Hardware and software advice needed for upgrading to guided imaging

[Thalestris24]

Hi Ian

Well done for getting some guided images! I think the theory of ISO is that it's best used at the number which gives closest to unity gain. That value varies from camera to camera but I think it's generally around 800. Probably anything between 400 and 1600 will be ok. Above unity gain the signal is amplified but so is the noise. However, bear in mind that your calibration frames have to be at the same iso as your lights. In the case of darks they need to be at the same exposure length, iso and close to the same temperature. For that reason, when imaging with a dslr I stick to the same iso and certain exposure lengths. That way you can create darks libraries which you can re-use.

Louise

[ianpwilliams]

Thanks Louise, frankly I'm amazed that it worked at all, let alone as well as it did!

The thing with the ISO is that as I gradually increased from 1-min sub to 5-min sub, the luminance peak obviously became higher. So if I had stuck with 800ISO going up to 5-min subs then the luminance peak would have been way too high. So that suggests that if sticking with 800ISO (or even 400), I would have had to limit my sub lengths to 2 or even 1 min in order to keep the luminance peak below 50%, and that would have meant sacrificing longer exposures in order to have a "good" ISO, which surely can't be right?

[Thalestris24]

Thanks Louise, frankly I'm amazed that it worked at all, let alone as well as it did!

The thing with the ISO is that as I gradually increased from 1-min sub to 5-min sub, the luminance peak obviously became higher. So if I had stuck with 800ISO going up to 5-min subs then the luminance peak would have been way too high. So that suggests that if sticking with 800ISO (or even 400), I would have had to limit my sub lengths to 2 or even 1 min in order to keep the luminance peak below 50%, and that would have meant sacrificing longer exposures in order to have a "good" ISO, which surely can't be right?

Hi

FWIW I do all my dslr imaging at iso400  which maybe slightly sub optimal for my 1100d. I have the same problem as you - light pollution! I find a cls-ccd clip filter helps but lp is limiting. However, it's possible  to 'expose to the right' and compensate via post processing. There's no substitute for dark skies but I can only dream!

Louise

ps  once you load subs into dss they will appear much darker showing there is still a lot of room to stretch the stacked image

[ianpwilliams]

I went as low as 100ISO last night. Should inmaybe try 400ISO next time and accept a high luminance peak for a long sub?

[Thalestris24]

Hi

Maybe - try it and see! Load up some of your subs into dss and see what they look like. As I say, it's a good idea to find the optimal iso and stick to it - makes life simpler and easier.

Louise

[ianpwilliams]

But am I right in saying that you can theoretically use any ISO and you won't be in danger of making your subs unusable, it'll just be that they might be less optimal in terms of signal to noise ratio?

Incidentally I did another session last night trying M31. I actually managed 10 min subs, although one shot has a hint of star trail, which may have been caused because small clouds were occasionally passing through, so maybe PHD2 lost its star for a moment. So I don't know what will be usable, but still I'm very happy with 10 min subs.

[Thalestris24]

But am I right in saying that you can theoretically use any ISO and you won't be in danger of making your subs unusable, it'll just be that they might be less optimal in terms of signal to noise ratio?

Incidentally I did another session last night trying M31. I actually managed 10 min subs, although one shot has a hint of star trail, which may have been caused because small clouds were occasionally passing through, so maybe PHD2 lost its star for a moment. So I don't know what will be usable, but still I'm very happy with 10 min subs.

I think that depends on your sky conditions but, as mentioned, using a larger iso than necessary is counterproductive and isn't giving you any more photons which is what really counts, but is amplifying noise. Using a lower iso than necessary is like discarding photons so isn't desirable either! As mentioned, it's much better to ascertain the optimal iso for your camera and stick to it.

Louise

ps for the 1100d I think it's iso 800

[ianpwilliams]

I believe the optimal ISO for my camera and conditions is probably 400 or 800 (I can always experiment and find out during my next session). But I guess a 5-min sub (and certainly a 10-min) sub would be over exposed at either of those ISO settings.

So it would seem that I would have two choices:

1 - use the optimal ISO setting, and settle for shorter subs (but more of them I suppose) in order to keep the luminance peak below 50%

2 - use a very low ISO setting, which would reduce the photons, but have much longer 5 or 10 min (or even longer?) subs to counteract the lower ISO

The only question then is which option would give better results. Perhaps the results would be similar either way.

Now that I am guiding I am obviously keen on doing long subs if I can, because that's the whole point (after all, 2-min guided subs wouldn't feel like much of a step up from 1-min unguided subs). However, longer subs can be screwed up by satellites or planes passing through the shot, not to mention clouds, which I'm sure affected a 10-min sub or two last night.

[Thalestris24]

Hi Ian

There's no point in taking longer subs for the sake of it. By using a lower iso than necessary you are taking longer subs than would be required by the optimal iso - so nothing gained. It's better to take lots of the optimal subs and stack them. As I mentioned before, you can probably take longer subs with advantage if you have a decent lp filter.

Louise

[ianpwilliams]

OK next time I'll try 800 ISO. It may mean that I'll be limited to maybe 3-min subs, because I estimate that at that length the luminance peak will drop at around 60%-70%, which hopefully won't be a problem. And hopefully 3-min subs will still be long enough in comparison to my previous 1-min unguided subs to make the guiding upgrade worthwhile.

I can consider buying an LP filter later on. I'll also look into dithering at some point so that I can forget about darks.

[Thalestris24]

OK next time I'll try 800 ISO. It may mean that I'll be limited to maybe 3-min subs, because I estimate that at that length the luminance peak will drop at around 60%-70%, which hopefully won't be a problem. And hopefully 3-min subs will still be long enough in comparison to my previous 1-min unguided subs to make the guiding upgrade worthwhile.

I can consider buying an LP filter later on. I'll also look into dithering at some point so that I can forget about darks.

As I say, you'll have a better idea as to how far you can push the sub length when you stack and process. You may still want dark subtraction to get rid of hot pixels.

Louise

[ianpwilliams]

So would it be a case of testing how far I can go (over multiple sessions) by sticking with 800ISO, then letting the luminance peak go higher each session, until I end up with a stack that is impossible to process? It would be a shame to have to use multiple sessions to find out, because clear skies are unbelievably rare round here at the moment. But if it's the best way...

I assume that the luminance peak is being pushed higher by the light pollution and nothing else? In which case an LP filter would be worth researching?

[Thalestris24]

So would it be a case of testing how far I can go (over multiple sessions) by sticking with 800ISO, then letting the luminance peak go higher each session, until I end up with a stack that is impossible to process? It would be a shame to have to use multiple sessions to find out, because clear skies are unbelievably rare round here at the moment. But if it's the best way...

I assume that the luminance peak is being pushed higher by the light pollution and nothing else? In which case an LP filter would be worth researching?

It's almost certain that it's lp distorting your histogram peak and shifting it to the right. There's no other way but trying it to see what your max exposure length might be. You can try maybe 1 x 2min, 4min, 6min, 8min for a particular target to see what seems best. Just load the single subs into dss to see how far you can stretch them.

Louise

[ianpwilliams]

I did actually try that on my first session - 1 min, 2 min, 3 min, 4 min, and then 5 min (followed by another 9 x 5 min). When you say see how far I can stretch them, would there be any way to quantify that so that I can compare the results with each other?

[Thalestris24]

I did actually try that on my first session - 1 min, 2 min, 3 min, 4 min, and then 5 min (followed by another 9 x 5 min). When you say see how far I can stretch them, would there be any way to quantify that so that I can compare the results with each other?

It's hard to quantify - it's a matter of judgement and experience really. However, you should be able to see what a reasonable sub length for a particular target is. If you load a particular sub into dss and it gives an image that's too bright then you know you need to shorten the exposure time. I tend to take my lights in multiples of 60s e.g. 60, 120, 180, 360, 480, 600s depending on the target, camera and scope. So if a particular target seems properly exposed at 480s, I'll take X x 480s subs of that target. Once you become familiar with your kit you just know what will likely be ok.

Louise

[ianpwilliams]

I was planning on taking some bias frames tonight to add to my guided stacks, but this thread suggests that bias frames aren't actually needed:

http://stargazerslounge.com/topic/92031-how-to-take-bias-frames-on-a-dslr/

What do you think?

[ianpwilliams]

I've attached screengrabs of some of my subs from my guided Orion session, which include histograms. I did the following (with a rough guess at the percentage where the peak falls to zero):

60 seconds at 800 ISO - 75%

90 seconds at 800 ISO - 90%

120 seconds at 400 ISO - 75%

180 seconds at 200 ISO - 65%

240 seconds at 100 ISO - 50%

300 seconds at 100 ISO - 50%

So assuming I do Orion again, and I think I should do to see what I can manage at 800 ISO (or possibly 400 ISO), I'll need to decide what I can try to get away with. The above results suggest that at 800 ISO my subs really can't be that long (surely the drop at 90% for a 90-second sub can't be usable can it?), but maybe 120 seconds at 400 ISO could be usable, with a 75% drop?

I am considering buying a light pollution filter by the way, but until then it would be nice to get as many photons as I can whilst still having usable subs.

[ianpwilliams]

Incidentally here is my first attempt at processing my Orion stack (10 x 300 seconds at 100 ISO). I must mention though that this is my first attempt at using PixInsight for nearly a year (having only had the demo version before now, and having bought the full version), and also only knowing the very basics as to how to use Pix (I really need to get some Pix research done). And it was my first ever guided session. And yes the centre is definitely too bright.

It's actually only a little better than my previous, unguided processed stack (12 x 30 seconds at 800 ISO), although the guided version does show more of the outer parts. But I know I can do a lot better than this with a guided setup, both out in the garden and processing here indoors.

[ianpwilliams]

Anyway this thread was technically about advice for hardware and software for guided imaging, which has since gone off on many tangents! So now that I'm upgraded I'll leave this thread here and I'll start new threads for my other questions.

Thank you all for your help with this, it's very exciting to be starting a new chapter in my astrophotography journey!