Exposure time

[Miguel1983]

Hi,

Since august 2018 i've started imaging, for now i always use my ED80 for deepsky and my Mak127 for lunar.

But that's not wat this is about, this just shows i'm still quite new to this and mostly it's trial and error as i go along.

I see some of you take 600sec subs, and i wonder is this to be preferred over say 2x300 sec exposures instead ?

Taking 600 sec subs is no problem, i got my tracking technique down, but isn't there a point up on witch there's nothing more to gain except noise ?

I usually don't go further than 300 sec, and make sure my histogram peaks (in Backyard EOS) are not in the third overexposed section.

Surely filters are a big factor in this, i know.

Thanks

[wxsatuser]

Once the histogram is detached from the lefthand side it should be ok whatever the sub length.

It does depend on your sky brightness and if an LP filter is in use.

Obviously the more subs the better the SNR will be.

 

[vlaiv]

It is always better to use less longer subs than more shorter subs if they add to same total imaging time.

This is due to read noise. All other noise sources depend on time (shot, LP, thermal), but only read noise depends on number of exposures.

However this dependence is not trivial one. Once read noise becomes much smaller than any other noise source - difference between SNR of short exposure stack and long exposure stack becomes also very small. On the other hand, there are things associated with long exposures that hurt your data collection. Saturation / clipping in bright areas is one of those things. Need for precise guiding is another. You increase likelihood of dropped frames due to excessive wind, cloud, airplanes passing by, whatever might happen - longer the sub, greater the chance for something to go wrong, and if it does - more imaging time wasted if you end up discarding that frame.

So what would be good sub length where you enter "diminishing returns" part? It really depends on your camera (CCD has higher read noise - therefore it usually needs longer subs), your conditions (Heavy LP will very quickly build LP related noise that will become dominant), etc ...

You can either do some sort of calculation / measurement, or just trial and error.

[ollypenrice]

A quick rule of thumb would be to say that fewer-and-longer subs will go deeper and more-but-shorter subs will give a smoother result. I work only by experiment and have found that, with CCD at a dark sight, the very long sub does go deeper. For this reason I often use 30 minute subs in narrowband and, on some targets, in luminance. When trying to find the outer limits of M31 I found a very clear advantage in 30 minute luminance subs over my previous set of 15 minute ones. I actually had less total integration time in the 30 minute set yet it went deeper.

Olly

[Miguel1983]

4 hours ago, ollypenrice said:

A quick rule of thumb would be to say that fewer-and-longer subs will go deeper and more-but-shorter subs will give a smoother result. I work only by experiment and have found that, with CCD at a dark sight, the very long sub does go deeper. For this reason I often use 30 minute subs in narrowband and, on some targets, in luminance. When trying to find the outer limits of M31 I found a very clear advantage in 30 minute luminance subs over my previous set of 15 minute ones. I actually had less total integration time in the 30 minute set yet it went deeper.

Olly

30 minutes ?

Like Vlaiv said, if a plane or satellite passes by...

[ollypenrice]

50 minutes ago, Miguel1983 said:

30 minutes ?

Like Vlaiv said, if a plane or satellite passes by...

Planes are very rare for me, satellites fairly common. What's the problem? They are both dead easy to remove. For a severe plane trail you make a stack with all the subs in it and then another with the plane trailed sub excluded.

In Ps do an initial levels and curves routine saved as an action to the full set and save. Now open the set with the plane removed and run the action on that set. Paste the original stretch with plane on top and remove the plane trail with the eraser. Flatten and save. This will add no more than three minutes to your processing time. In truth I rarely need to do this and just run 'remove line' in Astro Art on affected subs before stacking them in Sigma mode.

I had seen the upturned end of M31 (extreme left hand side in this example) only once in another image and tried several times to capture it. Only when I went to 30 minute luminance subs was I able to detect it.

Olly

[wxsatuser]

As Olly, planes and satellites are easy to remove, 12 subs or more and the right stacking method, they gone.

Most of my 600sec subs have two or more planes and sometimes planes and satellites.
Planes going into Gatwick cross west east right over our location.

[iapa]

So, in summary...... suck it and see?

Everyone's location, conditions and equipment are different.

Members can only make suggestions as to how to approach this.