Narrowband Subs

[Kaliska]

After stacking my Ha subs in PI (or DSS come to that), the resulting image always looks nice and smooth, but not for my OIII or SII. These always show a lot more noise, just wondering if this is a common occurence? 

[Tim]

Yes. There is usually many more times the signal in redder wavelengths.

The answer is longer or more subs, but preferably longer.

I dont use Sii anymore, just one for reference and then see how long a red sub will match the signal.

[RolandKol]

2 hours ago, Kaliska said:

After stacking my Ha subs in PI (or DSS come to that), the resulting image always looks nice and smooth, but not for my OIII or SII. These always show a lot more noise, just wondering if this is a common occurence? 

Ha is much intense signal in the "Sky"  as almost all the stuff out there is predominantly made from Hydrogen and signal to noise ratio is much higher in this spectrum,

plus, in addition, H spectrum is nicely separated from the Light Polution bands. 

In opposition, -  OIII and SII, are much weaker out-there, plus they usually overlap with all kinds of Light pollution bands. Usual recommendation, do not image OIII and SII while Moon  is Full, or if target is close to the Moon.

Of course, 3nm OIII/SII filters would help
I bought Baader NB bundle some time ago and I have no idea why, why they include 8.5nm OIII and 8nm SII, while Ha is 7nm in the bundle... Probably "price related" tactics, as there is no logic in it...

[RolandKol]

1 hour ago, Tim said:

Yes. There is usually many more times the signal in redder wavelengths.

The answer is longer or more subs, but preferably longer.

I don't use Sii anymore, just one for reference and then see how long a red sub will match the signal.

Do you mean, you use R filter instead of SII?

A bit confused... I am also quite new in AP, - but based on my small and quite recent experience, R filter does not stand close to SII band... R is always closer to H (if to compare stacks or even single longer subs).

[Tim]

There are very few targets with decent Sii. There is also significant crossover between Red, Ha, and Sii. Even when exposing for 60 minutes or more, a typical Sii sub is noisy, with very little Sii signal actually there. Some targets are an exception, IC410 and NGC 281 spring to mind. At any rate, for my tastes an equivalent approximation of a 30 minute Sii sub can be had with a 2 minute or less Red. The brightest Sulphur emissions are picked up readily through the Red filter.

If I want a scientifically accurate image (I dont) then I use the Sii filter, but the sad fact is that in the UK at least, spending precious clear sky time taking noisy, low signal Sii exposures for hour after hour is pretty soul destroying. If we had  better weather and longer clear spells it might be different.

[ollypenrice]

44 minutes ago, Tim said:

There are very few targets with decent Sii. There is also significant crossover between Red, Ha, and Sii. Even when exposing for 60 minutes or more, a typical Sii sub is noisy, with very little Sii signal actually there. Some targets are an exception, IC410 and NGC 281 spring to mind. At any rate, for my tastes an equivalent approximation of a 30 minute Sii sub can be had with a 2 minute or less Red. The brightest Sulphur emissions are picked up readily through the Red filter.

If I want a scientifically accurate image (I dont) then I use the Sii filter, but the sad fact is that in the UK at least, spending precious clear sky time taking noisy, low signal Sii exposures for hour after hour is pretty soul destroying. If we had  better weather and longer clear spells it might be different.

Point well made, but what about star size? (I don't do pure NB imaging sop this is a genuine question.)

Olly

[Kaliska]

 

On 05/03/2019 at 09:26, RolandKol said:

Ha is much intense signal in the "Sky"  as almost all the stuff out there is predominantly made from Hydrogen and signal to noise ratio is much higher in this spectrum,

plus, in addition, H spectrum is nicely separated from the Light Polution bands. 

In opposition, -  OIII and SII, are much weaker out-there, plus they usually overlap with all kinds of Light pollution bands. Usual recommendation, do not image OIII and SII while Moon  is Full, or if target is close to the Moon.

Of course, 3nm OIII/SII filters would help
I bought Baader NB bundle some time ago and I have no idea why, why they include 8.5nm OIII and 8nm SII, while Ha is 7nm in the bundle... Probably "price related" tactics, as there is no logic in it...

So, here are the two stacked images in HA and OIII. Both at -20, unity gain and 180sec. You recommend a longer exposure for the OIII?

[jimjam11]

This is a cmos camera? Longer might not be better per sub but more total integration will help if you are already swamping the read noise.

What is your median background level per sub above bias? If you aren’t swamping the read noise then you do need longer subs or a higher gain...

Very useful thread for cmos:

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

[Kaliska]

1 hour ago, jimjam11 said:

This is a cmos camera?

Yes, its a ASI1600 MM Cool and how do i determine the median background levels of both the sub and a flat (i'm still learning btw).

[jimjam11]

In SGP you can just dither the mouse in a dark area and it will give you the value in the statistics panel. It will also show you the whole image median level which is usually close. You then need to subtract the bias. You can get the bias amount by taking a bias frame (0.3s works well for the ASI1600) and doing the same thing in SGP.

For flats SGP has a flats exposure wizard. I just tell it to target 15000 ADU and it calculates the correct exposure. (N.B. You need to set the gain manually through the driver for this).

In PI you can use the Statistics module to get the background level. Alternatively you can use the readout but you then need to convert the number.

[Kaliska]

HA Light Info

HA Flat Info

OIII Light Info

OIII Flat Info

I use the Flats Calibration wizard before, but i do set it to 25000ADU. Is this too high?

[jimjam11]

I think 25k will be fine, you should be nowhere near clipping.

What exposure time was each sub?

What is your offset, gain for each sub and what is the resulting median adu for a bias frame (min 0.3s)?

 

[Kaliska]

34 minutes ago, jimjam11 said:

I think 25k will be fine, you should be nowhere near clipping.

What exposure time was each sub?

What is your offset, gain for each sub and what is the resulting median adu for a bias frame (min 0.3s)?

 

Both Ha and OIII lights are 180sec exp, preset is at unity gain, gain 139, offset 21

[jimjam11]

You want to measure your bias accurately but it will be in the realm of 21x16=336

Your median exposures are:

Ha=576 - 336 = 240

Sii=608 - 336 = 272

 

You want to be >= 450 adu above bias so you are somewhat underexposed and are therefore not working at maximum efficiency.

-You could extend sub length to 5m and see if that gets you 450 above bias. You could keep increasing exposure up to about 10m until you get to your target adu  above bias, I don't think many people go beyond 10m with the asi1600 because glows can become problematic.

-You could increase gain to 200 and leave the sub length at 180s.

 

I always use gain 200 for nb with my asi1600mm pro

 

[RolandKol]

36 minutes ago, Kaliska said:

Both Ha and OIII lights are 180sec exp, preset is at unity gain, gain 139, offset 21

Hmmm... Which scope? 

your fov looks like mine, 130PDS..... if the scope is the same....

We have very different results per sub....

I am not far away from you, Woolwich (which probably is just a bit brighter), and on the same Rosette nebula 180sec HA (Baarder 7nm), - I have around 1550 Median ADU (shown on SGP).

Just in my case, offset is 50, as ZWO updated their drivers and updated the default/preset Offset values.

[Kaliska]

I've got a TS 100 Quadruplet and its fairly dark here at a green zone Class 4 Bortle according to my Clear Outside app.

[RolandKol]

Hrrrr  even jealous!

If trees and roofs are not in the way and U F Os not visiting your FOV frequently... Try 600sec
Less subs to hold, less time to pre-process

Beauty!

I was quite fine with 300sec in my Bortle 8-9 area

Dropped down to 180sec as usually failed to collect 12subs per filter per session due to FOV obstructions....

 

[Kaliska]

The photo above is facing directly east and as you can see i'm quite enclosed, garage roof and high bamboo in shot with my house roof directly behind. planes , plenty of them for sure (Stanstead Airport is 15mins up the road).

But, as soon as i get a chance, I'll try out some longer test exposurers and see what results i get.

[Kaliska]

27 minutes ago, RolandKol said:

Hrrrr  even jealous!

If trees and roofs are not in the way and U F Os not visiting your FOV frequently... Try 600sec
Less subs to hold, less time to pre-process

Beauty!

I was quite fine with 300sec in my Bortle 8-9 area

Dropped down to 180sec as usually failed to collect 12subs per filter per session due to FOV obstructions....

 

The photo above is facing directly east and as you can see i'm quite enclosed, garage roof and high bamboo in shot with my house roof directly behind. UFO's, plenty of them for sure (Stanstead Airport is 15mins up the road).

But, as soon as i get a chance, I'll try out some longer test exposurers and see what results i get.

[RolandKol]

yep... you will probably benefit from higher gain just like me... As all the obstacles reduce the total amount of  subs per session drastically... plus Stanstead... plus weather....

Once I will finish with my camera spacing adjustments, I am planning to switch to gain 200, but I afraid of it, my skies shine almost as a LED panel....

Also, I still have a hope to catch Rosette for the last session on the old settings.

P.S.

Off topic
 Friend of mine offered his Roof in a suburban area of Argentina, -  as a remote observatory!
While visiting him, I was able clearly see Orion Nebula at 9PM in Feb, while street lights were still On!  Oddly, Upside Down, but clearly visible!

Sadly, Argentina is a bit too distant to sent all the toys away  

Plus my rig is not completely autonomous, - yet....
 

[glowingturnip]

I'm currently in the middle of processing an HOO version of the Rosette too.  I got a decent number of 10min subs of each, so the OIII isn't too noisy (though definitely noisier than the Ha).

However, I wanted to try to get a 'natural' balance of the three channels, so adding in the OO in about the right mix according to the signal that's there - I did it by two methods, firstly by adding it in what I think are the quantum-efficiency weightings of my camera, but alternatively by adding it so that the average intensity of stars only in each channel is the same (I'm assuming the stars on average would give equal amounts of Ha and OIII).  Both methods give roughly similar mixes.

Anyway, blah...  but the upshot is that for the Rosette, I reckon the OIII is *much* weaker than the Ha.  I've ended up going for Ha-only luminance since adding in any OIII was muddying the backgrounds.  In the chrominance I've not used it for much more than adding pleasant tonal detail to the red.  Makes me wonder why I spent a whole night capturing OIII really !

[glowingturnip]

what's that red thing sticking out of the front of your polarscope btw ?  And does your other half go nuts at you crunching around in that slate gravel at 3am ?  My wife sure does when I do it ? 

[RolandKol]

1 hour ago, glowingturnip said:

I'm currently in the middle of processing an HOO version of the Rosette too.  I got a decent number of 10min subs of each, so the OIII isn't too noisy (though definitely noisier than the Ha).

However, I wanted to try to get a 'natural' balance of the three channels, so adding in the OO in about the right mix according to the signal that's there - I did it by two methods, firstly by adding it in what I think are the quantum-efficiency weightings of my camera, but alternatively by adding it so that the average intensity of stars only in each channel is the same (I'm assuming the stars on average would give equal amounts of Ha and OIII).  Both methods give roughly similar mixes.

Anyway, blah...  but the upshot is that for the Rosette, I reckon the OIII is *much* weaker than the Ha.  I've ended up going for Ha-only luminance since adding in any OIII was muddying the backgrounds.  In the chrominance I've not used it for much more than adding pleasant tonal detail to the red.  Makes me wonder why I spent a whole night capturing OIII really !

I am even a bit jealous reading you are able to finish 10min exposures on OIII... in my case 5mins = close to overexposed!

And usually I end up deleting half of my OIII subs due to quality... And even whats left, not even close to quality of HA.

So I end up de-noising and smoothing OIII as much as possible and sharpening HA, which I use to bring back the final sharpness. 

not yet perfect... but makes background a bit smoother...

 

[Kaliska]

On 22/03/2019 at 15:56, glowingturnip said:

what's that red thing sticking out of the front of your polarscope btw ?  And does your other half go nuts at you crunching around in that slate gravel at 3am ?  My wife sure does when I do it ? 

It's a Polemaster, which doesn't get a lot of use these days since I had the permanant pier installed. The scope is setup permanantly and covered with a TeleGizmos 365 cover (TeleGizmo)  with a small heater and silica underneath to keep and moisture away. So the only gravel crunching is done at the start of the night as the small computer pictured on top of the scope is Wi-Fi and all the imaging is done from inside the house and left to run throughout the night (weather permitting of course).

 

My final image didn't come out too bad in the end.

 

 

[RolandKol]

Extremely nice one
I was quite happy with my "Depressed Rosette" initially

not anymore