Back to forums

Few good shots on few good nights ... where to optimize?

Nikon D810a Equatorial Mount Acquisition Deep Sky
7 replies•208 views
Ben avatar
Ben avatar
Hi all.

I was wondering if some of you here could help answer a question I get when I look at subs' details on images I like on this site: how do I best optimize the very few nights I get every year with my setup?

First, context on location and gear:
 - I live in Christchurch, New Zealand, and local weather makes it difficult to find a good stable imaging night when I'm available  (5 - 6 on a very good year).
 - I found a site 600m AMSL a one-hour drive away from home which puts me above frequent humidity over the Canterbury Plains, but is quite exposed to wind. I can minimize the effect on my setup via local cover, but if gusts get >10 - 15m/s it gets tricky …
 - my equatorial mount is a Skywatcher NEQ6 with annoying periodic hiccups in tracking, ruining a high proportion of shots (>66% if sub is >300", >40% if sub is <120"smile and I don't have the budget for another one. I modded the motors to be belt-driven, and the purchase of an ASI120Mini for parallel autoguiding helps in reducing but not eliminating these problems.
 - I shoot with a Nikon D810A and a telephoto lens (Sigma 120-300mm F2.8 or Nikon 600mm F4), with or without 1.4x or 1.7x teleconverter. I'm working on modding the 600mm to use 2" H-alpha and OIII filters I recently bought.
 - typically I would shoot with 180" or 300" subs, smile200ISO and the lens aperture a smidget less than fully open.
 - I use PixInsight and DxO PhotoLab for processing and editing but doubt I'm sufficiently patient or committed, or smart enough to fully understand and apply lengthy, technical, difficult processes. smile

Now to the question: considering what's above, could I switch from few usable smile00" subs to using loads of short subs (smile0" or even less?) and maximize the output of a given night without compromising on signal? Is a DSLR a hindrance to such technique, or the sensor type is irrelevant and it's down to night conditions and processing? And finally, will it take to sacrifice precious nights for trial and error based on guidance you may offer?

Anyway apologies for the wordy description, hopefully it's clear enough … smile
estabrook avatar
estabrook avatar
I'm shooting in Bortle 8/9 skies on an Alt-Az mount.  I usually have two suitable nights per month, but my sessions tend to be short–about 2 or perhaps 3 hours of imaging.   I've been using short exposures exclusively (usually 20 seconds) to beat light pollution and circumvent tracking limitations.  This month alone I've had what for me are great results on Helix, Cocoon, Pacman, and Wizard.  Hope this helps!
Well Written Insightful Concise
andrea tasselli avatar
andrea tasselli avatar
To answer your question: yes it would work and it doesn't depend on it being a DSLR or cooled CMOS/CCD camera. What matters is the read-out noise and the D810's is low enough to use 30s-60s exposures, especially if you use the lens fully open (or thereabout).
Well Written Helpful Insightful Concise
dkamen avatar
dkamen avatar
You would need more subs is all. The D810 is ISO-invariant meaning read noise is always the same (in other cameras higher ISO reduces read noise).

So SNR of a single sub is proportional to its duration and SNR of an integration of n subs is proportional to the square root of n.

For example, to get the same SNR as a 180 second sub you will need to integrate about 36 × 30 second subs (36 = square of 180/30). Or about 1100 seconds. That is ignoring everything else apart from read noise. In practice 36x30 second subs will be far superior to the single 180 second sub for many reasons (less thermal noise, better star shape, dithering, better dynamic range) 

So you need to experiment a little. A night of full moon or even partial clouds is a good opportunity as you want to check star shape,not do real imaging. You are looking for the maximum sub duration that gives you comparable SNR to a single 180 second sub within the same circa 350 seconds that it currently takes you to obtain it (since the other 170 go to a wasted sub). Provided of course 90% of the subs are keepers which is what you actually need to test.

Based on what you describe, about 70 seconds should do it and are in fact the optimal sub duration. But you might have to settle for less (and accept a longer total integration time) if your mount is too cranky or the wind too weird. Like if you lose 20% of subs in 70 seconds, go 60 or 50 or whatever gives you +90% keepers.

Cheers,
Dimitris
Helpful Insightful
Ben avatar
Ben avatar
Thanks for your inputs, much appreciated (and reassuring!). smile
One has to wonder how I never clicked muuuuch sooner to change how I image anything … go figure …
Ben avatar
Ben avatar
@dkamen What you told me above applies to RGB as well as narrowband, or only RGB?
dkamen avatar
dkamen avatar
Hi Ben,

yes but narrowband has so low SNR to begin with, particularly with a DSLR. People do 300 or 600 to get good results so 180 would be the short sub in that case.

Cheers,

Dimitris
Ben avatar
Ben avatar
Hi Dimitris, once again thank you for your advice smile