Seeking C&C for My Shot of the Sadr Region

John Favalessa:
A super image Harlan!  👏🏻  great dedication.  Framing is great.  I love the bold saturation, my personal preference.  I pixel peeped at full res...nice stars corner to corner.  The only think I could critique if I had to is the star color, not any red or gold, but that's the nature of NB stars, (but you have to pixel peek to notice).   So I give this two big thumbs up.  -john

Thanks John!

I've been reading about using a different filter to get RGB stars - something like the Antlia Triband RGB Ultra. Is this the proper way to do it, or is there a cheaper solution with similar results?

Thanks again,

Harlan

John Favalessa:
...in hindsight I should have used zero gain instead of 100...to get better star color...

I'm doing mono + filters so my question may or may not be germane for me but why would a lower gain give better color (unless you are over-saturating the stars in which case just reduce exposure time?)?

I have it as well.  That's actually an aspect I still do not quite understand.  The spec sheet says that 0 vs 100 gain has a dynamic range in stops of about 13.8 vs 13.5 (eyeball estimate from the graph).   Yet the FW(e-) does look like a large change (can't read it well, but will accept it is 4x).  I'm unclear how the dynamic range can vary so little with full well varying so much.  But the read noise takes a dramatic drop from about 3.6 to 1.5 or so while gain goes from about 0.8 to 0.28. 

So you get 2.8 times the gain with 2.4x the noise. 

I just don't know how to interpret the DR change, but in experimenting I stopped imaging broadband at zero and just do everything at 100, and have seen no degradation and significantly shorter imaging. 

Can you (or someone) explain the DR vs FW?

Linwood Ferguson:
I have it as well.  That's actually an aspect I still do not quite understand.  The spec sheet says that 0 vs 100 gain has a dynamic range in stops of about 13.8 vs 13.5 (eyeball estimate from the graph).   Yet the FW(e-) does look like a large change (can't read it well, but will accept it is 4x).  I'm unclear how the dynamic range can vary so little with full well varying so much.  But the read noise takes a dramatic drop from about 3.6 to 1.5 or so while gain goes from about 0.8 to 0.28. 

So you get 2.8 times the gain with 2.4x the noise. 

I just don't know how to interpret the DR change, but in experimenting I stopped imaging broadband at zero and just do everything at 100, and have seen no degradation and significantly shorter imaging. 

Can you (or someone) explain the DR vs FW?

Dynamic range is a tricky thing. It is defined as the ratio of the full well to the read noise. So what you have to do is look at the effect of gain on both of these.

As you increase gain, full well will decrease. No way around this unless you increase the bit depth. So if you have a 16 bit output file and the gain increases, the signal adu/e increases, so for the same photon input you get more signal into your 16 bit value. The full well in terms of electrons, the number of photons captured, must go down, since the signal per electron increases. 

As you go up in gain many cameras decrease the read noise. And often when the camera switches modes there is a big jump to lower noise. So dynamic range around that gain value may be about the same, as the full well decrease is matched by the big noise decrease. At higher gain, both full well and dynamic range decrease, as the full well decrease is faster than the noise decrease at those higher gains.

This means one should really only go to higher gain to reduce read noise, and thus also to reduce the sub exposure time. If you don't reduce the exposure time you will loose full well, even though your dynamic range at some higher gain may not be reduced. Of course, this only matters if you are saturating the stars, if you are not saturating you could even increase gain further to get more read noise decrease, as long as you keep within the full well.

Personally I look at the histogram on my subs to check if my stars are saturating. Of course, brighter stars will saturate the core no matter what low gain and short exposure. So I also use Nebulosity (free version) to actually check what is saturating, if it is just the cores of a few bright stars that is fine, as long as other star cores are not saturated, and thus will have good color.

Hope this helps
Rick

Your camera has similar behavior to mine, I have a 2600. Generally I would never recommend running this camera at Gain 0. There is a huge difference in read noise going to gain 100, noise is bad in an image, so why not take advantage. The dynamic range is the same, so yes, if you are saturating you just decrease your sub exposure time. If you are not saturating you are good to go, or even to increase your gain until you are getting close to saturation. The only disadvantage is more subs to process, but this actually helps using rejection methods in your stacking, to remove satellites, etc. But also shorter exposures are less prone to errors in tracking, wind, seeing and satellites. There are some subtle things too, there is something called quantization error. So by expanding the scale of adu/electron, you get more levels. So at 1 electron adu, if the system adds or subtracts say 0.6 adu equivalent noise, you can only get 0 or 1 as outputs. If you are at 10 adu/electron, then the 0.6 adu of noise is now 6 adu. So now you get 4, 10, and 16. So at least for the fainter signals you can get a more representative signal out, which is very important if you want the noise to cancel, as quantization reduces the randomness.

Rick