR and lens although I couldn't get any where near the resolution, and I can't really process it to my satisfaction.

But this is the deepest M31 image I've come across, it becomes difficult to distinguish the outer parts of M31 from the IFN (dust in our own galaxy).

I like to think of deep images as being both revealing and deceptive, as they have to compress regions of enormously varying brightness into a narrow visual range. Shallower images and visual observations can compliment them well, to give a more complete view of an object.

[dph1nm]

On 14/02/2017 at 20:06, Knight of Clear Skies said:

It was good fun to play with this data, it wouldn't have been possible to manipulate the plates like this when they were first taken - if you wanted a mosaic you'd have had to lay them out on a table.

Just to be clear, the original plates are 6 degrees across, so M31 fits on one plate quite nicely! It is the digital version which limits you to 60 arcmin boxes.

NigelM

[Owmuchonomy]

Very interesting post. 

[Knight of Clear Skies]

On 16/02/2017 at 13:51, dph1nm said:

Just to be clear, the original plates are 6 degrees across, so M31 fits on one plate quite nicely! It is the digital version which limits you to 60 arcmin boxes.

NigelM

Thanks for clearing that up, didn't realise they were that big. I looked at the scope's aperture and focal ratio to work out its focal length (somewhere over 2000mm I believe) but I didn't realise just how big the photographic plates used were, 11 or 14 inches or many times bigger than a full-frame camera.

I haven't come across a noticeable plate boundary yet, but possibly they have been calibrated to match when digitized or I haven't browsed enough.