Staring at the Sun, 15 :: home :: |
7/29/2010. The huge filaments and the new sunspot (AR 11092) rounding the limb got me out early to shoot through a lot of air mass, but a lot of very still air mass. As threatened, I shot 100 frames. I did that in two groups fifty, the second being slightly offset from the first. My intent was to use the combination to help hide dust. As it turns out, a modest amount of polar misalignment provides plenty of movement so that Registax averages dust into near invisibility anyway (see those nearly vertical dusky streaks near the top of the solar image?). That's what the first set of fifty frames looks like, using most of what I've learned since I got serious about staring at the sun about six weeks ago. Two barlows, ISO 400. Double stacked Lunt LS60. 50 x 0.2 seconds. Canon 50D. Maxim script to isolate red plane from the CR2 datafile, stretch, and save as 16 bit TIFF. Align and stack in Registax using multiple alignment points (the spot, the gauzy portion of the big filament, and the thinner end of the small, sharp loop in middle lower right). The already impressive result benefitted from a Focus Magic pass followed by a little local smart sharpening and a despeckle pass in Photoshop. It's presented here at half scale and colorized. Here's a monochrome version. In this optical arrangement, the solar image is 2,500 pixels across in a single Bayer-plane. So it's 5,000 pixels across on the sensor. The 50D's sensor has 4.7 micron pixels. 5e3 pixels * 4.7e-3 mm = a solar diameter of 23.5mm. Which implies an effective focal length of about 109 * 23.5mm. Call it 100 * 25 = 2,500mm. The aperture of the external etalon is 50mm, so we're working here at about F50 when all is said and done. We're spreading 30 arc minutes or 1800 arc seconds across 2500 (red sensitive) pixels, for an effective pixel scale of 0.8 seconds per pixel. At 50mm, the diffraction limited resolution (via Dawe's limit) would be about (116/50) or 2.3 arc seconds. So at this EFL, we're getting a reasonable to generous oversampling for aggressive post-processing.
|
:: top ::