Question regarding binning, gain and exposure settings + polar alignment

[rapture91]

Hello there,

at the moment I have several questions with respect to my EAA technique. I always try to improve things to get the best out of my equipment.
Starting off, my setup includes a 10“ f/5 GSO dobsonian over a single-axis EQ platform in combination with a Starizona Nexus 0.75x coma corrector/focal reducer (turning it into f/3.75) and the ZWO ASI533 MC color (uncooled). Additionally, I use the Baader Neodymium Moon & Skyglow filter for galaxies, open and globular clusters as well as reflection nebulae. For emission and planetary nebulae, I apply the Optolong L-eNhance. The faintest star I can usually see with direct vision is Eta Ursae Minoris with a magnitude of 5.0. This would translate to a Bortle level of either Bortle 5 (20.46 mag/arc sec^2) according to lightpollutionmap.info (which I would estimate too optimistic) or Bortle 6.4 (18.6 mag/arc sec^2) according to Sharpcap‘s Sky Background Calculator. So let‘s call it a Bortle 6 sky.

So far, I‘ve used a gain of 300 with 4 or 8 s exposures in bin 2 mode, since this was generally recommended for EAA on the CloudyNights forum.
However, in my recent images, I noticed some stars being plain white, which indicates pixel saturation.

I want to do EAA, so get a nice image in a maximum Integration time of 15 mins per object. Furthermore, I discovered post-processing some of the images is really fun and I‘d like to get the most out of the 16 bit FITS files I save for each object additionally to the „save as seen“ version. I would not bother having to wait a few more minutes to get a comparable picture with say lower gain values as long as post-processing would be easier.

Now my questions are the following:

1) Given my maximum possible exposure time of 8 s and the Bortle 6 skies, what combination of gain and exposure time would you recommend to me in order to
- have a pleasing EAA experience
- be able to sufficiently stretch the image afterwards while not saturating bright stars
Where should the histogram peak be located? More at roughly 1/4 or 1/2 away from the left end?

2) Should I image unbinned or in bin 2 mode?
When going unbinned, my resolution would be 0.83 arc sec/pixel. So far, I’ve used bin 2. However, I’m not sure whether that led to a loss in resolution and I would rather like to go in unbinned…

3) When comparing my image object X when using the Baader Neodymium filter with the image of the same object of another EAAer who is using very similar equipment (10“ dob, EQ platform, 3 sec exposures at 350 gain, Baader MPCC —> which would be similar to my Nexus CC), I noticed that his stars are much brighter and show way more well defined diffraction spikes.
I was wondering whether the Bortle 4 level he‘s imaging at is the sole explanation for this difference?

4) With respect to polar alignment, does the PA tool in Sharpcap work with equatorial platforms now? I once read that Tiago Ferreira was in contact with you, Robin, to make this work?
If so, I would like to try this feature

5) When using the L-eNhance, I understood that the contribution from light pollution noise should be smaller than that of read noise AND I have to achieve a solid SNR. So in this case, should I use the maximum exposure time my platform is capable of (8 s) in combination with high gain like 400?

Thanks in advance for your answer(s)!

Independently of that, I plan to do comparisons w/ gains of 150 and 300 in combination with 4 or 8 s exposures for the Baader Neodymium and see how it goes.

CS
Johannes (from Germanyj

[admin]

Hi Johannes,

phew! - lots of questions

1) Your skies at about bortle 6 or so in combination with 3.75 micron pixels on a colour camera and f/3.75 scope give an estimates 7e/pix/s light pollution level with no filters. If you image at gain 100 or above then your read noise is 1.5 or less, so applying the rule of thumb for exposure estimation of 10 times read noise squared divided by light pollution, we find 3.2s is the minimum exposure. Now, the use of filters will change this - just a bit for the Baader, but quite significantly (factor of 10?) for the L-Enhance.

Beyond gain 100 where the big drop in read noise is, there isn't much to be had by using higher gains - at most a factor of 2 reduction in exposure length.

It would be worth trying out the SharpCap smart histogram function to see what it recommends as giving the maximum dynamic range, but I suspect in the region of 150 to 250 gain would be fine.

2) Binning - binning is entirely software on CMOS cameras, so there is nothing gained by binning in SharpCap except smaller image files. You can get *exactly* the same results by doing the binning in post-processing if you want. If image file sizes are not an issue I would go with 1x1.

3) Lots of things can affect the sharpness of stars, but to begin with I would check that focus is accurate and that collimation is just right (very important with fast newtonians. I doubt that the darker skies will be the cause of this. Note that image stretching tends to blow out the stars (unless handled separately) - SharpCap live stacking tends to have this effect. Really the best approach is to separate the stars from the background, process them separately and then re-combine to avoid this problem

4) Yes, there is a 'low rotation' option for polar alignment that will give results with a small rotation of about 15 degrees (from memory). It may be somewhat less accurate, particularly if the pole is out-of-view of the image when aligning. It's certainly worth trying.

Hope this helps!

Robin

[rapture91]

Hi Robin,

thanks so much for the detailed reply!

1) I will most likely try either 4 or 8 s at a gain of 150 or 200 when using the Baader filter. Regarding the L-enhance, according to your recommendation, it would be best to aim for 8 s exposures at high gains of 300 - 400, since light pollution in this case should have a smaller effect than read noise on the total noise.

Would you think that when spending no more than 15 minutes on a target, a lower gain of 150 should still be sufficient to acquire enough data?

2) I will probably switch to bin 1. Would you say that one always benefits from the extra data/resolution or that seeing usually limits the details the camera can resolve?

3) my polar alignment routine so far consists of centering the true North Pole in my RACI finder, then rotating the platform to its end position (roughly 15 °).
I then watch the movement of the North Pole and adjust the south bearing in height as well as rotate the platform.
I think this procedure should be similar to the Sharpcap PA method.
Does the Sharpcap method use Polaris or the true North Pole for alignment?

Thanks again & CS
Johannes

[admin]

Hi,

remember that the gain setting doesn't actually make any difference to the number of photons that you will capture during an observing run (15 minutes or 15 hours) - the total number of photons per pixel purely depends on the total time and the f-ratio of the telescope. The reasons we adjust the gain are to try to ensure that each sub-exposure has enough photons per pixel for the random stochastic (shot) noise to be much bigger than the read noise, so that in the final average the read noise of the camera has minimal effect. Different gains also affect how quickly brighter areas saturate.

For the binning, I would expect that usually there will be limited extra information because of seeing limitations, but there will be a little and there is no downside to it beyond the larger files, so it's best just to go for 1x1 unless you are very significantly oversampling.

SharpCap's polar alignment routine plate solves the sky near the pole to work out the position in the image of the true pole (adjusted for date due to precession, etc). The rotation in RA step allows SharpCap to work out the current pointing position of the RA axis (the point that appears to stay still). Once both are known, adjustment is simply putting the two in the same place.

cheers,

Robin