First M74

[Russe]

Hi guys,

last weekend's photosession has produced my first ever go on M74. Not easy to process. Hope you like it.

10 hours 45 min data

Canon 450D (unmodded), guided in 300s exposures at ISO800

SW 130PDS on HEQ5.

Don't judge too harshly...

[jambouk]

Too harshly?!?! It's a little beauty! You myst be over the moon (well under it at present but you know what i mean).

I can't comment on any of the technical aspects, but from a simple outsider it looks smashing. I hope you are going to print it out and stuff it on your wall and admire it!

Well done and thanks for sharing. Did you encounter any problems collecting the data with your kit, or processing it, which might help others with similar kit and aspirations?

Great work.

James

[wayneh]

Have to say i would be over the moon with this my self

[Astrosurf]

That's a lovely image. I must try that object myself!

I would try to bring down the green a little though. What processing sw did you use?

Alexxx

[Russe]

PS CS5.

Thanks for the kind comments. Will check on the green. Colour blindness doesn't help but I use "colour samplers" in PS. Thought I had corrected it...

[Russe]

Too harshly?!?! It's a little beauty! You myst be over the moon (well under it at present but you know what i mean).

I can't comment on any of the technical aspects, but from a simple outsider it looks smashing. I hope you are going to print it out and stuff it on your wall and admire it!

Well done and thanks for sharing. Did you encounter any problems collecting the data with your kit, or processing it, which might help others with similar kit and aspirations?

Great work.

James

The problem with processing is that when stretching the data, the outer arms of the galaxy are just above background noise. So in order to control background noise, one needs to put an anchor into curves exactly at the background noise (and one below) and stretch only above. The difficulty is that some colour in the galaxy is below the same colour in the background noise. The other problem I had was the bright star in the lower left of the picture - even when stretching the stars separately, this still needs more control, which I've achieved by using layer masks with specifically partially hiding the overly stretched star. As well, I've stretched the galaxy separately and feathered it in. For sharpening I tend to use two high pass filters in blend mode "overlay" for 8-10 and "light colour" (I think that's what it is at least, the one right underneath overlay) for 40-60 and feathered it in.

I think I'd like to try the difference in going for 10min subs rather than 5min next. Definitely avoid ISO1600 I'd say. And it definitely needs a light pollution filter (at least in Cardiff)...

Guiding was easy with a finder guider setup - there were 3 bright enough stars to choose from.

[SnakeyJ]

Very nice work Russe and nicely processed and presented - a lot of dedication to capture 10hrs 45 minutes of data!

[Astrosurf]

Did you line up the peaks of the three colour chanels in the Histogram? I find that helps.

[Russe]

What I do is, I threshold the pic looking for the black and the white point (but don't apply threshold to the image). There I put 2 (3x3px) colour markers which show how skewed (in colour terms) the image is. Then I level the image in each individual colour and bring the black point to 20-25 for each channel and should the image be underexposed, I'd bring the white in each colour to 253-255.

After each of the following 2 curves I check again with lvls and the histogram, whether one colour or the other starts to emerge too much and I correct this either with lvls or specific colour curves for areas that go into the wrong colour.

In the end I check the colour histogram, whether one colour is obviously out of sync.

[glowingturnip]

some interesting post proc techniques there - going to try that 2 layer high pass thing

[Russe]

some interesting post proc techniques there - going to try that 2 layer high pass thing

but do this only in the end - as well - don't do 100% opacity - play with that back and forth and partially hide areas you don't want sharpened that much - I do it for stars and galaxy separately

[glowingturnip]

yep - have been using the high-pass + overlay & transparency, but not seen the light colour bit

[Russe]

was it light colour? I just can't remember right now - but it's right underneath overlay. This takes the edge of the slightly harsher 40-60 high pass filter.

I've noticed as well, that when sharpening the star layer this way - the 40filter works better than the 10filter. Gotta play with it.

With the stars alone, obviously, I do match colour / saturation to bring out more colour.

[Astrosurf]

What I do is, I threshold the pic looking for the black and the white point (but don't apply threshold to the image). There I put 2 (3x3px) colour markers which show how skewed (in colour terms) the image is. Then I level the image in each individual colour and bring the black point to 20-25 for each channel and should the image be underexposed, I'd bring the white in each colour to 253-255.

After each of the following 2 curves I check again with lvls and the histogram, whether one colour or the other starts to emerge too much and I correct this either with lvls or specific colour curves for areas that go into the wrong colour.

In the end I check the colour histogram, whether one colour is obviously out of sync.

Eeek! Didn't get most of that! I'll shut up!

[Russe]

Eeek! Didn't get most of that! I'll shut up!

Lol!! If you like I'll put up a few screenshots?!

[Astrosurf]

Yes please! My PS ability is horrendously limited.

[Astrosurf]

Is this what you mean by black and white points?

http://myastroimages.com/Astro_Imaging_Tutorials/Learning_To_Use_PhotoShop_Curves/

I just use my eye with Curves. The above technique is much more accurate.

[Twisted Lip]

If its not too much to do screenshots that would be ace interested in how you PP'd it certainly.

Btw, it is a cracking image, you don't see too many M74's come up which shows how hard they are to get and how much time you have to put into getting them!

Will

[Russe]

Is this what you mean by black and white points?

http://myastroimages.com/Astro_Imaging_Tutorials/Learning_To_Use_PhotoShop_Curves/

I just use my eye with Curves. The above technique is much more accurate.

Similar. I'll show it in a bit what I mean.

[Russe]

If its not too much to do screenshots that would be ace interested in how you PP'd it certainly.

Btw, it is a cracking image, you don't see too many M74's come up which shows how hard they are to get and how much time you have to put into getting them!

Will

I won't be able to show the whole lot, since it turned into experimenting at one stage, but I can certainly show a couple of steps.

Keep in mind though - I'm a blumming beginner too!!!

[robbie c]

Hey Russe that's a great image well done 

[Russe]

First: the original stack straight out of DSS:

this then has been cropped to remove any extra areas (you could theoretically stack in a way that you get rid of the extra parts, but I like doing that by hand

so first - threshold - don't apply it - first move the slider all the way to the right (white - the picture will be black) and start bringing it down one by one until smth appears on the picture - that's where you put your first colour sampler (1 in the pic)

then you move the slider all the way to the right (the pic would be white now) and start moving it up until smth black appears - that's where you'll put colour sampler number 2

now cancel threshold and you are left with two colour sampling points for the whitest and blackest point in the pic

now make a copy of background and move the curves up a lot to bring up the outer part of the galaxy - there you'll place colour sampler nr. 3 - cancel these curves (I usually put colour sampler 4 in the center of the galaxy - but that's up to you...)

now you'd have 3 or four colour sampling points, allowing you full control over your changes

now look at colour sampler 2 in both pictures - change levels in each colour channel and bring them down to the same number (not to 0 though)

those are two stretches to separate the background from the dark parts of the galaxy

And now - any further curve stretches should be in each colour channel - anchor the background (shift + ctrl click on colour sampler) - put a 2nd below (otherwise it'll curve the curve downwards - choose a specific number on the curve to raise (I arbitrarily took 55 up to 75) and bring it down on a certain point (I took 168 for no reason) - and put one above for the same reason as the one below. Remember the numbers and do the same stretch in each colour.

And finally a levels after these curves - note background colour sampler 2 is all 23 (an acceptable background) and colour sampler 3 clearly brighter.

[Russe]

This I repeat several times and every 2-3 curves I work on a new layer (this will help greatly with star control in a tidbit).

[Russe]

Now I make my life easier by saving a curve stretch that I have clicked in for each colour channel - once you know you control the background - you could repeatedly do the same stretch.

Every two stretches I'd start a new background layer.

I think roundabout after 2 stretches like this I'd go for the gradient exterminator (no false advertisement...)

[Russe]

couple more stretches - take care with colour control (& galaxy is becoming noisy) - background will be controlled later

and now we're left with a couple of problems

a bloated star and white dots in the center of some of my stars (see marquee selection)