Is there anybody on here using the zwo asi1600 mono cooled ?.

[Wirral man]

As above I need some rough guidance as last night everything ran like a dream until today I processed the data and its awful 4 hours wasted on m101 

[vlaiv]

I am, or to be more precise, I would be using it if it were not for the bad weather all summer

How can I help?

[Wirral man]

1 hour ago, vlaiv said:

I am, or to be more precise, I would be using it if it were not for the bad weather all summer

How can I help?

Hi I have bortal sky of 5-6 I just need to know what gain and offset and exposure time to use for lrgb and narrowband I dont understand all the complex maths involved just need the settings so I can get a decent  image for once thanks

[vlaiv]

Here you go:

Offset 64

Gain 139

Exposures in range of 1 to 4 minutes - use shorter exposures for star color and to avoid saturation (depending on your processing workflow), use longer exposures for Lum (if no much saturation of target) and narrowband.

Calibrate your frames - use following calibration frames / technique:

1. Darks (matching temp, settings and exposure length)

2. Flats (depending on flat panel / source - look into ADU value being around 3500 scaled or  ~55000 in software that does not scale back to 12 bit - most will display it like that in 16 bit)

3. And matching flat darks - same exposure and settings as flats

Calibrate method:

Master_Flat_Dark  = stack of flat dark subs

Master_Flat = stack of (flat - Master_Flat_Dark)

Master_Dark = stack of darks

Calibrated frame = (light - Master_Dark) / Master_Flat

In each stage use 32bit precision (for creation of master calibration files, and when calibrating and stacking subs).

Take as many calibration files as you can - I personally use 256 of each.

[Wirral man]

26 minutes ago, vlaiv said:

Here you go:

Offset 64

Gain 139

Exposures in range of 1 to 4 minutes - use shorter exposures for star color and to avoid saturation (depending on your processing workflow), use longer exposures for Lum (if no much saturation of target) and narrowband.

Calibrate your frames - use following calibration frames / technique:

1. Darks (matching temp, settings and exposure length)

2. Flats (depending on flat panel / source - look into ADU value being around 3500 scaled or  ~55000 in software that does not scale back to 12 bit - most will display it like that in 16 bit)

3. And matching flat darks - same exposure and settings as flats

Calibrate method:

Master_Flat_Dark  = stack of flat dark subs

Master_Flat = stack of (flat - Master_Flat_Dark)

Master_Dark = stack of darks

Calibrated frame = (light - Master_Dark) / Master_Flat

In each stage use 32bit precision (for creation of master calibration files, and when calibrating and stacking subs).

Take as many calibration files as you can - I personally use 256 of each.

That's great so do I use the same gain and offset for both lrgb and narrowband sorry to sound dumb but I'm sick of wasting clear nights also with narrowband I just expose for longer ?

[vlaiv]

Just now, Wirral man said:

That's great so do I use the same gain and offset for both lrgb and narrowband sorry to sound dumb but I'm sick of wasting clear nights also with narrowband I just expose for longer ?

Yes, just the same settings but different exposure length. You can use even the same exposure length, but it is better to go with longer exposure length for narrowband because of reduced LP levels when doing narrowband. Difference will be in achieved SNR for given total imaging time. It is always better to go longer exposures, but in case of LRGB difference is so small that other benefits of shorter exposures win (like avoid star saturation to get proper colors, or less strain on guiding and less wasted time in case of wind / cable snag  .... ).

[david_taurus83]

4 hours worth of data should yield something. How long were your exposures? Post up a sub?

[Wirral man]

4 minutes ago, david_taurus83 said:

4 hours worth of data should yield something. How long were your exposures? Post up a sub?

Iv deleted them in anger lol but I had the gain set to 72 offset to 15 and 60sec exposure and they were just dark with no data I had to really pull on the curves to get any picture which just brought out noise and horizontal lines it was a right mess

[david_taurus83]

Ah the simplicity of ISO on a DSLR ?

[vlaiv]

7 minutes ago, Wirral man said:

Iv deleted them in anger lol but I had the gain set to 72 offset to 15 and 60sec exposure and they were just dark with no data I had to really pull on the curves to get any picture which just brought out noise and horizontal lines it was a right mess

Yes, that can mess up the frames - using as low offset as that tends to cut out the faint detail (embedded in read noise). Its a bit like clipping histogram to the left when processing the image - large areas of black appear - due to read noise and the way CMOS works - these will be visible as pronounced black lines (probably).;

[Wirral man]

4 minutes ago, vlaiv said:

Yes, that can mess up the frames - using as low offset as that tends to cut out the faint detail (embedded in read noise). Its a bit like clipping histogram to the left when processing the image - large areas of black appear - due to read noise and the way CMOS works - these will be visible as pronounced black lines (probably).;

Yes exactly will run with the settings you suggest and try yet again thank you

[Wirral man]

13 minutes ago, david_taurus83 said:

Ah the simplicity of ISO on a DSLR ?

Yes I so miss that...I really do lol was at the point today where I thought I should have just bought a colour model instead 

[david_taurus83]

http://www.stark-labs.com/help/blog/files/GainAndOffset.php

Worth a read.

[Wirral man]

2 minutes ago, david_taurus83 said:

http://www.stark-labs.com/help/blog/files/GainAndOffset.php

Worth a read.

Omg just had a quick glance at that link Haha think you need to understand quantum physics to make sense  well I do anyway...no but seriously I will give it a good read till it kinda makes sense as I would love to get a full understanding of this cheers bud!!

[david_taurus83]

And this.

 

https://www.google.com/url?q=http://www.astropixel.gr/uploads/7/8/3/5/7835053/scientifically_determining_ccd_gain_and_offset.pdf&sa=U&ved=0ahUKEwijsdy-vqTdAhWlIMAKHe54DocQFggWMAI&usg=AOvVaw26u2L2J_Yu7UnHTJ9LJcK5

[kens]

I do much the same as vlaiv. Main differences:

Use offset 50 at all gains

Gain 75 for L, gain 139 for RGB, gain 200 for narrowband

At gain 75 expose the histogram peak to 1100+, at 139 to 1200+, at 200 to 1400+ (if possible) - all in 16 bit values. Typically, under my skies, that means 30s exposures in LRGB and 3-5 minutes NB

The flats, darks and flat darks are most important. Flats I expose to 20000 (16 bit) and I take 40 of each

[Wirral man]

Great keep this info coming I need good results I spent so much on this camera and the kit filter wheel and filters and not getting the enjoyment that I did as with my simple DSLR and star adventurer 

[ngwillym]

check out this thread on Cloudy Nights

https://www.cloudynights.com/topic/573886-sub-exposure-tables-for-asi-1600-and-maybe-qhy163/

 

Neil

[Wirral man]

8 hours ago, ngwillym said:

check out this thread on Cloudy Nights

https://www.cloudynights.com/topic/573886-sub-exposure-tables-for-asi-1600-and-maybe-qhy163/

 

Neil

Hi bud I have seen this before and it made no sense as I'm orange band at f6.4

and at 139 gain it says 15secs?????. Far to short surly 

[vlaiv]

48 minutes ago, Wirral man said:

Hi bud I have seen this before and it made no sense as I'm orange band at f6.4

and at 139 gain it says 15secs?????. Far to short surly 

Math behind it is solid. What the author of that table had in mind (and I did not check their calculations, but have no reason to doubt them) is that when LP is dominant source of noise, read noise that makes all the difference in many short subs vs few longer subs tends to have very small contribution. Let me show you on a simple example:

Let's suppose we have two sites, one with such LP that it contributes with 10 photons per second per pixel, and other one that only has LP of 1 photon per second per pixel. And let's compare 1 minute exposure and 10 seconds exposure on those two sites.

1 minute:

Site1 total lp signal will be 600 photons, that means LP associated noise will be square root of 600 or ~24.5e, since read noise is ~1.6e (at unity gain) total noise (from LP and read noise) will be square root of sum of noise squares, or sqrt(24.5^2 + 1.6^2) = 24.55

Site2 total lp signal will be 60 photons, LP noise will be ~7.75 while LP+read noise will be 7.9

We might say that read noise contributes 24.55 / 24.5 = 1.002 or 0.2% total noise increase on site1, while it contributes 7.9/7.75 = 1.019 or 1.9% on site2

10 seconds:

site1 LP 60e, LP noise 7.75e, total noise 7.9, increase again 1.9%

site2 LP 10e, LP noise 3.162, total noise 3.544, increase (1.12) of 12%

So going from 10 to one minute in heavier LP changes ratio of (+read noise) / (no read noise) from 1.9% to 0.2.%, while same transition in low LP creates difference of going from 12% down to 1.9% - and this is considerable difference.

Simple fact is that if you are in heavy LP - it is the most dominant source of noise, and it scales with exposure duration. All main noise sources except read noise scale with exposure length, and target signal scales with exposure length, so read noise is only thing that differentiates many short subs vs few longer subs (same total exposure time). And if read noise is small compared to any other noise - you can go shorter. This is the reason CMOS sensors can go with shorter exposures vs CCD sensors - read noise in CMOS is around 2 while in CCD is at least 3-4 times bigger (depending on model). It is also one of the reasons for longer subs with narrowband imaging - narrowband tends to minimize LP related noise, so read noise again becomes relevant and you need long exposure in narrowband to collect enough LP signal for it's noise to become dominant.

But, bottom line is that while you can go with such short exposures (and there is minimal impact on final quality of your stack) - you don't need to go with such short exposures - you need to find balance between how short you want to go with your exposures to balance out things like pixel full well depth against saturation of bright stars, guide errors and wasted subs, amount of data you are willing to store and process ...

Hence my general recommendation of 1 to 4 minutes - if you look at those tables you will see that from most time and in most scenarios it fits well with their calculation. Yes you can use 15s subs and it will be almost the same SNR when you complete your stack, but you will need to gather 4 times as much frames, spend 1-2 seconds between each exposure in frame downloading, if you dither - there is another amount of time wasted and not capturing data, .... In my view 1-4 minutes is kind of best balance for all factors.

 

[ngwillym]

As Vlaiv says - its a matter of finding a sweet spot for you and your kit - fundamentally its  a trade off between sub-exposure length and stacking time. As an example, on myC11 @f/10 (2800mm) I can managed unguided exposures of ~20 secs, but I still expose each object for >60 mins - it just means my PC has to work harder stacking 100's of subs - but I'm used to that from my planetary & solar work where I regularly stack 1000's of subs.

There are a couple of minor benefits of shorter subs -I)  less subs are ruined by stray satellites/aircraft etc wandering through the frame ii) other frames affected by bad seeing can be ignored in the stacking process and iii) shorter subs can mean that stars don't saturate so easily.

 

I use Sharpcap as my capture - it has a 'brain' function that measures the sky background, and using calibration data from the camera, does the maths for me, so it will tell me the minimum exposure needed for the shot noise to overcome the read noise.

Neil

[Lead_weight]

I use those tables, and the math is sound. The short exposures are to ensure you don't overexpose the stars...not necessarily the object you're imaging. But in most images after stretching you'll see you have a few clipped stars, so you can expose longer so long as you're willing to blow out (pure white centers) a few of the larger stars in your photo.

I've experimented in doing short exposures of 30 seconds (almost double the recommended setting for my light pollution and F stop), and I did 750 images, which took about 4 hours of processing time just to calibrate the frames an an 8-core iMac Pro. My conclusion was that I was willing to go no less than 1 minute per sub exposure and would probably never again do 30s exposures.

I'm in a red zone and tend to do Gain 76 for LRGB with 60second subs.  And Gain 200 for narrowband with 3 minute subs. You will really struggle to get good narrowband results without upping your gain and exposure time...especially for OIII and SII. You can see some of my images on Astrobin. I post all my gain and exposure time for all my images.