Details count...
Aug. 29th, 2000 11:11 pm![[personal profile]](https://www.dreamwidth.org/img/silk/identity/user.png){kind=small}
alexpgp...especially when they're major.
I spent a restful 15 minutes outside earlier this evening, looking up at what passes for a night sky south of Houston, hoping to catch sight of the Mir space station as it whizzed by nearly directly overhead. Too bad my timing was off by an hour. Ah, well.
At the very least, I learned how to do a 16-second manual exposure with my digital camera. Since I was pointed in the general direction of Scorpio anyway, I decided to take a photo or two just to see how much detail I could pick out of the night sky. After transferring the shots to my computer, I compared them both against a program called Distant Suns, which was published by Virtual Reality Labs about a decade ago, and A Field Guide to the Stars and Planets.
I found, as I expected, that a 16-second exposure reveals more stars than you can see just looking up at the sky. A telephoto shot narrows the field of view to the extent that stars are no longer points, but elongated ovals, which is caused by the camera rotating slightly along with the earth (to which the camera is fixed), which causes the stars to appear to move in the sky. This effect is really noticeable with that really big star, the Sun, which appears to move in the sky, too.
In looking closely at the images, I also found that there were a lot of blue and red pixels in the photo, and I'm not quite sure they are the result of starlight. There are some points of light that are fuzzy, and others that are neither shown by the program or displayed in the star charts in the book. Then again, there are stars that should be there that don't appear in the image, and some dots of light in the image that are too bright, given where they are. I suspect that the photosensitive CCD element in the camera doesn't do well at very low light levels, resulting in some types of stars registering "too well," while others register hardly at all.
I tried manipulating the images, cropping one of them to eliminate some flare in the corner, then experimenting with various levels of JPG compression, finally getting a good-looking image at around the 30% level. After resizing the image from 864 pixels across to 250 pixels across, the final product stores in an amazingly small 7K, whereas the original image was 3.6 MB (though I cropped about one third of it). The only problem with the whole process is that the compression does funny things to the image, creating stars where there are none in the original image, and vice versa. Resizing also affects the quality of what's on the photo, so for the time being - assuming this technique holds any promise at all - I'm going to be stuck with native, huge BMP files.
Cheers...
I spent a restful 15 minutes outside earlier this evening, looking up at what passes for a night sky south of Houston, hoping to catch sight of the Mir space station as it whizzed by nearly directly overhead. Too bad my timing was off by an hour. Ah, well.
At the very least, I learned how to do a 16-second manual exposure with my digital camera. Since I was pointed in the general direction of Scorpio anyway, I decided to take a photo or two just to see how much detail I could pick out of the night sky. After transferring the shots to my computer, I compared them both against a program called Distant Suns, which was published by Virtual Reality Labs about a decade ago, and A Field Guide to the Stars and Planets.
I found, as I expected, that a 16-second exposure reveals more stars than you can see just looking up at the sky. A telephoto shot narrows the field of view to the extent that stars are no longer points, but elongated ovals, which is caused by the camera rotating slightly along with the earth (to which the camera is fixed), which causes the stars to appear to move in the sky. This effect is really noticeable with that really big star, the Sun, which appears to move in the sky, too.
In looking closely at the images, I also found that there were a lot of blue and red pixels in the photo, and I'm not quite sure they are the result of starlight. There are some points of light that are fuzzy, and others that are neither shown by the program or displayed in the star charts in the book. Then again, there are stars that should be there that don't appear in the image, and some dots of light in the image that are too bright, given where they are. I suspect that the photosensitive CCD element in the camera doesn't do well at very low light levels, resulting in some types of stars registering "too well," while others register hardly at all.
I tried manipulating the images, cropping one of them to eliminate some flare in the corner, then experimenting with various levels of JPG compression, finally getting a good-looking image at around the 30% level. After resizing the image from 864 pixels across to 250 pixels across, the final product stores in an amazingly small 7K, whereas the original image was 3.6 MB (though I cropped about one third of it). The only problem with the whole process is that the compression does funny things to the image, creating stars where there are none in the original image, and vice versa. Resizing also affects the quality of what's on the photo, so for the time being - assuming this technique holds any promise at all - I'm going to be stuck with native, huge BMP files.
Cheers...
