Quick question about exposure length

[Marky1973]

After getting over the shock of a clear night and no moon, I've rushed out and got everything set up and the kit is happily chugging away in the garden....

So, set it up doing 120 second exposures at 800ISO as that had the histogram exposing to about 1/3 - probably a bit more than it is supposed to be but it feels slightly counter-intuitive to me to not be going for the longest exposures I can manage.....so, the question is, with my DSLR, is it always better to be going for an exposure length that gets the histogram in the right place or the longest stable exposure you can get?

I'm sticking to shorter exposures tonight as my alignment was fraught with RA setting ring problems that didn't seem to be right when trying to polar align, so it probably isn't as accurate as usual.....but that got me thinking about exposure length and the histogram....and whether I should be going for as long as possible or not....2 minutes just feels short to me.....could manage that on my old EQ3-2 without guiding! ;-)

Probably rambling now.....sorry for the daft question.....

[Pompey Monkey]

I tried for some long exposures tonight. And it it has all gone pear-shaped so far, despite having a tonne of concrete under my mount! ...

[D4N]

It depends on whether you are over exposing your stars by increasing the exposure length. I found a UHC-S filter allowed me to expose for longer without blowing out highlights.

/Dan

Sent from my iPad using Tapatalk

[cjdawson]

Don't forget that Astro photos are pictures that contain mostly dark grey - space.

As D4N said.  The point is not to overexpose the stars themselves.   Ideally the histogram would show data over the whole range of an image.  In the case of Astro photos, that means lots of dark colours hence in the bottom 1/3. and a little spread out over "almost" then entire histogram.

The brightest stars would have data that is almost right at the top of the range, and the deepest black would have data right at the bottom.

Underexposure means that you miss data as it's off the bottom of the scale and so not picked up.

Overexposure means that you have captured so much that it's gone off the top of the scale.

I very much doudbt that nature has anything where a star is 255,255,255 in it's RGB colour. Or 0,0,0 for that matter, so if you have these two colours in your image, you know that it's either under of over exposed.

The trick is to get as close as you can to both ends without touching.

Stacking images doesn't increase the range of the image, what it does is allow you to use several images to increase the noise to signal ratio.  The more signal that you get the better the image will look.

[Marky1973]

So is it more important to get the hours in and stack photos of the longest possible/sensible exposure for your skies.....obviously 20 minutes exposures with a DSLR during a full moon is going to be a bit daft) as long as you aren't clipping the histogram at either end? That's how my brain works......

Now, if I could just have more than one night every 6 weeks to practice.......

Cheers

Mark

[cjdawson]

Marky1973, that's how I understand it.  Also, just because you take the images about 6 months apart from each other, doesn't mean that you can't use them in the same stack.

The idea is that if you are not clipping either end, then you can stretch the histogram.  The closer that you can get to filling it completely the less that you'll have to stretch the image.  Also the further up the scale that the main block of the data is, the lighter the general image is.  Depending on the object, you might want a very light image - for example M31 where the image fills most of the view.    But for something like M57, which is a small object, you'll want a darker overall image as the image will be mostly star-field rather than object.

[Marky1973]

Cheers - that's kind of what I thought as well but, by nature, I can be a bit "by the book" sometimes and worry if I am not following strict rules with stuff like this....time for more experimentation I guess!

[kendg]

When I went to a seminar on AP they said always expose to the right. It was quite amazing what could be pulled out from a raw file that looked to be way over exposed.

[wimvb]

When I went to a seminar on AP they said always expose to the right. It was quite amazing what could be pulled out from a raw file that looked to be way over exposed.

This makes sense. You can Always subtract a constant (or gradient) background to get a wider range of your data, which you then can stretch. But you can't subtract noise from noise. Better images tend to be on the bright side, but have lower noise. I.e. high SNR even if the S (signal) part may contain quite some background.

[ollypenrice]

I don't image with a DSLR but with deep welled CCD cameras. However, I think that the concerns about over exposing are frequently overstated. From a dark site and with cooled cameras I find that very long exposures work well and that not many targets over expose easily. I suspect that light pollution and thermal noise are the main exposure limiters for DSLR users. If a target is not over exposed in the linear image then it isn't over exposed. It just needs the right processing.

In CCD from a dark site I'm not averse to using 30 minute luminance subs. I found that this was the only way to find the outer glow around M31, for instance. https://ollypenrice.smugmug.com/Other/Best-of-Les-Granges/i-xbvjFDF/0/X3/M31%20Outer%20HaloLHE-X3.jpg

In my view nothing beats experimentation. It's amazing how easily orthodoxy sets in. I wouldn't worry about challenging it!

lly

[Marky1973]

I don't image with a DSLR but with deep welled CCD cameras. However, I think that the concerns about over exposing are frequently overstated. From a dark site and with cooled cameras I find that very long exposures work well and that not many targets over expose easily. I suspect that light pollution and thermal noise are the main exposure limiters for DSLR users. If a target is not over exposed in the linear image then it isn't over exposed. It just needs the right processing.

In CCD from a dark site I'm not averse to using 30 minute luminance subs. I found that this was the only way to find the outer glow around M31, for instance. https://ollypenrice.smugmug.com/Other/Best-of-Les-Granges/i-xbvjFDF/0/X3/M31%20Outer%20HaloLHE-X3.jpg

In my view nothing beats experimentation. It's amazing how easily orthodoxy sets in. I wouldn't worry about challenging it!

lly

Lovely photo Olly!

I'm definitely going to challenge it - I had a crack at Bode's/Cigar on Tuesday night - just shy of 3 hours of 120 second subs and I have struggled to pull much more out of it than I did a year ago with less than an hour of 60 second exposures....I admit my processing still needs work but I thought I should be able to see more, even with my basic abilities....

Onwards!