How long is long enough? - Forums

Topic starter

Mar 28, 2022 · Edited: Mar 28, 2022

I am just wondering how long I should spend imaging a single target. And before you say it, I know I am asking how long is a piece of string. I spent the last few nights capturing the Crab Nebula and could only get a couple of hours a night due to where it was in the sky and neighbours trees etc.. The clouds are now back for the next couple of weeks, so no chance to improve on it right now. My question is: do you think spending maybe 15 hours on the crab nebula would improve the image that much over my 6.75 hours image? FWIW: I'm using an 8" F4.5 newtonian, ASI1600MM camera and am imaging from my back garden in a suburban setting.

Here's my 6.75 hour image:

Crab Nebula SHO using a cheap newtonian

Engaging

D. Jung

Mar 28, 2022 · Edited: Mar 28, 2022

You need 4x times the signal to get double the snr. So if you want to double the SNR of your 6h image, you need to collect 24h of data and for the next doubling you need 100h….
If you account for variable conditions through different nights and you wanted to only go with comparable data you need to spend even more exposure.

Helpful Concise

Rafał Szwejkowski

IOTD Staff

Mar 28, 2022

Narrowband I'd go for 4h per channel as baseline. I think 12 hours is where I consider a SHO project mostly complete. Of course it's nice to shoot more if you really like the target and the intermediate results.

Helpful

Leonardo Landi

Mar 28, 2022 · Edited: Mar 28, 2022

In my opinion it depends on the subject you're shooting. As guideline, I assume one night per channel if I shoot narrowband, and 2 nights of luminance and 2 of RGB if I shoot LRGB. I'm in a suburban area too, my SQM is around 18.50. This is my baseline. If I like the target, or it is very faint ecc I can double the time .
I like to integrate cumulatively the single nights when I'm working on a project. When I have a sufficient quality of data, and can get rid of noise easily with MureDenoise, I stop acquiring data and start working on postprocessing.

Helpful