Question on maximum exposure time

[Galaxyfaraway]

I live in a relatively light-polluted area (suburbs of London) and am trying to work out the maximum time for exposures I can get in my area. 

I took the first image attached is without an LP filter, at 30 secs and ISO3200 and the 2nd image I took yesterday, when I got the LP filter (IDAS D1), at 101 secs ISO4000 and I think I actually got more detail from the 1st image. I can't seem to pull out more from the 2nd image after stacking and level adjustments etc. Both seem pretty washed out though. There was some moon light yesterday towards the end so perhaps it would make all the difference to wait till there is no moon at all?

How should the histogram look to get a successful exposure and am I at a dead end with regards to the light pollution and maximum exposure times?

I thought with an LP filter I should be able to increase exposures by quite a bit.

Any advice? thanks,

gfa

[laser_jock99]

A general LP filter like the IDAS is good for wavelength specific light pollution, it won't help against a broad band sources like moonlight and the ever increasing white LED street light menace. LP filters are fast becoming less effective as the LP spectrum broadens.

Regarding the exposure lengths- that is just something you'll have to experiment with to ascertain where sky glow starts to wash out real signal.

Also your ISO settings might be a bit high? Most people stick to around 800-1000 to keep camera induced noise down. 

[Galaxyfaraway]

Is there not a way to tell from the histogram in the camera whether a photo is overexposed for stacking purposes? I do it to check for normal photography but with astro, I'm not certain how washed out it's supposed to be until the signal disappears.

I read somewhere the reading is supposed to be in the first left quarter of the histogram. I think mine are all towards the right.

Mine is a full frame camera Nikon d750 and ISO noise handling is usually some of the best there is so I am trying to push the iSO as high as possible so that I can do shorter exposures.

[Galaxyfaraway]

Another thing: I can't really stay up all night to take hours of data and can do perhaps a couple of hours at a time. How do I go about stacking & combining data taken from various sessions? Do I stack each session from each day separately and then somehow combine them or can I just take all the subs and stack them as if taken from one session?

Similarly if i need to change settings (for example to get the core of M31 separately). I am still not quite clear about this.

thanks, gfa

[Joel Shepherd]

I'm no expert in Photoshop and even less so in teasing detail out of RGB, but my crude attempts pulled a lot more out:

That was done simply by pulling up the dark levels on each of the RGB and luminance channels individually. I'm sure you can get much better results with some practice but there is some useful data there to work with.

If you poke around the Imaging forums here, there are some extensive tutorials on using Photoshop for this sort of work: plenty of stuff on the web too. Set aside a few hours to start working through some of them and I'll bet you'll be pleased with the results.

(This doesn't directly answer your question, except to say that 30 seconds is probably not your upper limit.)

[Galaxyfaraway]

Ah thanks. I posted the two lights just to check whether they are ok in their original form (unstacked, un-levelled, unprocessed etc) or over-exposed because if I am not setting my camera settings correctly, there is no point taking hours of over-exposed lights. I don't know what people's light raw files look like (you only see the finished photos, mostly) so I wasn't certain whether this is anywhere near the right ball park.

I am not asking how to process the images (yet). Just want to find out how to take proper subs and what to expect realistically in terms of exposure length, given where I live.

[Galaxyfaraway]

To answer myself, I just found this very helpful tutorial on addessing the multiple sessions issue here: 

[steviemac500]

I thought that you should aim for between 20-40% on the histogram on the camera.

[Joel Shepherd]

Got it. Typically, raw lights aren't going to be very inspiring, so don't let their bland appearance throw you too much. As Steviemac500 suggested, they are on the bright/white side. I'd probably experiment with lower ISOs and longer exposures, both of which will help improve your signal to noise ratio. As long as you don't go so low to get hard black pixels (no light whatsoever) you should be able to stretch the data into something worthwhile.

[KyleStoke]

I would personally lower the ISO and take as many subs as possible, not sure if you have a guided setup etc but i would set it to say ISO 800 then take subs for the maximum amount of time before any trailing becomes apparent then as many of them as your time allows, stack them all in DSS with calibration frames and see where that leaves you.

[wxsatuser]

This is a good spot for the histogram with Canons

If your Nikon just get at least 25% of the histogram, that will give you a starting point.

Drop your ISO to 800 and no more than 1600.

With the LP filter you should do at least 2 to 3minutes if not more.

[dph1nm]

Some things to note:

* If you have a lot of light pollution then you don't need long exposures - once the sky noise dominates the camera read noise you don't gain anything by taking longer subs. This happens pretty quickly if you have LP! In fact longer exposures are worse, as you just saturate all the stars.

* ISO doesn't change the camera's sensitivity to light, so the total exposure doesn't need to change when you change ISO. However, lower ISO will cut down the amount of saturation you get in your subs (it won't, however, change the exposure length needed to overcome read noise - unless you have a Canon, as they have rubbish read noise at low ISO! Nikons should be OK).

NigelM

[pipnina]

There is much to be gotten from these RAW images. I got this from just the JPEG: http://i.imgur.com/x66wABF.pngthe RAW would look much better.

So I'd guess that your current level of exposure is fine. In fact if you go down to 1600/800 ISO and double/tripple the exposure you'd get much more. Try taking one image of each and see what end up best on your computer.

[Galaxyfaraway]

Many thanks - exactly the responses I needed...

I did attempt stacking and posting the batch from the first file in this thread and got a bit of data out of it: http://stargazerslounge.com/topic/253287-first-go-at-andromeda/?hl=%2Bgfa+%2Bfirst+%2Bgo#entry2762618

But didn't post the raw stacked Tiff, which is here (from the first session):

https://www.dropbox.com/s/sudgr1a76yxemlb/Andromeda%202%203x%20drizzle%2016%20bits.TIF?dl=0

You can probably do much better in terms of extracting detail?? (Although I tried.)

And with LP filter from another session (is it ok to combine two such different sessions into one: one without LP filter and one with LP?):

https://www.dropbox.com/s/z5xp7wm7irqjd68/Autosave.tif?dl=0

Last question (doubt it, but hope springs eternal!):

would taking shorter exposures, but hours, days or weeks of them, give me the same detail as taking longer exposures? (but less of them?). So for example would 10x10 minutes subs give me the same detail as 200 x 2 minutes or similar? Or does it not work that way?

[pipnina]

Many thanks - exactly the responses I needed...

I did attempt stacking and posting the batch from the first file in this thread and got a bit of data out of it: http://stargazerslounge.com/topic/253287-first-go-at-andromeda/?hl=%2Bgfa+%2Bfirst+%2Bgo#entry2762618

But didn't post the raw stacked Tiff, which is here (from the first session):

https://www.dropbox.com/s/sudgr1a76yxemlb/Andromeda%202%203x%20drizzle%2016%20bits.TIF?dl=0

You can probably do much better in terms of extracting detail?? (Although I tried.)

And with LP filter from another session (is it ok to combine two such different sessions into one: one without LP filter and one with LP?):

https://www.dropbox.com/s/z5xp7wm7irqjd68/Autosave.tif?dl=0

Last question (doubt it, but hope springs eternal!):

would taking shorter exposures, but hours, days or weeks of them, give me the same detail as taking longer exposures? (but less of them?). So for example would 10x10 minutes subs give me the same detail as 200 x 2 minutes or similar? Or does it not work that way?

I think there was another thread a while back about whether 30 1 second exposures or one 30 second were better. I think single 30s won by a fair way. (much less background noise).

[Vox45]

I think there was another thread a while back about whether 30 1 second exposures or one 30 second were better. I think single 30s won by a fair way. (much less background noise).

There it is

[ollypenrice]

I find that the shooting of multiple exposure lengths is only very rarely necessary. M42, certainly. Maybe the shallower well depth of DSLRs might make it more useful than it is for me but the core of Andromeda is a real challenge and I didn't find any more useful detail in short ones than in long ones. On the other hand, multiple stretches of the same data, combined in Layers, can be very effective.

This will take you through what I think is the best way to process M42 and multiple exposures. http://www.astropix.com/HTML/J_DIGIT/LAYMASK.HTM

Olly