Total Intergration time

[FenderGreg]

Hi,

is there some kind of maths into how much intergration/subs makes a difference?

I originally did 2 hours with 10min subs on the cone nebula and it came out looking great. But I thought I’d go for it again and see how much more detail I could bring out.

I added another 2 hours 20mins, again with 10min subs but after a quick process I see almost no difference

[michael.h.f.wilkinson]

Theoretically, to double the signal to noise ratio (S/N), you need to quadruple integration time, so doubling the integration time should give a modest increase in S/N, in the order of 40% improvement. This means you can sharpen a bit more, without noise exploding, but not much. Mathematically S/N scales with the square root of integration time.  This is one reason people may go for extremely long integration. Also note that the sky background (mathematically also just signal) also adds to the noise in the same way, so picking out a faint signal against a bright background can be very hard indeed, as the noise from the object of interest may be swamped by the noise from the sky background.

[FenderGreg]

1 hour ago, michael.h.f.wilkinson said:

Theoretically, to double the signal to noise ratio (S/N), you need to quadruple integration time, so doubling the integration time should give a modest increase in S/N, in the order of 40% improvement. This means you can sharpen a bit more, without noise exploding, but not much. Mathematically S/N scales with the square root of integration time.  This is one reason people may go for extremely long integration. Also note that the sky background (mathematically also just signal) also adds to the noise in the same way, so picking out a faint signal against a bright background can be very hard indeed, as the noise from the object of interest may be swamped by the noise from the sky background.

Thanks for the reply. That does kinda make sense.

so if I’ve done 2 hours I might not notice much difference until 8?

What about if the subs were shorter thus increasing the volume? Same rule really due to overall time?

[michael.h.f.wilkinson]

13 minutes ago, FenderGreg said:

Thanks for the reply. That does kinda make sense.

so if I’ve done 2 hours I might not notice much difference until 8?

What about if the subs were shorter thus increasing the volume? Same rule really due to overall time?

Total integration time is the main issue. More, shorter subs would mean a (marginal) increase in noise due to more read-out noise, but on modern cameras that is hardly noticeable. Going to a darker site would help much more. Faster optics also help (for a given FOV). Here's an integration of 12 h at F/2 using a Samyang 135 mm F/2 lens, Canon 550D (modded), and Optolong L-eNhance filter.

The L-eNhance filter helps reduce background noise due to my Bortle 4-5 skies. The high speed of the optics help grab loads of light.

[FenderGreg]

These were my two. First pic 2 hours, second pic 4 hours 20mins.

Would have loved to bump this to 8 hours like you mentioned but the weather has been terrible since 💩

[ollypenrice]

A noise reduction tool as powerful as Russ Croman's Noise Xterminator, which does minimal damage to resolution, is worth - and now I'm sucking my finger and holding it up in the wind - a trebling of integration time. Worth a try...

Olly