To Pier 3 Members: Exposure Sugestion

Hi Shane

Pier 3 is a light bucket and too long exposures risk saturating pixels on many targets. The difference between 100 X 5 minute and exposures and 200 x 2.5 minute exposures in SNR is negligible. Take a look here and play with it to see how little difference it makes

https://snrcalc.vercel.app/home

Additionally longer subs expose you to more risk of failure - clouds, tracking, aircraft etc. You stand to lose less with shorter subs.

Undoubtedly it takes longer to process shorter exposures but only in initial calibration and stacking.

With the quality of modern processing software, CMOS sensors (which do not need long exposure like CCDs did) and the fact that p3 is an absolutely photon guzzler I would suggest that very long subs are not needed. Lots of subs are always good, and you can always add more.

Excuse my brevity I am on mobile.

Hello Pete, 

This is interesting and has some relevance to recent discussions on optimising our submissions. 

The app you refer to was partly trialled here at roboscopes by Brian Morgan the developer. I spent some time liaising with him during the testing and at the time learned a lot. Most of the discussion was via email and therefore absent in the post here on the forum. 

As an absolute novice when it comes to admin stuff I am unable/incapable of displaying the link to the post here. A search on "New online tool" will show you all those found, with the original post at the top.

This works brilliantly in theory, but as in life will only get you so far. Highly recommended for personal use as the majority of the parameters need only be set up once. On one's own setup you have far more control and therefore the results the app gives will be far more accurate. Understandably at roboscopes we cannot control the altitude when capturing data for each filter. We also have a more restricted range of exposure times to choose from to optimise it even more. etc 

What I did learn was that it was most important to establish the minimum exposure time per filter at an 'average altitude'  for the target. The app enabled that to be done and was a real eye opener. Whilst you are probably correct to say that you may gain less than you think by doubling exposure time, if you use too short an exposure you'll need far more integration compared to the optimal one. Also, as you'll see if you can find the time to read the post on the forum, a key element is using the surface area brightness using the recommended app, which is what I had to download on my computer. I think it was C2a. 

I need to stop here. 

The main things I took from that exercise was that, it's better to use an exposure time that is longer than the optimum, as using a shorter time can result in needing double the integration time. Demonstrated to me by Brian. The signal/noise ratio can end up being far too low. It's not just the low read noise of the camera you have to consider. There's a lot more to just using the optimal ratio of RGB subs, but that is a good part of it. 

The best I think we can do is make use of our own experience or first put in a shorter test submission for any job that will likely need many hours, or seek advice from our guru. Direct message or put in a help desk request to Steve. I know he'll thank me for that suggestion. :) 

Cheers and CS, 

Ray 

Thanks for the detailed reply of your thoughts mate. I am on the move so excuse my brevity. I think much of what you are saying can be attributed to CCD imaging. CMOS works in a different way, and in some cases shorter subs are actually beneficial vs. longer subs.

Take for example Lucky Imaging. It aptly demonstrates that better images can be created with tiny subs than longer ones, becasue you can try to beat the seeing. 1 second subs are actually a little long for this technique.

https://www.astrobin.com/418782/

https://www.astrobin.com/6cjy0l/

https://www.astrobin.com/3jurgg/B/ (11 hours or 1 and 2 second images!!!)

Extreme example, but it shows what is possible. Ultimately what is important is that you have enough data, not enough 5 min sub frames.

Saying that if you try to process lucky imaging with 1 second subs your processing PC will heat up the atmosphere to the point you will worsen the seeing for everyone else :) So there's a compromise to be had.

As I say I'm out and about so please excuse the short response.

Peter,

Those are great examples. Carsten Dosche and Stephane Gonzales (Exaxe) are others I've looked at after I had watched a video posted a year ago on the subject. The image of M57 by Stephane is really amazing given its short integration time.

You've already pointed out one downside when processing the data. There is of course, the overhead on imaging time. I know that you are not proposing going that route, I hope. 

Thinking back to early 2021 when liaising with Brian, of the 2 piers I know we used, pier 1 still used a ccd with the 8300 chip at 2x2 bin. I believe it wasn't swapped to a cmos until a few months later. The other pier, pier 14 had a recently installed cmos 6200 sensor. So yes, the reference to needing a minimum amount of imaging time may possibly have been about pier 1. My knowledge of ccd and cmos in those early days wouldn't have led me to question it.

Moving away from discussing RGB targets which is what that app was designed for, I still believe that on those dimmer emission narrowband targets it's better to stick to the exposure times advice on the equipment pages. Dimmer needs longer. Sounds a bit like me, as often it takes me longer to understand things. :) 

Thanks for the links to those other 'lucky imagers', I'll have a peek later. 

Cheers and CS, 

Ray