Ok, guys, what am I doing wrong? I'm just playing with it, trying to learn how to use it. Apparently, it's easy to use. I must be doing something wrong then. I've taken 2 shots (yes, only 2, just playing) of M8. I've shot them in RAW (CR2) and convered them to 16 bit TIFF files using Canon's DPP. Said TIFF files are in colour. When I open both files in DSS and register/stack them, the result is mono, and has a LOT of noise. Yes, I know I haven't done any flats/darks etc (forgot to last night), but I still feel that the amount of noise being generated by DSS is horrenous. And I can't for the life of me figure out how to get the final stacked shot to appear in colour either. For reference, I've attached a jpg of one of the TIFF shots - if you look closely you can see some colour, but it's a bit harder to spot cos of the resizing etc. I've also included a jpg of the final output of DSS.
Deep Sky Stacker will handle your CR2 files. Don't convert your images or anything, just open Deep Sky Stacker and load the images through the open files command. Then look for the recommended settings, select those and check all tabs so you know where you saved it and that is it. Or should be anyway, works for me!!
1) Reads in CR2 raw files.
2) Stack as Median (default is Average)
3) Align RGB in final Image (Tick on first page of stacking settings)
4) Save final image as 32bit Rational TIFF (defaults to 16bit which gives an almost mono feel to the picture)
After DSS is done, save the file into 16-bit TIFF format, then load this into DPP , play with curves and then convert and save as JPG.
You can have a look at what I am doing, I tried to explain the DPP post processing here (scroll down close to end of page until you see explanation in pictures): http://www.iceinspace.com.au/forum/s...t=48651&page=2
EDIT: If by accident you used "Per channel background calibration", this will have "colourless" appearance of the background.. and in LP areas will screw-up the colour balance of the whole image.
EDIT 2:
The noise you see is normal consequence of only two images in a stack.
To have noise visibly reduced, you have to have at least 5 images in stack.. but not more than 20, going above this number there is not much difference, but the processing only takes longer.
Thanks guys, will play with raw and see what happens. I still don't see why there's so much noise from the stacked pair, especially when noise in the individual images is far less. I'll put it down to my lack of understanding and knowledge on astro imaging ;-)
I only tried a bit of imaging last night because a mate who'd come over said 'go on give it a go'. Since I've had problems with the focuser on the Equinox, I've really lost a lot of the interest that I had in imaging initially. I was pretty happy to be able to focus last night, nay, I was REALLY happy. I'm limited to ten seconds it seems, 15 seconds results in elongated stars due to poor alignment.
Next time I get the scope out, I'm going to give it a try again.
What's happening with the focuser??. Too much play in the draw tube?? You may need a better focuser assembly, or you can tighten the one you have up a bit to make it a little more robust.
mmm no go for this lad. Stacked image has less details than the actual RAW files themselves. I'll double check the settings later in DSS, maybe I still have something set wrong.
What's happening with the focuser??. Too much play in the draw tube?? You may need a better focuser assembly, or you can tighten the one you have up a bit to make it a little more robust.
It just simply won't grip properly if the OTA is pointing even remotely upwards. Even with a 2" eyepiece (no camera). My mate noticed it last night. You basically have to put pressure on the bottom of the 2" diagonal as you're trying to focus in order for it to grab. I've tried tightening etc, does S.F.A. I was so surprised that it focused last night with the camera on. The last time I tried, it only not only refused to come to focus witih the camera on, but when I took the camera off and put the diagonal and eyepiece back on, I couldn't gain focus with it either. I can only think that the internal focusing unit is damaged. I prolly should return it to Andrews, especially since I raised it as an issue within 2 months of buying the unit.
If it's still under warranty, I would return it. If not and you're still having problems I would get a replacement. A good Moonlite or Feathertouch focuser would be spicky.
I still don't see why there's so much noise from the stacked pair, especially when noise in the individual images is far less.
Dave
I do not think so.
Try to increase the contrast of the individual CR2 image to the same extent as a stacked image and you will see it contains LESS noise.
The important thing to remember is that the S/N (Signal to Noise ratio) is what we are talking about here and what we are increasing by stacking a number of individual frames into a single image.
Bojan is right, what you are trying to do by stacking is increase the signal to noise ratio. By only having 2 light frames and no dark frames, you havent really accomplished that in any major way.
As it says in the opening lines. Why combine?
The answer is simple: only to increase the Signal to Noise Ratio (SNR). Is the resulting image more luminous? No.
Is the resulting image more colorful? No.
The goal of combing many images into one is only to increase the SNR. The resulting images are neither more luminous or more colorful but they contain much less noise which will let you stretch the histogram a lot more which will give you more freedom to bring back colors and details.
Thanks Trevor - I did read that bit in the FAQ/help file actually, but it didn't make sense to me at the time. Despite a busy day @ work and feeling tired from it all, it sunk in just now. :-)
Dave
PS you still doing ciscos etc? My bosses are wanting me to learn them...I'd be happy to if I actually had the time @ work lol
If it's clear on the weekend, I'm going to give it another crack at M8 and try and get 30 or so 10 second subs. Hopefully I'll get something out of it that's usable. I'll try and remember darks and flats as well. Thanks again everyone for your advice.
Nooo, don't be confused! The easiest way to think about it is to treat your flat frames just like light frames. You take darks for lights, so take equal length exposures for your flat lights.
Some people will tell you to take bias or offset frames to remove sensor readout noise (1/4000 or 1/8000 second exposures, the shortest your camera can do), but, in my experience they tend to ruin or add artefacting to my images. I also figure that the bias is also recorded in the dark frame, so, shouldn't really need it. Might work for others, just not for me!
Hope that helps.
Regards,
Humayun
Quote:
Originally Posted by dpastern
Oh now I'm confused...more research required by moi.
David, do not worry about Darks and flats for the beginning.
Just play with DSS stacking to obtain the feeling of what it does.
As for darks, they are just exposures of the same duration taken with lens cap on.
If you set your camera to aumatically compensate for long exposures (this will be on special functions menu, Long exposure noise reduction), the camera will take darks for you and do all necessary math on image file.
So you just carry on with stacking.
However, this method (LNR) is inferior to taking separate darks and stacking them in DSS (it is more noisy).
Taking Flats are very easy task, but perhaps not necessary at the beginning. They will compensate for so called "vignetting" which is un-even illumination of image frame (lighter in the centre, darker at edges).
