Quote:
Originally Posted by mithrandir
When are we likely to see a smarter method of producing alignment info for our own images than trial and terror? Will we need to plate solve them first?
|
Hi Andrew
The json display method needs to know the coords of the corners of the picture. This information is included in the headers of the fts images from the DSS that Stellarium can download and display. The DSS display is not yet activated in stellarium but it is in the Virgo version that is available on the internet.
Converting images to this form will be a problem for most because at the moment you must know the exact centre, size and offset angle in order to calculate the corner coords.
The way I find the centre is by locating an identifiable star as close to the centre as possible and which direction north is. I then rotate the image in photoshop to make north squarely up (within a degree or two) and crop the image so that the size is 512 x 512 pixels using the known star as the centre of the image. This gives me a basic texture that is much easier to manipulate.
I next locate two more stars with known coords to calculate the separation. Stellarium has a way to do this easily. From this I can calculate the size of the texture quite accurately using maths.
I then apply these details to another program I wrote in Qbasic that calculates all the necessary coords and writes the data to a file in the format required for an insert into the textures.json file. I am reworking this program to include the size and angle calculations but it is not finished yet.
Because the photos are linear applying them to a spherical display requires quite a bit of trigonometry to find their corners depending on how far they are from the poles. At the moment my program works pretty well for textures under 1 degree square and within 30 degrees of the celestial equator but outside this errors start to creep in. I think this is mostly due to the precision of the calculations. In any case the textures are close enough for fine tuning by hand. See the mosaic I made of the Eta Carinae area.
Also take a look at the two textures IC1805 and IC1848. These are very large areas that took me quite some time to locate the necessary stars, size and angle because I had no information at all other than the names. Once I found this information I used it to fine tune my program but it still needed some manual manipulation for the final display.
Barry
You'll soon hear if it does.
