exposure time and LP

[planetdnb]

hi all,

I have recently had a couple of attempts at m33 my first go was 20 subs at 2 mins. second go was 12 subs at 10 mins.

I processed each session in the same way and the results were almost identical - both pretty lame.

what I want to know is, is exposure time dictated by LP?

if my results for 10 min subs are the same as 2 mins do I need a LPF in order to get more detail?

I live in a village between 2 towns. skies are fairly dark but still some LP.

thanks

[adamsp123]

LP will have some affect on max exposure, but it may be that M33 is a right swine of a subject to get a decent result especially with a DSLR I have found, try some trial runs on a different subject M1 maybe or the Horsehead Nebula.

I personally have tried various LP filters with my DSLR but recently I didn't bother with any filtering at all to see how it performs....BTW it is modded so it has no IR filtering at all... results here...

http://stargazerslounge.com/imaging-deep-sky/120071-quick-orion-running-man.html

http://stargazerslounge.com/imaging-deep-sky/120366-quick-horsehead.html

[planetdnb]

should i not expect to see better results from 2 hours of 10 min subs than 40 mins of 2 min subs? In fact if I am correct I didn't actually get the 40 mins of 2 min subs as the fog rolled in. It's possibly more like 20 mins.

[adamsp123]

How I understand it, the more subs the less noise, the longer the sub more detail can be gained, however other factors come into play with long subs, LP, guiding/tracking errors etc, try and go for a longer a sub as practical with these things in mind.

And yes I would expect 2hrs of 10 mins would be much better than 40 mins of 2 min subs, unless LP swamps the data.

[MartinB]

This is complex territory and the more you look into it the more confusing it becomes. Here is a very simple crude method.

Expose until the background brightness is about 1/30 full saturation - aound 8-10/256 in PS.

That's all there is to it but if you want to delve further below is a more detailed explanation but is still a bit of an over simplification! Don't worry about all this stuff if you are just starting out, just stick with the above.

Your images will contain 2 kinds of noise, shot noise and read noise. Read noise is a constant amount of noise present in each exposure. It doesn't increase with longer exposures. Shot noise increases with length of exposure.

Shot noise is removed by stacking. To all intents and purposes read noise isn't removed by stacking.

Signal is the stuff we want. The goal is for the signal to predominate over the noise. Signal increases with exposure time and, fortunately this rate of increase is faster than the rate of increase in the shot noise. So, signal to noise ratio increases with the length of the sub. In fact it increases by the sqaure root of the increase in exposure time. If you want to double the signal to noise ratio of a 1 hour sub you would need to expose for 4 hours (ignoring read noise and light pollution etc)

The longer your exposure the smaller the proportion of noise from read noise but after stacking the read noise will still be there.

Think of LP as a form of constant noise. Unlike shot noise it increases at the same rate as the signal. Therefore it has a constant effect on the signal to noise ratio. It doesn't matter what your exposure length is. Until you start to saturate your chip LP doesn't flood out our signal so you could expose as long as you want provided you stay well clear of a saturated chip. However there are disadvantages of long exposures - subs are more prone to satellite trails. passing clouds and episodes of poor tracking. Also stacking works best the more subs you have. So on this basis short subs are good.

However, short subs contain a higher proportion of read noise which isn't removed by stacking and this is bad! What we need to do is make the read noise such a low proportion of overall noise that it really doesn't matter any more. This point is achieved when the LP effectively washes out the read noise (LP will wash out read noise since read noise doesn't increase with exposure time). Once you have reached this point you might as well stop your exposure and get lots more of the same length. It isn't that the long subs are detrimental, just that long subs are vulnerable to being messed up and you don't have as many to stack.

How do you know when the read noise is being drowned out by the LP. There are calculators which you can use if you can measure your sky glow and other parameters such as dark current, camera gain and full well depth etc. However, in practice you can get close by getting the background to about 1/30 of your saturation level. This will be around 10/256 in PS. Obviously the time of exposure required to achieve this will depend on your LP, the air conditions and the declination of the target.

BTW 8 bit can show 256 levels of brightness (hence the 256 in PS although 255 is more correct because 0 counts!), 4096 levels for 12 bit and 65536 for 16 bit. When using a 16bit CCD chip and software that supports 16bit you would aim for 2000 ADUs. Transferred into PS using a lossless 16bit format would then give you a level of around 10

[dph1nm]

The effectiveness of an LP filter will depend on the ratio of sky to object. If you already have a quite a dark sky you may actually make things worse by using an LP filter.

It won't make any difference between the 2min and 10min stacks, unless the 10min stack is saturated (but that would be obvious just by looking at it). All that should matter is the total exposure time which went into each stack.

NigelM