Not a great image, but trying to look for ideas about colouring this image correctly. I have tried every trick I know and cannot seem to lift this image. Mind you it is only 32 minutes of data.
25, 2.5, 2.5, 2.5
Calibrated and stacked in CCD stack.
Longer subs obviously should give better data and longer too.
Any suggestions would be welcomed. Collimation needs attending to on the RC also. Guiding has been an issue too. This image I had to do several sharpening runs just to make something of the data and now it looks well average.
You are being a bit hard on yourself that is a good result for learning new gear.
A couple of critiques:
1. Framing. To my eye it would be better to have M42 at that angle a bit more over to the right. That way the extended nebulosity will flow into the diagonal of the image. This image doesn't show that extended nebulosity but it could have. So if it did the image would have been off balance framing-wise. I usually look at someone elses image to frame the object before I do anything.
2. 25 minutes. Not sure what that means - one exposure of 25 minutes or several at 2.5? I suggest you use a standard exposure length for your camera/scope/mount. The shorter the exposure the easier on tracking but the less you will get of faint details. I use 10 minutes. I used to use 15 minutes. You have to get the signal above the noises of the camera to show anything. As pointed out there is an exposure calculator for different CCDs. 10 minutes is a handy length though.
3. M42 needs a unique approach because the core is so bright. You take a long exposure image of say 2 hours worth of LRGB at 10 minutes each.
Then you take a core shot of LRGB at about 15-30 seconds each. The core details are usually burnt out above 30 seconds but check it for your system and camera by test. Otherwise just take say 5 or more runs of LRGB at 15 and 30.
Then you process each shot separately then you replace the burnt out core of the long exposure with the shorter exposure core shot you took and blend the 2 to hide the join. There are a few ways to do this. There are examples on the net if you google it.
4. Sharpening is best done by selective sharpening. I almost never use
unsharp mask. Its like a blunt instrument that is too harsh very easily.
Again google selective sharpening. There are free tutorials. I think Don Wade has some on that. It basically consists of making a duplicate layer, setting it to overlay, run high pass filter to suit (perhaps 4 or 5 pixels),
layer/mask/hide all, set the background/foreground tool in the tool box to white/black now rub with the brush on the areas you want to come through in the underlying sharpened image and not areas you don't.
You can reverse the foreground/background to cover up areas you accidentally rubbed on (eg stars can look overly sharpened).
When finished layers/flatten image.
Seems like you got a lot of things right though in that image, cooling, autoguiding wasn't too bad, callibration etc.
Paul, I think you may be expecting a bit to much. You have to remember you only have 7.5 minutes of colour data to work with.
I would be inclined to say your main problems are Focus/collimation. Even with the OSC M42 requires quite a lot more exposure to fill the folds in the nebulosity with colour and detail.
This being one of the hardest targets to capture and process correctly I would suggest you persevere on something a little easier like M20 until you find your feet with LRGB.
Colour balance can also be a bit of a pain when using filters and Don Goldman has a good tutorial on his site about calibrating your colour filters with G2V stars. It's worth a look.
It is a huge learning curve when using filters and it will start to come with time. This is one of the reasons I went back to a OSC as I can easily get a reasonable image in one evening without having to spend hours and hours in photoshop. My work committments and patience ran out. No doubt with practice your end result will be better but try not to be to impatient and have everything right and in your favour, by this I mean focus , collimation and seeing. The sensitivity of the CCD will highlight any such problem.
I notice you are using CCD stack, which I have never used but being an image stacking program I expect it has provision to combine the LRGB images and set colour factors. This is where the G2V star calibration will help.
Greg the images were 5 minute subs for the Luminance and 25 secs for the colour. Both you and Doug have hit the nail on the head, not enough time at the scope. I do need to check collimation. Although CCD inspector said 4" out, I think it is more than this. Focus was good as I checked this with a Bahtinov Mask and it was perfect, so I think collimation needs some work.
My previous attempts with OSC were pretty good generally but there is a lot to learn I must admit on LRGB. 10 minute subs on all and then shot subs for the core just like OSC.
I thank you for taking the time to respond. Next time I hope to have a better image.
Perhaps you are still handling the CCD camera as if it were a DSLR. DSLRs are almost always CMOS chips (Nikon used CCDs but now uses CMOS).
They are also heavily amped up with the ISO compared to CCDs (or at least I think so). CCDs do best with long exposure, much longer than probably is feasible with a DSLR. You have to get the signal above the noise sources to be viewable in an image.
They definitely thrive on longer exposure times. I thought my 10 minute subs were a little on the short side but its a balance between exposure, flexure and possible eggy stars.
10 minutes is handy as my dark library only needs to be 10 minutes at 1x1 and 10 minute dark at 2x2.
So I usually use (and so do a lot of others) 10 minute 1x1 for luminance and 10 minute RGB at 2x2. RGB does not need 1x1 as gradations in hue are not sharp like they are in luminance. Some use 1x1 for RGB (I've done this too) for objects where ultimate resolution is important like Globs (where you tend to use shorter exposures like 5 minutes or 4 minutes for Omega Cent or 47 Tuc).
That way you can also desaturate the RGB to turn it back into black and white and add it to your luminance for a bit more signal. I haven't done that much but Marcus did that with his recent lovely image to good effect.
Also aim to do as much exposure time as you can. The really top images you see are usually over 10 hours total. That is not always practical and some images turn out better than others anyway. So on an object you are pretty sure is going to come out nicely (you can tell from the first luminance downloads) it is better to image far fewer objects and for far longer than lots of objects and for short times (I am not that good at this as imaging time can be limited and there are usually several objects of interest to image).
I plan to go for much longer images soon once I move. Perhaps 12.5 hours as normal with the odd 20 hour job on something worthwhile.
Greg.
Quote:
Originally Posted by Paul Haese
Thanks guys for the imput.
Greg the images were 5 minute subs for the Luminance and 25 secs for the colour. Both you and Doug have hit the nail on the head, not enough time at the scope. I do need to check collimation. Although CCD inspector said 4" out, I think it is more than this. Focus was good as I checked this with a Bahtinov Mask and it was perfect, so I think collimation needs some work.
My previous attempts with OSC were pretty good generally but there is a lot to learn I must admit on LRGB. 10 minute subs on all and then shot subs for the core just like OSC.
I thank you for taking the time to respond. Next time I hope to have a better image.
Last edited by gregbradley; 21-09-2009 at 01:46 PM.
That is an interesting point Greg. I had thought that this camera was more sensitive than a DSLR. And; like you say I should not be treating the camera as a DSLR. So 10 miute subs it is. I just have to get my off axis guider co-operating so that guiding is better. Not sure what is going on there as I thought I had guiding nearly sorted with the guide scope. I have some bad egg stars. Perhaps another look at polar alignment and guide parrameters.
One of my mounts is a Tak NJP and has quite low PE to start with.
I have always gotten best results with 1 second autoguiding exposures.
Others have said that is too fast and its better to have longer guide times. That is not the case with my mount. The argument against it is the mount could chase the seeing. But if everything is turned down and not too sensitive I find I get the best results.I did find once or twice that 2 seconds was a bit better when the seeing was poor (in summer) I find 4 secs the PE has built up too much.
Where I image the seeing is usually fairly good and that helps too.
I also have the aggressiveness turned right down and use the lowest correction rate. So its gently gently so its not correcting the last overcorrection all the time.
I also find with guide stars that some give better guiding than others. don't pick a double or don't pick a guide star with another similarly bright star nearby.
I take a 1 sec image at full frame, then I click on subframe (CCDsoft) and drag a rectangle around my guide star of choice. Now the frame is only a partial frame of the full frame (I don't want a brighter star coming into the frame and confusing the software).
I get X Y errors routinely under .1 pixel with the occasional .25 but usually in the range of .05 to .20 pixel errors. This will give me round stars.
Also check no cable drag and your scope is well balanced. Sometimes you get bad errors and its confusing and its because a cable is dragging on the scope.
The first thing to check with bad guiding errors is balance and then polar alignment. An accurate polar alignment makes everything easier including your go-tos.
Greg.
Quote:
Originally Posted by Paul Haese
That is an interesting point Greg. I had thought that this camera was more sensitive than a DSLR. And; like you say I should not be treating the camera as a DSLR. So 10 miute subs it is. I just have to get my off axis guider co-operating so that guiding is better. Not sure what is going on there as I thought I had guiding nearly sorted with the guide scope. I have some bad egg stars. Perhaps another look at polar alignment and guide parrameters.
Ah good point Greg, I noticed the other day that the scope is out of balance in Dec with the finder scope in place. It is a sizeable amount too. I will try the specs you suggested on guiding. I found the em400 is very accurate and on wide field track pretty well with round stars for 10 minutes easily. Try to go back the basics and get it all sorted again. Thanks so much for your help, much appreciated.
I think your LRGB images are coming along nicely, this is a good M42 just needs the sharpening modified as Greg has suggested.
If you are guiding off axis within the camera there is little chance differential flexure is causing those star shapes. If your guiding readouts indicate the autoguiding is going well then you should be getting nice round stars. It is most likely a collimation issue or as I have mentioned before perhaps even some variable astigmatism? The stars in your Lagoon shot look ok..?
Its a fine image, it just seems you havent mastered curves in PS. I stretched it a bit, and theres lots more to see there (although youve clipped the black). Ive followed your fustration in apparently dark images, but its just PS stretching skills, the datas there .
Mike yeah that is always a problem with off axis. The OAG is supposed to stop this but it was almost as bad, so collimation might need addressing for sure. Thanks for the vote of confidence on the LRGB imaging. I think this is gonna take me a long time to get results I will be happy with.
Fred, you are absolutely right. That looks like my base image but I wanted to hide the noise, so I curved down and ended up clipping the image just to hide the noise. More subs and heaps more time is what is needed for sure.
Really appreciate all your comments. Greg I gave that technique a try, I will have to write it down and use it again. Like you said a subtle sharpening routine, not a broad brush stroke.
Yep that is a flare. Peter Wards RCOS gets these too. I have seen an image of his of the Horse Head that has this type of flare. Not worried about this.
Gotta agree with Dennis re: stars. Maybe a bit too much deconv & white clipping?
