Are my lights too light?

[StuartT]

I'm shooting a pile of lights of M27 this evening, but I am worried the sky is way too bright. Surely if the sky is a similar brightness to the actual target, then no amount of stretching is going to separate them? Or have I got that wrong? These are shot at 60sec ISO1600 with an Astronomik CLS filter. The moon is still only 10% and has practically set (and these images are East in any case)

 

Edited June 15, 2021 by StuartT

[Davey-T]

Can't ope jfif files can you load a PNG ?

Dave

[StuartT]

7 hours ago, Davey-T said:

Can't ope jfif files can you load a PNG ?

Dave

Hi. Here is the original RAW file and a PNG version too. Thanks

L_9428_ISO1600_60s__29C.CR2

[Stuart1971]

They look fine to me, I did this target a while back, and my subs were a bit brighter than yours, and this was my resulting image, it’s from an Unmodded canon 1000d DSLR and an 8” Meade SCT..also I was using an Astronimik CLS LP filter so my subs had the same green / Blue tinge to them…..

 

Edited June 16, 2021 by Stuart1971

[david_taurus83]

That's just the colour cast from the filter. You can sort it out in processing by balancing the RGB channels. When you stack them all together they won't come out that bright anyway.

On a side note, 29°C is a bit on the high side for a DSLR. Are you making sure the screen is off and flipped out if applicable?

[StuartT]

Thanks both. It wasn't so much the colour of the sky that was worrying me, but just the sheer brightness of it. If stretching lowers the background brightness of the sky, will it not also lower the brightness of the nebula too? In other words, how would photoshop 'know' I only want to darken the sky?

35 minutes ago, david_taurus83 said:

That's just the colour cast from the filter. You can sort it out in processing by balancing the RGB channels. When you stack them all together they won't come out that bright anyway.

On a side note, 29°C is a bit on the high side for a DSLR. Are you making sure the screen is off and flipped out if applicable?

David, I am not sure what you mean by 29C... ? where are you getting that from? Although it was a very warm day, it had cooled down quite a bit by midnight

Edited June 16, 2021 by StuartT

[Stuart1971]

28 minutes ago, StuartT said:

Thanks both. It wasn't so much the colour of the sky that was worrying me, but just the sheer brightness of it. If stretching lowers the background brightness of the sky, will it not also lower the brightness of the nebula too? In other words, how would photoshop 'know' I only want to darken the sky?

When you stretch an Astro photo, you only stretch the areas you want, so in your case you would stretch the light areas, such as the nebula, and not the dark areas like sky background….you will get a curve in PS stretch tool, and you must alter the part of the curve that relates to the nebula…

[david_taurus83]

36 minutes ago, StuartT said:

David, I am not sure what you mean by 29C... ? where are you getting that from? Although it was a very warm day, it had cooled down quite a bit by midnight

29C is in the filename of the CR2 file. That's the recorded temperature of the camera at the time of capture. I found that above 20°C with a Canon DSLR that it starts to get noisy. You may need to take darks to try and remove thermal noise.

[ollypenrice]

6 hours ago, StuartT said:

In other words, how would photoshop 'know' I only want to darken the sky?

 

Because you'll tell it! 😁 Here's how.

1) Your background, once colour balanced in Pixinsight's DBE, came out at a Photoshop brightness value of 11 to 12, so certainly not too light. I like a background sky of 22 or 23 per channel. This is unstretched. Alternatively, in Curves, if you alt-click  Ctrl-click with the cursor on the background it will show you your present background value:

 

2) Put 4 colour sampler tool markers on different parts of the background, set to sample at 5x5 average (top toolbar) and stretch in levels till your background comes up to 22-23 or so. If you look at the four colour sampler boxes on the upper right of the screen grab you can see the input and output values, so the first marker  has gone from Red 11 to Red 20 in this case. G has gone from 12 to 22... and so on. Now you can stretch in Levels beyond this point and just bring in the black point to clip the background back to 22 but, at some point, this will overstretch your background and make it noisy. This is when you go to step 3, though I went there directly from here for the sake of this demo.

 

3) Open curves, alt-click on the background to make a fixing point on the curve. Put a second fixing point below that to avoid changing the background at all. Now lift the curve above the background in a series of gentle iterations, experimenting with the shape of the curve you make. Essentially you want to stretch harder just above the background than you do higher up, where you risk stretching the stars too much.

 

I think your initially posted image included a screen stretch or 'visualization' even though you hadn't asked for one. You managed to get a remarkably gradient-free background sky!

Olly

Edited June 16, 2021 by ollypenrice
correction

[StuartT]

1 hour ago, david_taurus83 said:

29C is in the filename of the CR2 file. That's the recorded temperature of the camera at the time of capture. I found that above 20°C with a Canon DSLR that it starts to get noisy. You may need to take darks to try and remove thermal noise.

Ah! I didn't realise that. I think APT must have kindly included that info in the filename. I did shoot darks, flats and bias frames so maybe it will all come out in the wash. 

But is there any way I can stop the sensor heating up like that in a DSLR? When I am shooting, I fold the LCD screen inwards which I think switches it off. I do this to save battery life.

 

20 minutes ago, ollypenrice said:

Because you'll tell it! 😁 Here's how.

1) Your background, once colour balanced in Pixinsight's DBE, came out at a Photoshop brightness value of 11 to 12, so certainly not too light. I like a background sky of 22 or 23 per channel. This is unstretched. Alternatively, in Curves, if you alt-click with the cursor on the background it will show you your present background value:

 

2) Put 4 colour sampler tool markers on different parts of the background, set to sample at 5x5 average (top toolbar) and stretch in levels till your background comes up to 22-23 or so. If you look at the four colour sampler boxes on the upper right of the screen grab you can see the input and output values, so the first marker  has gone from Red 11 to Red 20 in this case. G has gone from 12 to 22... and so on. Now you can stretch in Levels beyond this point and just bring in the black point to clip the background back to 22 but, at some point, this will overstretch your background and make it noisy. This is when you go to step 3, though I went there directly from here for the sake of this demo.

 

3) Open curves, alt-click on the background to make a fixing point on the curve. Put a second fixing point below that to avoid changing the background at all. Now lift the curve above the background in a series of gentle iterations, experimenting with the shape of the curve you make. Essentially you want to stretch harder just above the background than you do higher up, where you risk stretching the stars too much.

 

I think your initially posted image included a screen stretch or 'visualization' even though you hadn't asked for one. You managed to get a remarkably gradient-free background sky!

Olly

Thanks for this comprehensive explanation (though unfortunately, I can't follow what you are doing). It's good to know my sky ain't too light 🙂

My initial image wasn't stretched in any way. It's just what the camera captured

 

Edited June 16, 2021 by StuartT

[david_taurus83]

On the 600D you can flip the screen out and switch it off. Tjis helps it cool down. Can't remember exactly but pressing the menu button on my 6D cycles until the screen is off.

[ollypenrice]

1 hour ago, StuartT said:

Ah! I didn't realise that. I think APT must have kindly included that info in the filename. I did shoot darks, flats and bias frames so maybe it will all come out in the wash. 

But is there any way I can stop the sensor heating up like that in a DSLR? When I am shooting, I fold the LCD screen inwards which I think switches it off. I do this to save battery life.

 

Thanks for this comprehensive explanation (though unfortunately, I can't follow what you are doing). It's good to know my sky ain't too light 🙂

My initial image wasn't stretched in any way. It's just what the camera captured

 

Maybe check out 'Levels and Curves' via Google since these are the key processing tools. It's easier to show these steps in videos, of course.

If there was no stretch the bright background just comes from the unbalanced colour channels. Not a big deal. You can see this in Photoshop if you look at the histogram of the image you first posted:

First understand the graphs, though. The x axis goes from dark on the left to bright on the right. The vertical axis shows the number distribution of pixels at a particular brightness. Looking at green and blue, and starting at the left, we begin with nothing. There is no graph because the entire histogram is too far to the right. In Levels you need to move the left hand (black point) slider to the right, initially just to the start of the graph.

The first little bit of thin black line is the background sky. The main pedestal is the nebula. The thin line after the main pedestal is just the stars.

To balance the colours you need to get the top left of each pedestal (shown by red lines) to more or less the same distance from the left. You can only move the pedestals one way - to the left - using the black point slider in Levels. Unfortunately red is already a bit far left so longer exposures might fix that. Don't worry about the sky being too bright. Once balanced it certainly isn't. It's about half the brightness it needs to be at the end. The low red signal probably comes from your filter but it's done well against LP.

 

Then keep on adjusting the black point in levels till your histogram peaks align. A neutral sky, equally bright in each channel, is always my starting point for an image processing job.

Olly

Edited June 16, 2021 by ollypenrice
clarification