|
引用 | 編輯
aa
2014-01-15 19:48 |
1樓
▲ |
|
不過他的拍攝資訊我看不太懂也
I shoot the planets with my 14" f/4.5 (native focal length is 1600mm). I add a 5x barlow which gets me close to 9200mm (isn't not preciesly 5x but varies) Then I add a monochrome Flea3 camera from PointGrey Research. To get color I shoot with filters to give me Red, Green and Blue images. Each color image consists of capturing many frames of each color as fast as I can. I shoot the red at 120 frames per second (FPS), then the same for the green and bit slower for the blue as it's not as sensitive. So for example the red is running at 120fps for 30seconds. That nets me ~3,600 images (stored in .avi video format), again for green and then blue. Wait on second and start another sequence of R, G and B, 30s, 30s, and 30s. So each color image is 90 total seconds of capturing. I capture back to back and if you can believe it, the moon Io (right one) moves so fast in 90seconds I had to adjust the color channels in the final image. I capture many RGB runs back to back for a couple of reasons. Firstly, I use a very sophisticated program called WINJupos to derotate (a method of combining individual images that have rotated to a different central meridian or longitude) all of them. This increases the signal-to-noise present in each of the single RGB runs. I also capture many RGB runs in order to make a cool animation of the moons moving and planet rotating. Raw frames as promised: Nearly the worst of a single sequence of red filtered images Best of 3600 frames in Red: Normally I stack the best 75% of the 3,600 frames in AutoStakkert which uses a process called 'lucky imaging' After stacking I sharpen using some deconvolution tricks in AstraImage. I also combine the filtered monochrome frames to a composite RGB. Then derotation win WINJupos. Then photoshop magic like denoise, color, high pass filters, rgb channel alignments. I have yet to look at the features on Ganymede to see if they're accurate or if I just messed up in my processing treatment.  x0
|
|