Sub exposure length with Ha 7nm filter

[BrendanC]

Hi all,

I've started doing some Ha imaging with my trusty little modded EOS1000D to create HaRGB images and getting some decent results (to my eye anyway).

However, I've been thinking about exposure times and histograms. I have a darks library that goes up to 300s sub lengths, and so I've been doing Ha subs at 300s. But, I've noticed that the histogram is very much packed to the left, obviously because of the less light let through by the filter.

So, I know a lot of this game is 'try it and see' but does anyone have a view on whether I could/should consider longer sub lengths when shooting Ha, to bring the histogram further to the right and improve SNR? And presumably if I do that, I'll also need to create a darks library for them (which would be painful, creating a darks library for each temperature at, say, 600s+)?

Also, is there any 'mathematical' way of estimating by how much the sub length should increase? I did once come across a page that said a 7nm filter lets through X amount less light so you need to increase your sub lengths by X to compensate, but have never been able to find it since.

Thanks, Brendan

 

Edited January 17, 2022 by BrendanC

[Budgie1]

If it's just the darks that you're concerned about, have you tried stacking your 300s Ha subs both with & without Darks to see what the difference is? 

I stopped using Darks with my DSLR and only calibrate with Flats, Dark-Flats & Bias frames because I couldn't see any difference in the final stack. That saves either taking the Darks during or at the end of a session or have to build a library for all temperatures.

I don't know how this would work when using an Ha filter on a DSLR, but may be worth an experiment, just to see?

[BrendanC]

Interesting idea, I'll try that, thanks!

I have tested without darks before and noticed a difference, which is why I use them now. But they're a real pain and I'd very much like not to have to use them. Looks like I'll have to do some testing.

[The Lazy Astronomer]

First up, you're sacrificing a lot of resolution by using a Ha filter with a colour camera, although I suspect you know and understand that already.

You should expose until your sky background level swamps the read noise of the sensor. What this actually means in terms of sub length depends on many factors, including: light pollution level, optics used, sensor QE, pixel size and read noise.

The formula I use for calculating optimum minimum exposure length is:

Exposure = C*rn^2/P

Where:

C is a factor applied to determine how much additional noise will be present in the image (I use a factor of 10, which equates to 5% extra noise)

rn is sensor read noise in electrons

P is the light pollution rate in electrons per pixel per second

All of the above is taken from a talk given a few years ago by Dr Robin Glover (Sharpcap creator). There is a video of it on YouTube.

He's also made a calculator for estimating your light pollution level: https://tools.sharpcap.co.uk/

[BrendanC]

Yep, I know I'm losing resolution. It's just a way of being able actually to do something during those periods when the skies are clear and the Moon is inevitably big and bright!

Thanks for this, and I think I've seen that Robin Glover presentation on YouTube. I went through a load of his stuff to arrive at 240s or 300s being optimal for my use case for RGB subs, but I'm just not sure how this affects Ha subs.

[BrendanC]

Just bumping this one gently to see if anyone can confirm that yes, I should really ideally be exposing Ha subs until the histogram reaches further across to the right, between 1/3 and 1/2 way across. I can see that it's going to be a balance between light capture and thermal noise, but all I need to know is, should I be doing this? I think the answer is yes. If someone could just say 'yes', I'd feel much better.

[The Lazy Astronomer]

Your calculated time for broadband seems a little long (it's in high read noise ccd territory), but if that is indeed correct, then yes, you'll likely benefit from exposing for longer in Ha.

[newbie alert]

What iso are you using, as you can up that as well as longer subs

With my 100d i did 10min subs, with I think iso 800

[ollypenrice]

You can, in my view, calculate till you are blue in the face but, until you've tried all the options, you will (quite rightly) wonder if you fed everything necessary into you equations.

Experiment!  Some of my own experimental results disagree with the relevant arguments from calculation.

Olly

[BrendanC]

I thought that would be the answer! As with everything in this maddening, wonderful hobby, 'it depends'!

Right, time to experiment...