Whirpool Galaxy last night

[ryandsimmons]

I am slowly starting to get the hang of this I think now.

Last night did 30 frames of 45s, some flats, and unfortunately only 5 darks (as I was getting rather tired and had to get up early this morning)

I was just wondering if you guys had any comments on obvious issues that I should be looking at?

The obvious one I can see is vignetting (I learned something!) which the flats didn't seem to get rid of properly, next time I will take loads more to try on that point.

There is also some streakiness with the picture.  I will have a look at the individual frames tonight to see if I can figure out where that is from, possibly condensation on the scope?

But the main thing for me is that I think I captured the Whirlpool Galaxy at least okay.  Still not at the point where I am entirely happy, but it's something I can show my wife with pride and justify all the cash spent on the equipment!

Quite scary how much appeal imaging has.  From the first time I tried it a couple of months back I haven't actually had an observing session, any clear skies are an exciting attempt to get some lovely pics!

A better res picture is here:

[ryandsimmons]

Damn, can' edit.

Anyway relevant info:

C8 on a CG5 mount.

Canon EOS1000d camera (unmodded)

ISO 8000

30 frames of 45s, of which 26 were stacked.

5 darks

3 flats

Stacked in Deepskystacker (all default settings.)

RGB mountain things adjusted to align in DSS, all other post processing done in Photoshop.  All I did in post though is adjust the levels slightly and tweak the curves.  I don't actually have a clue what else to do just yet.

[carastro]

Well done, it's addictive isn't it.

As you rightly said, vignetting and flats need to be addressed and it would help if the object of interest could be centred.

You say you have done some flats, but it's possible you didn't do them correctly, they can be a bit tricky at first.  If you over expose them they won;t do anything to improve the image.  With a DSLR this is relatively easy.  I used to set my camera to 100ISO and AV and pouint at a - not too bright - diffused light.  In my case I was able to leave the camera set up until the following day and then I would point at the not to bright sky with paper over the front to diffuse the light.  If you are not able to leave your scope and camera set up, you can use a white computer screen, or some people make light boxes.  

This is mine, inside a cardboard box with an EL panel (electro-luminescence panel from Earlsmann), inside and sheets of paper to dull.  AV will take to right length sub and the histogram should be around 1/3 - 1/2.

HTH

BTW you can check if your flats are OK, by taking the master flat into Photoshop and stretching the brightness and contrast )play around with it) so see if you can reveal vignetting and dust bunnies), if you can then your flatsa re OK, if you can't then they were not done correctly.  

Carole 

[ollypenrice]

Firstly well done and, yes, it is hideously addictive!

To make your flats work is the first thing. Done properly they should cure the vignetting, as Carole says. The key things with flats are;

-  Expose till the hisotgam peak is about a third of the way to the right. (You can usually check this easily on the camera.)

-  Take a set of bias frames (the shortest your camera can take, with no light getting in) and use these as 'darks for your flats.'  This stops flats from over correcting.

-  Take at least 20 flats and 50 bias. (Your shortage of subs is not the cause of your flats not working, though. Three would work if they were correct but would bring noise into the image.

 You can buy electoluminescent panels for flat making as well. They need diffusers (like sheets of typing paper...)

A credible flat, once stretched, looks roughly like this with a bright centre, dark corners and some dust bunnies.

The streaks in the image are not of optical origin. They are typical of many DSLR cameras. It would probably help to take longer sub exposures and, after that, to learn the dark secrets of image processing to reduce them!

Olly

[Stargazer33]

That is a very good effort! Much better than my first efforts with a galaxy. In fact better than any galaxy that I have taken!

I'd say it definitely looks like you are getting the hang of it.

[carastro]

Whoops I seem to have omitted the photo of my light box in my earlier thread:

[carastro]

Take at least 20 flats and 50 bias

Olly, I have never taken as many as 50 BIAS, and normally only take 15 flats.  Please can you explain the reason why so many are needed, it could be that addressing this issue might bring about improvement to my images, but I do like to understand why.

Sorry to hijack thread, but I am sure the original poster will find this informative as well.

Thanks

Carole

[ryandsimmons]

Yep, the more info the better for me.  It's not so much a learning curve as a learning cliff I find, with every step being insanely obtuse for me. 

By the way, are there any other targets people would highly recomend?  I am in my garden in an estate so surrounded by houses,, meaning I try for high up targets only.  Whirlpool is great as it is easy to see and get a reasonable picture plus is at the zenith.  I would like to try something else tonight though.  I did try a 3 minute test exposure of the Sombrero but got very little so I assume that is a tricky target?  The Crab I did some test shots last week and I think is a more advanced target as well.

[IanL]

The streaks are due to a lack of dithering between frames.  With a DSLR you're going to get a fair bit of pattern in the background which isn't really that obvious in a single frame (it will tend to look like random noise).  When you stack the images, the repeating elements in the pattern will line up and suddenly you'll spot streaks or similar.

In this case there is a steady diagonal drift in the top-right <-> bottom-left direction. This is most likely due to being a bit off in your mount's polar alignment; even if you get a better polar alignment you're still likely to get this kind of effect so that isn't the solution.

Instead what you need to do is dither between frames, i.e. re-point the mount slightly between each exposure (the recommended amount is so that the target moves perhaps 12-15 pixels between each frame).  The movement needs to be random each time, so a slightly different movement distance and in a different direction each time.

This doesn't get rid of the pattern, but distributes it in such a way that it's effect is reduced by the stacking process, plus any remaining pattern ends up looking like random noise rather than streaks.

If you are guiding, dithering is easy as you can set the guiding program to do it automatically between each exposure.  If you aren't guiding (which I don't think you are) then maybe take a few exposures, pause the camera, use the software or handset to jog the view a bit (remember 12-15 pixels each time) and then take a few more exposures, repeat.  You don't want the target to be moving a long way across the image between the start and the end, you want it to 'wander around' randomly within the same area.

[ryandsimmons]

Ouch!  That is an added complication I really didn't want.

I am not guiding, I had a brief look into it and decide that I would try to master unguided first.  Baby steps and all that.  (And an empty wallet!)

I hate trying to move the image in the frame as I do not have a laptop and it's a nightmare.  Hence I left the image of the whirlpool to the side rather than trying to fully align last night.  What I have to do is GOTO the target.  Take a 3 minute exposure.  Look at the camera while crawling on the floor as the scope is near vertical, try to zoom in and hopefully see a fuzzy in the image, and if it is there I just leave it.  Trying to align the fuzzy is difficult as I cant remember which direction is which on the handset and every movement needs another 3 minute exposure and a cricked neck and wet knees.

Mind you I suppose for the purposes of the above I can simple do a brief press in a random direction on the handset and as long as I vary them it shouldn't move the image too much around the frame.

[ollypenrice]

Olly, I have never taken as many as 50 BIAS, and normally only take 15 flats.  Please can you explain the reason why so many are needed, it could be that addressing this issue might bring about improvement to my images, but I do like to understand why.

Sorry to hijack thread, but I am sure the original poster will find this informative as well.

Thanks

Carole

Noise goes down as the square root of the number of images in the stack. That means taking four times as many will reduce the noise to half of what it was. I'm sure that 15 will be fine and adding another five probably wouldn't make any difference, but ours is a game of signal to noise, so every little helps. I feel, as do many others, that bias are so easy to take that you might as well take plenty of them. The reason they make good darks for flats is simply that they are almost identical to darks taken at the same exposure time as the flats. (When exposures are very short thermal noise doesn't build up significantly so there is very little difference between a one second dark and a bias.)

It's a matter of debate how useful it is to take more than a certain number of images for stacking. Tony Hallas puts the number quite low, around 30 from memory. The authors of Astro Art suggest that diminishing returns set in at around 40. (Again from memory.) If you want to use a sigma clip stacking algorithm then it will really start to work well when you get to a dozen or so. I know from experience that 35 light frames beat 20 but beyond that I've never explored. Now that I'm into the wonderful world of 30 minute subs I doubt that I'll ever be in a position to find out!!

Olly

[glowingturnip]

very nice pic !  maybe get the Gradient Xterminator plugin for Photoshop - has a free trial period, and should get rid of all that vignetting for you.

By the way, are there any other targets people would highly recomend?  I am in my garden in an estate so surrounded by houses,, meaning I try for high up targets only.  Whirlpool is great as it is easy to see and get a reasonable picture plus is at the zenith.  I would like to try something else tonight though.  I did try a 3 minute test exposure of the Sombrero but got very little so I assume that is a tricky target?  The Crab I did some test shots last week and I think is a more advanced target as well.

maybe M101 ?  Similar size and brightness to M51 and in the same area of sky

[ollypenrice]

very nice pic !  maybe get the Gradient Xterminator plugin for Photoshop - has a free trial period, and should get rid of all that vignetting for you.

maybe M101 ?  Similar size and brightness to M51 and in the same area of sky

It's a bit optimistic to think that Grad X will deal with vignetting. It can and will help, as will the more sophisticated DBE in Pixinsight. However, both systems work by creating a general gradient map and flats don't produce a general map, they produce a specific one. Dust bunnies will be included. Also, systems like Grad X and DBE struggle if you have nebulosity in one part of the image and not the other. They need you to tell them what is and is not background sky. That's fine if you have a reasonable distribution of background sky but if you don't.... And, finally, imagers seeking the faintest signal steer clear of any GradX or DBE filtration because only flats can really tell true signal from false. I really think that getting to grips with flats (which can be a battle!) is worth the effort.

Olly

http://ollypenrice.smugmug.com/Other/Best-of-Les-Granges/22435624_WLMPTM#!i=2266922474&k=Sc3kgzc

[glowingturnip]

totally agree, flats are definitely best practice.  I take loads of flats straight after the lights (if I remember !   )  and use DBE very carefully.

For this image though, seems the OP's flats aren't working for some reason - using GradX could rectify that to some extent and produce quite a decent end result without having to bin everything and start all over again.

@ryandsimmons - the flats you took *should* have removed that vignetting for you, but looks like something has gone awry somewhere. What do the flats look like, are they over or under exposed ?  How are you using them in processing ?

Cheers

[ryandsimmons]

I tried doing flats again last night.  This time I took around 15 and just threw them into DSS as it had no vignetting at all.

I think I tracked down the problem, when I looked at the individual frames for the above I noticed that there was a variation in the vignetting on each one, by walking back into the garden and setting up the scope in the same position I beleive that this was caused by the neighbours upstairs light being turned on for some frames which bounced around inside my dew shield and lit up one side of the image.

So on the plus side I found the cause, on the downside though it means my area of sky that is good for AP is further reduced.