I've been using my STF-8300M for about a year now. It has the FW8 wheel and I am using the Baader LRGB 36mm filters with it. Over the last year I've noticed that the camera consistently loses the blue on the fainter objects. For brighter objects like stars, the RGB filters all seem to have about the same response and the stars come out white. But for fainter things, like the blue glow around M31 and other galaxies the blue is simply missing from the images. I've tried increasing the blue values when stacking and processing but all that does is throw off the white balance without giving me the fainter blue that I am expecting. I 've also tried exposing blue longer than red and green, which does help a little but really not very much and has the unwanted side-effect of causing blue halos around all the white stars, even the small faint ones. It is my practice to use the same exposure time for all three color filters, typically 4 or 5 minutes but I have gone as short as 2 and as long as 8. The anemic low-level blue response seems to be consistent regardless of exposure time. As mentioned above, but worth repeating, the blue response seems ok on bright objects, like stars for example. It also did a good job on the blue portion of M57 (ring nebula). But its when the blue is faint it seems that the camera ignores it. My guess is that there is a non-linear response going on and when blue drops below some intensity level it becomes attenuated. But that's just speculation based on what I'm seeing and the actual cause could be something else. So, anyone out there have any ideas? Dave
It's all about how much light is transmitted all the way from the object to the chip, and the CCD chip's sensitivity. Most of what you are experiencing is "normal". From the data sheet (http://www.kodak.com/ek/uploadedFil...lutions/Datasheets(pdfs)/KAF-8300LongSpec.pdf) you'll see the chip's quantum efficiency is 33% 40% 33% for RGB. But if you look at the graphs, you will see wider wavelength range for green and red than blue. Green will need the least exposure time, blue and red will need more. Next, the transmissivity of the filters (eg how much light at a particular wavelength range) gets through will be the next factor that makes for a longer exposure time. So you need the data on the filters you have. You'll find some filters pass less than their mates, so that makes for a longer exposure as well. Next, the wavelengths of the emissions from a nebula like M57 or M27 are quite strong in OIII (greenish-blue), compared to glow from hot young blue stars in some galaxies. Add all that up, and your blue exposures need to be substantially longer than the green.
Hi Colin, Thanks for the info. I looked into the filters and from the graphs I saw that the filters transmit basically the same amount of light across the spectrum. At most it looks like there is only 1 or 2 percent difference and that varies depending on the specific wavelength even within each color. I also checked the Kodak specs and indeed, the blue response on the sensor is lower than green but it is roughly equal to red. Also noted that the red area occupies more bandwidth then blue. However, when the camera is used with the filters, the excess red bandwidth is cutoff and each color is collected across approximately 100nm of spectrum. There is some variance here though, blue seems to be about 150nm wide, green maybe 110 nm wide and red about 125nm wide. So, linearly speaking, the blue filter casts the widest net (but if using a log scale for wavelength its not so clear-cut). Anyway, the numbers don't seem to add up for me based on the actual response I am seeing. From the filter and sensor graphs, yes I should need to gather more blue, or give it more weight when processing. But not all that much, maybe 10% or 20% difference would seem appropriate. So I've tried both longer exposures and boosting the blue after the fact and still have almost no low-level blue where I would think it should show up. Instead I seem to get almost no blue at all unless its at or near saturation. Sort of an all or nothing response. So... is there a test you or anyone knows of that I can conduct to determine if its the camera, the filters or maybe just my processing technique that's causing this to occur. Below is a crop of NGC 2403. The exposures are 5 minutes each for LRGB. The stack is about 20 frames from each filter. The blue was boosted about 20, maybe 25% in post processing. As a result of the boost, many of the white stars now sport a stylish blue halo. But the galaxy is still missing its lovely blue arms... (same thing in other galaxy images, blue arms are missing or come out greenish or reddish but not ever bluish). Edit: also should add that this happens with 3 different imaging scopes. I have a 4in APO, 8in Mak-Newt and an 11 inch SCT. So... pretty sure its not caused by the scope.
One approach would be throw a camera lens on, use a color test pattern and a "daylight" (~5000K) light source. Then calibrate against that test pattern, and you will know the levels for each of your RGBs. You could also image some stars with known colour - Al Kelly has an old article with a list here: http://www.kellysky.net/artdraf7.htm Another tip would be to adjust the focus for each color - often there is a slight offset needed for each filter and for the scope. Lastly, the atmosphere affects this quite a bit - the higher the elevation, the less air you are going through, so the color extinction is different than when shooting something lower down.
Hi Colin, I will try the stars with known color approach. The camera lens, test pattern and light source is a good suggestion too but I don't have a camera lens or test pattern handy. I do have some "daylight" lamps so if the star test gives me trouble then I can look into that method further. The filters are supposed to be parfocal, at least according to Baader. But it never hurts to check... So far I've been running them all at the same focus. Yes, the atmosphere affects things for sure. Because of my location, obstructions mean that I have to point pretty high up anyway to avoid the trees and buildings. Even so, on white stars I often see some color separation (eg: usually a bit red on one side and bit green or greenish blue on the other side). Sometimes aligning the RGB planes corrects this but sometimes its uneven across the field so I just leave it. Its not usually a big deal except in winter when the air causes the color to really separate. The article you referenced is quite interesting. Dated yes but still good to have some more background. Lots of the white stars in the list should be visible from where I am so I can give them a shot, maybe even tonight.
Just a nuance for you - the filters may be parfocal, but not all instruments focus colors to the exact same point - usually its very very close with quality scopes, however there is a bit of discrepancy across the wavelengths.
That is referred to as spherocromatism. Schmidt-Cassegrains have this to some extent. It was quite significant in the old-style Meade 16".
