Another, what Gain/offset/exposure should I use for asi1600

[blinky]

Hi,

Im trying to figure out what settings to use for RGB and Narrowband on my ASI1600 - there are a few interesting but confusing posts on CloudyNights about this but I cant figure out what to do!  I have a red filter, offset 12, gain 75 image, exposed for 60 seconds. The scope was an 8" F4 Newt and my skies are pretty good, I can just about see the milky way so not fantastic but Im about 5 miles outside Edinburgh in a smallish village.  So.….I should be aiming for about 20x the read noise for the sky background yeah?  So, the read noise, when looking at the graph in the manual, for gain 75 looks to be about 2.25, so multiply that by 20 and you get 45.  So is 45 the magic number I should be aiming for in the Median of the image stats in SGPro?  If so then Im way overexposing as my median is 368.  

The other thing though is offset, I think its too low in this image as when looking in Pixinsight the histogram looks black clipped....

And finally, am  I best just to image at gain 75, or do people use different gain depending on filter?  I was using gain 300 for the Ha filter the other night and going with 2min subs.

So:
Should I drop my exposure time down lower till I get to 45 for the background?

Should I increase the offset?

Should I stick with Gain 75 for all filters or increase it for narrowband?

ngc7662_60sec_1x1_r_Red_offset_12_gain_75.fit

[blinky]

Pixinsight histogram

[david_taurus83]

Do you have Sharpcap Pro? It has a Samrt Histogram routine where it will take images of your sky conditions and recommend some exposure times, gain and offset settings based on your cameras read noise. It's very good. I have used it to determine that around 90s at a gain close to or around unity is good for LRGB where I am and around 10 minutes for narrowband. It tends to recommend a low offset for the ASI1600. I ignore this and set offset to 60 and gain at unity 139. Saves doing loads of different darks. Before you use it you need to run a routine to measure your cameras read noise. Tenner for the year and takes the guesswork out of things. And you get the best polar alignment method available that works in a couple of minutes! Set yourself an hour aside and have a watch of Robin's talk at the Practical Astronomy Show earlier this year.

 

 

[blinky]

Quote

Yeah thanks - watched that and its really interesting but a shame he never had time to go through the gain and offset settings!

 

[jjosefsen]

21 hours ago, blinky said:

 

The rest is somewhere on youtube, gain, offset, etc.

Here:

 

[symmetal]

On 04/08/2019 at 10:33, blinky said:

Im trying to figure out what settings to use for RGB and Narrowband on my ASI1600 - there are a few interesting but confusing posts on CloudyNights about this but I cant figure out what to do!  I have a red filter, offset 12, gain 75 image, exposed for 60 seconds. The scope was an 8" F4 Newt and my skies are pretty good, I can just about see the milky way so not fantastic but Im about 5 miles outside Edinburgh in a smallish village.  So.….I should be aiming for about 20x the read noise for the sky background yeah?  So, the read noise, when looking at the graph in the manual, for gain 75 looks to be about 2.25, so multiply that by 20 and you get 45.  So is 45 the magic number I should be aiming for in the Median of the image stats in SGPro?  If so then Im way overexposing as my median is 368.  

Hi blinky,

 In your calculations you haven't used consistant units. Read noise is in electrons (e-) while your sky background is in ADUs (16 bit).

To convert read noise to ADU (12 bit for ASI1600) divide read noise by the gain in e- per ADU. So read noise in 12 bit ADUs:    2.25 / 2 = 1.125

Multiply this figure by 20 :    1.125 * 20 = 22.5

Add the offset which is in12 bit ADUs :    22.5 + 12 = 34.5

Convert 12 bit ADUs to 16 bit ADUs by multiplying by 2 ^ 4 = 16 :     34.5 * 16 = 552

So your sky backbround ADU is 552 to achieve a  20 * Read Noise figure. Your current ADU figure of 368 is lower so you need to increase the exposure duration to reach 552.

In this thread there is some useful info, and at the end I included an excel spreadsheet to calculate the sky background ADUs for various read noise swamping calculations.

The offsets recommended by Zwo are too low and do cause black clipping.

I would just start at unity gain for everything until you get more familiar with the camera as any benefits/disadvantages with different gain settings can be quite small and taking many subs will be your best way of getting the best image. I use gain 139, offset 56 (to avoid black clipping) for everything. Another advantage of just using one gain setting is less sets of darks needed. 

Alan

[Adreneline]

1 hour ago, symmetal said:

I use gain 139, offset 56 (to avoid black clipping) for everything. Another advantage of just using one gain setting is less sets of darks needed.

+1 for these settings and advice.

I have found they work a treat for me and I have been delighted with the results from 120s and 180s NB exposures. In my particular location shorter exposures have also meant less wastage (compared with typical CCD exposures of 300s+) as I am right on the flight path for EMA - a busy place at night!

Good luck.

[symmetal]

46 minutes ago, Adreneline said:

+1 for these settings and advice.

I have found they work a treat for me and I have been delighted with the results from 120s and 180s NB exposures. In my particular location shorter exposures have also meant less wastage (compared with typical CCD exposures of 300s+) as I am right on the flight path for EMA - a busy place at night!

Good luck.

When I started with the ASI1600 I too used 2 x unity gain for narrowband and 0.5 x unity for LRGB as often quoted as 'optimum' on CN forums. Too often I ended up forgetting to set the right gain as SGP has three places where you can set the gain and 2 for the offset, with different priorities. The SGP option for adding the gain and offset to the filename never seemed to work for me so had to manually put it in the 'suffix' field to know what the settings were (hopefully).

Setting just one gain/offset in the Ascom driver only did away with the need for all this palaver. 

I use the 10 * RN ^ 2 setting for the sky background swamping level rather than the 20 * RN as that seems to be accepted as more 'realistic' and with my fixed gain/offset of 139/56 it's 1386 ADU for everything.

Alan

[blinky]

5 hours ago, symmetal said:

Hi blinky,

 In your calculations you haven't used consistant units. Read noise is in electrons (e-) while your sky background is in ADUs (16 bit).

To convert read noise to ADU (12 bit for ASI1600) divide read noise by the gain in e- per ADU. So read noise in 12 bit ADUs:    2.25 / 2 = 1.125

Multiply this figure by 20 :    1.125 * 20 = 22.5

Add the offset which is in12 bit ADUs :    22.5 + 12 = 34.5

Convert 12 bit ADUs to 16 bit ADUs by multiplying by 2 ^ 4 = 16 :     34.5 * 16 = 552

So your sky backbround ADU is 552 to achieve a  20 * Read Noise figure. Your current ADU figure of 368 is lower so you need to increase the exposure duration to reach 552.

In this thread there is some useful info, and at the end I included an excel spreadsheet to calculate the sky background ADUs for various read noise swamping calculations.

The offsets recommended by Zwo are too low and do cause black clipping.

I would just start at unity gain for everything until you get more familiar with the camera as any benefits/disadvantages with different gain settings can be quite small and taking many subs will be your best way of getting the best image. I use gain 139, offset 56 (to avoid black clipping) for everything. Another advantage of just using one gain setting is less sets of darks needed. 

Alan

Thanks for that, I’m still confused though, how come the excel spreadsheet shows 1283 for your sky background at gain 76?

[blinky]

It's because my offset is 12!  So, I should increase my offset to 60 to avoid clipping and at gain 75 I should be aiming for around 1,300 ADU in 16bit - Simples!

Edited August 5, 2019 by blinky

[symmetal]

44 minutes ago, blinky said:

Thanks for that, I’m still confused though, how come the excel spreadsheet shows 1283 for your sky background at gain 76?

Those calculations are with offset 56. If you use offset 12 you have to subtract (56 - 12) * 16 = 672 from those figures.

The 20 * RN then becomes 576 and the 10 x RN ^ becomes 611 ADU.

I read the graph as the noise at gain 76  being 2.2e- while you read 2.25e- hence the slight difference in the calculated values.

You deduced that just as I posted. 

Alan

Edited August 5, 2019 by symmetal

[blinky]

Sorry, we both replied at the same time!  Can you look at my last reply and confirm its correct?  If so Im going to create a master dark and bias library tonight for gain 75 offset 60 and gain 200 offset 60 and just use those for LRGB and Narrowband

[symmetal]

Yes. you were correct. 

Just to confirm the sky b/g ADU at gain 75, offset 60 is 1347 using 10 x RN ^ 2 and 1312 using 20 * RN.

At gain 200, offset 60, it's 1633 for 10 * RN ^ 2, and 1888 using 20 * RN.

At higher offset values this becomes a more dominant factor so the sky b/g ADU differences are not so greatly influenced by gain or calculation method used. I'd use the 10 * RN ^2 values personally. The numbers are just figures to aim for. If your image sky b/g ADU is around those figures that's all you need to do. For RGB you can either have different exposures for each colour to achieve these figures, or choose 1 exposure for all three which is near enough not to make much difference. Luminance will be around 1/3 the RGB exposure duration.

For narrowband it's unlikely you'll be able to expose long enough to achieve these sky b/g ADU figures unless you have bad light pollution. I generally choose 480 secs for Ha (at unity gain). In my bortle 3 skies I'd have to expose Ha for over an hour to achieve the 'recommended' ADU values.  With you using 2 x unity gain a shorter exposure may be better (240 secs or so) allowing you to take more frames in the same overall time to overcome the loss of bit depth caused by using higher gain.

It's commonly stated not to use bias with the 1600 as there are reports of some odd results, though personally my bias frames look 'normal' I just use the bias (or short exposure) to analyse the histogram to see if it's black clipping.

When you do some LRGB test exposures to determine the optimum exposure time wait until at least an hour after astro dark so that the sky brightness is fairly constant. Then use these exposures for all your subs. If astro dark has just arrived you'll find these exposures will give a b/g ADU which is too high but after a while they will settle down to the 'right' value.

If you create your dark library before knowing the correct exposure duration you may have a lot of unused darks, unless you use dark scaling. Good luck anyway.

Alan

Edited August 5, 2019 by symmetal

[blinky]

Thanks very much for all the information - Im going to create a dark library tonight, of 1,2,3,4,5,6,7 min subs - At least its just a one off exercise whilst its till light, I will stick them in my dark garage and shoot them tonight

[Starflyer]

23 hours ago, blinky said:

Thanks very much for all the information - Im going to create a dark library tonight, of 1,2,3,4,5,6,7 min subs - At least its just a one off exercise whilst its till light, I will stick them in my dark garage and shoot them tonight

CMOS sounds complicated Craig 🤪

[blinky]

Think I have it sorted now, and cmos means I can take shorter images so guiding does not have to be quite as good!