Two questions on DSO imaging

[Dave Smith]

1. When it is necessary to image the same object on different nights, I appreciate the need to have fresh darks and flats, but how do you combine the images from each night. Do you just throw the unprocessed tif files from DSS back into DSS to combine or do you process each one first or do just a basic stretch or what?

2. To image an object such as M42 where there is a wide range in brightness I notice that people take a variety of time exposures, e.g. short to get the trapezium details and long to get the faint outer regions. How do you combine those subs e.g. using DSS?

Many thanks for any ideas on these.

Dave

[Parallax]

Dave

To answer number 2 first, have a look at these tutorials. Go to Advanced Image Processing with Photoshop. Your question is answered about 3 minutes in.

As for number 1, I think that it is six of one, half a dozen of the other, but someone else might be along to confirm or deny.

Steve

[ollypenrice]

Hi Dave,

Assuming you get close enough on the composition for DSS to handle the alignment then you could throw everything in, but that would presumably mean that your darks and flats would have to be the same. I don't think (but am not sure) that you can apply a certain subset of flats to a certain subset of images. Personally I move heaven and earth not to disturb my optical train between nights but that is easy for me here.

Or you can make two stacks and combine them, controlling the wieghting in accordance with the number of subs per stack. I do this in AstroArt, Registar or even Ps, but that is a pain. (Maybe I don't know the best way.) You could use DSS to do that, too.

High Dynamic Range Layer Masking in Ps is not too hard but too long to explain here. Send me a regular email address to les.granges(AT)homail.co.uk and I will send you the instructions that I give out here. I think I have some practice files as well but a failed backup disc may have done for them.

Olly

Edit Sorry, crossed in the post with Steve.

[Demonperformer]

1. DSS allows lights, darks and flats from each night to be added separately - there are tabs along the bottom. If they are all the same exposure length/ISO, you can put you bias master in the main group and this adds another tab, where you store your first set. When you start adding to this tab, a new tab is opened. I know it gets at least as far as tab 8, but presumably there is a limit somewhere.

2. Not sure

[Dave Smith]

Thanks Steve, those tutorials are very clear and useful.

Dave

[Dave Smith]

Thanks Olly, if your instructions are a different approach to the video recommended by Steve then my email address is dave@astrosnaps.co.uk

With regard to Q1, if you use any of those methods do you process the images first or work with the separate tiffs.

Dave

[Dave Smith]

1. DSS allows lights, darks and flats from each night to be added separately - there are tabs along the bottom. If they are all the same exposure length/ISO, you can put you bias master in the main group and this adds another tab, where you store your first set. When you start adding to this tab, a new tab is opened. I know it gets at least as far as tab 8, but presumably there is a limit somewhere.

2. Not sure

Thanks DP. I will look into that. It could be just what I am after.

Dave

[martin_h]

http://stargazerslounge.com/imaging-discussion/119533-adding-extra-data-how-done.html

[Dave Smith]

Thanks Martin, I saw the beginning of that thread but somehow missed the later posts.

Very useful. It looks as if I am now sorted.

Dave

[martin_h]

[ollypenrice]

I've just watched the layer masking video and consider the technique a bit primitive because it misses a really good trick. Or two, really. The technique I learned uses the short exposure itself to create the layer mask, which is more subtle than using the selection tool because the mask follows the brightness contours more precisely. Also in their method you don't know what your work in progress looks like till you combine the images. There is a way to have a 'finished product' preview open in real time as you work on the profile of the mask. That is an awesome advantage.

In fact the CS3 version of my notes is short enough to paste here. See wwhat you think.

Olly

Objective; to combine images of different duration in order to cover a high dynamic range (HDR), a problem always encountered in M42 and often in galaxies. In Photoshop 7 it was easy, or at least familiar. Things have changed in Ps CS3.

-Shoot 3 exposure durations. I used 10 secs, 60 secs and 5 mins. I think maybe 12 or 15 secs might have been better. The Tak is about f5 so adjust accordingly.

-Open 2 images which we'll call the Long and the Short.

-Select, Copy and Paste the short onto the long. It is best to have co-registered them first but you can do it now; make the top layer only 50% opaque and push it around till it ligns up. Then go back to 100%.

- In Layers click on Layer Mask icon, the little rectangle with a circle in it.

-NOW GO TO CHANNELS. This is what's changed!

-Alt click on the white mask rectangle in the layer mask channel and have only that channel visible (using the eyeball symbol. De-eyeball the other channels.)

-Ctrl V to paste the short onto the white mask screen. It shows in grayscale.

-Go to Wndow>Arrange>New window for Short. In this window you will see what your work in progress looks like but you will work on the gray mask image, not the new window. However, you will be looking at the new window while mutilating the mask.

- In Filters put a huge Gaussian Blur on the mask image. About 30?

- In Curves grab the graph and haul it around like crazy while looking at the result in the new window. You will soon see what changes have what effects. When you get it right the Short comes through the mask in the right places to fill in the burned out sections of the Long image.

- In Colour balance or Curves you can match the colours of one or other image as selected to get a seamless blend. Keep looking at the window showing the work in progress.

-When happy, Flatten Image and repeat the process for your third data set.

[ollypenrice]

PS Mine came out like this.

[Dave Smith]

Thanks Olly, I will have a play with that. I'm hoping I can still do this in CS2.

Dave

[Dave Smith]

Thanks again Olly. I have finally managed to get that to work after some difficulty although I don't have suitable images to see the proper effect. I've saved those instructions for future use.

Dave

[ollypenrice]

Thanks again Olly. I have finally managed to get that to work after some difficulty although I don't have suitable images to see the proper effect. I've saved those instructions for future use.

Dave

Dave, if you can advise me as an end user on how to improve the instructions please don't hesitate.

Olly

[Dave Smith]

Dave, if you can advise me as an end user on how to improve the instructions please don't hesitate.

Olly

The main mistake I made was following this instruction:-

"

- In Layers click on Layer Mask icon, the little rectangle with a circle in it."

I just didn't notice that icon in the layers "box", I clicked on one of two icons on the left of the screen and had some very strange effects. I think your instruction is clear - it does say "In Layers" - it is just me not ever having used layer masks before. After I clicked on the correct icon all was fine.

Dave

[martin_h]

Have you seen this ?

Compositing 2 Different Exposures via Layer Masks

[Dave Smith]

Have you seen this ?

Compositing 2 Different Exposures via Layer Masks

Thanks Martin. It looks the same as Olly's method but more detailed and with pictures which certainly helps me.

Dave

[ollypenrice]

Excellent link, well done Martin. I wish I'd had that while learning this technique!

It is indeed the same as mine but 20X more detailed.

Olly

[Beyond_Vision]

In answer to question 1 using DSS,

If the temperature is within a few degress between sessions and you havent moved the optical train you can load all the frames into DSS and select intersection in the stacking paramters to create a stack of the parts of the images which overlap.

If you have disturbed the optical train and/or have new darks due to the temperature being different you can stack the images in groups. Select the first set of frames as per normal then you will see a tab at the bottom of the page called group 1 select this then select your second lot of light frames and calibration frames. This can be repeated for multiple sessions.

There is a stacking mode called Entropy Weighted Average (High Dynamic Range) this may help with stacking images which are of a different exposure length. Although the layering techniques in photoshop will produce better results but take longer to do.

Regards

Kevin

[martin_h]

I thought it was - hope it helps:)

Excellent link, well done Martin. I wish I'd had that while learning this technique!

It is indeed the same as mine but 20X more detailed.

Olly