#1  
Old 04-07-2022, 11:18 AM
bluesilver (Peter)
Registered User

bluesilver is offline
 
Join Date: Jul 2018
Location: Australia, Tasmania
Posts: 426
asi2600mc very noisy images

Hi, Not sure if i have got this posted in the correct section here or not.
I have just had a good clear night last night to try and do some imaging of NGC 6744 and also NGC 6334
Camera is asi2600mc, scope is Sykwatcher Esprit 150
All images are 2 minute exposures at gain 100 and -10c
I stacked 50 lights, 50 Flats, 50 Bias frames
Light frames were done in APT at 22000ADU
Stacked in DeepSky stacker and processed in Startools

As you can see the noise is very very bad,
NGC 6334 has some very bad sort of white glow effect going on also.
I took the flats and bias right after i finished taking the lights, so nothing was disturbed.

Something is way off here, I am not sure if it is a noisy cable form the camera ( standard ZWO cable )
Not sure if i should of included Dark Flats and Darks also in the stacking process?

In DeepSky Stacker, under Stacking Parameters, under Light, you have 3 options:
No Background Calibration
Per Channel Claibration
RGB Channels Background Calibration
( I selected the last one, I think this is what it is at the default settings anyway )

The you have an options button with:
Linear
Rotational
( i selected Rotational, again i think this is the default setting )

Then you have RGB Background Calibration Method:
None
Minimum
Middle
Maximum
( I selected maximum, think this is the default setting )

When all stacked and finished, i saved it as a TIF 16 bit file

Open in Startools as a Linear from OSC/DSLR
Using the wipe function and tried both basic and vignett options.

So something is at a miss here or i am stacking it all wrong.
Any advice or suggestions would be very much appreciated.
Attached Thumbnails
Click for full-size image (Catpaw11 22000ADU.jpg)
97.3 KB124 views
Click for full-size image (Spiral Galaxy11 2200ADU.jpg)
120.8 KB125 views
Reply With Quote
  #2  
Old 04-07-2022, 11:26 AM
Drac0's Avatar
Drac0 (Mark)
Registered User

Drac0 is offline
 
Join Date: Feb 2020
Location: Nowra, NSW
Posts: 522
Looks like I'm mistaken with this camera.
Cheers,
Mark

Last edited by Drac0; 04-07-2022 at 11:58 AM.
Reply With Quote
  #3  
Old 04-07-2022, 11:39 AM
AdamJL
Registered User

AdamJL is offline
 
Join Date: Sep 2010
Posts: 2,100
Are you able to upload the stacked files somewhere? Your images are quite small here; it's hard to see noise which requires a close up view.

That said, the Wipe function in ST shouldn't be used as a metric for noise. It purposefully does a very very aggressive stretch of the image so you can see any issues. Personally, I really don't like it how it does it, even though I understand why it does.

The first image though, looks like a (very bad) bad flats problem. Are you sure your flats were taken with the same settings and without the image train changing?

Also, ditch that flat cable. It's the first rule of thumb with ZWO anything... those flat cables are better used as linguine than actual data transfer cables. It likely won't have caused your issues, but you're just delaying the inevitable by using it.

And lastly, stick to Lights, Flats, and Bias only with the 2600 cameras (and 268s). You don't need to do darks and dark flats, your flats problem will get worse (speaking from experience as I've seen it happen to my own images)
Reply With Quote
  #4  
Old 04-07-2022, 11:51 AM
Drac0's Avatar
Drac0 (Mark)
Registered User

Drac0 is offline
 
Join Date: Feb 2020
Location: Nowra, NSW
Posts: 522
Quote:
Originally Posted by AdamJL View Post
And lastly, stick to Lights, Flats, and Bias only with the 2600 cameras (and 268s). You don't need to do darks and dark flats, your flats problem will get worse (speaking from experience as I've seen it happen to my own images)
That's interesting. Learn something new every day. Do the bias do enough to deal with hot/stuck pixels?
Reply With Quote
  #5  
Old 04-07-2022, 12:48 PM
bluesilver (Peter)
Registered User

bluesilver is offline
 
Join Date: Jul 2018
Location: Australia, Tasmania
Posts: 426
Thanks for the replies,
Yes 100% sure all flats and Bias were taken right after the Lights were done.
So that part is done correct then, just Flats and bias.
I wanted to make all setting weren't changed at all.
I have had this light glow issue now on a few attempts, i thought i was a flats issue, so lowered the ADU from 30000 to 22000 for this run to see if it helped, afraid it didn't though.
I uploaded the files on here to jpeg and reduced them by 25% as they were a tad large.
Tried uploading the same version but at 100% but it won't allow it here.
I am not sure where else to upload the full version to yet, will have to look into that one.
Will look at replacing that flat cable, If they make a shielded USB A to B cable, that would be best, If so i would also look at replacing the cable inside the mount to a shielded version also.

I am now thinking if it could possibly be a smoke haze for a neighbors fire,
Was sitting at about 1 degrees outside when these were taken.

Next thought was to do a back to back comparison with my DSLR, run an image session with the DSLR then swap it straight over to the asi2600mc and run that , but the specs say that the asi2600mc should be far superior in all areas.
Reply With Quote
  #6  
Old 04-07-2022, 12:57 PM
AdamJL
Registered User

AdamJL is offline
 
Join Date: Sep 2010
Posts: 2,100
Quote:
Originally Posted by Drac0 View Post
That's interesting. Learn something new every day. Do the bias do enough to deal with hot/stuck pixels?
It's only these modern sensors. A couple of other cameras use the same sensor design, and they really don't benefit from darks or dark flats. Mainly because there's just little to no noise anyway.

As for hot pixels, I haven't seen any in my 2600MC or 268M, but I haven't looked too closely. Dithering is how I combat that!

Quote:
Originally Posted by bluesilver View Post
Thanks for the replies,
Yes 100% sure all flats and Bias were taken right after the Lights were done.
So that part is done correct then, just Flats and bias.
I wanted to make all setting weren't changed at all.
I have had this light glow issue now on a few attempts, i thought i was a flats issue, so lowered the ADU from 30000 to 22000 for this run to see if it helped, afraid it didn't though.
I uploaded the files on here to jpeg and reduced them by 25% as they were a tad large.
Tried uploading the same version but at 100% but it won't allow it here.
I am not sure where else to upload the full version to yet, will have to look into that one.
Will look at replacing that flat cable, If they make a shielded USB A to B cable, that would be best, If so i would also look at replacing the cable inside the mount to a shielded version also.

I am now thinking if it could possibly be a smoke haze for a neighbors fire,
Was sitting at about 1 degrees outside when these were taken.

Next thought was to do a back to back comparison with my DSLR, run an image session with the DSLR then swap it straight over to the asi2600mc and run that , but the specs say that the asi2600mc should be far superior in all areas.
Can I ask how you're taking your flats? Are you using a man made source or the sky?
I would check for light leak in your train. If you're getting it more than once, it appears your camera might be picking up stray light from somewhere, as flats are usually taken with a much brighter source and so light leak can show up easier.
Spend some time covering various parts of your scope, check for internal reflections, etc.
I don't envy you on this stage, because I'm still trying to fix some internal reflections of my own and it's not fun.

For image hosting, if you have OneDrive or similar, you can upload them there. Failing that, I'm happy to host them on my own OneDrive if you want. Send me a PM and I can send you the link (and then share it here, so others can also play with the files)
Reply With Quote
  #7  
Old 04-07-2022, 01:11 PM
The_bluester's Avatar
The_bluester (Paul)
Registered User

The_bluester is offline
 
Join Date: Feb 2011
Location: Kilmore, Australia
Posts: 3,342
I have actually found the opposite with regard to darks, with really good flats (And your master flat and master bias should be 32 bit not 16 bit) you can sometimes see a slight brightening in the corners at high stretches unless a master dark is used along with a master flat. When I first got my 2600MC I used master bias only to calibrate both lights and flats, but soon moved to master bias for the flats to generate a master flat, and that master flat and a master dark, to calibrate lights. Dark-flats are not very useful for me as I normally shoot sky flats so the exposure times vary widely, in the range of 0.2 to 40 seconds per filter.

For the OP, you should be able to create master dark and master bias frames and re-use those for months, you don't need to shoot bias frames to go with each set of lights at the time you shoot them. So long as you use the same gain and sensor temp master bias and master dark frames are good to reuse for months with these cameras. Flats are the ones you don't want to disturb the image train for, to make sure any vignetting and dust bunnies match up with the lights.
Reply With Quote
  #8  
Old 04-07-2022, 01:15 PM
xelasnave's Avatar
xelasnave
Gravity does not Suck

xelasnave is offline
 
Join Date: Mar 2005
Location: Tabulam
Posts: 16,866
Have you done a stack with no flats?

I am trying to include flats in my game but with them I am getting crazy results similar to your Cats Paw.

I dont know what I am doing incorrectly and in time will sort it out but I have the same camera ...however try a stack with no flats.

alex
Reply With Quote
  #9  
Old 04-07-2022, 01:21 PM
bluesilver (Peter)
Registered User

bluesilver is offline
 
Join Date: Jul 2018
Location: Australia, Tasmania
Posts: 426
Quote:
Originally Posted by The_bluester View Post
I have actually found the opposite with regard to darks, with really good flats (And your master flat and master bias should be 32 bit not 16 bit) you can sometimes see a slight brightening in the corners at high stretches unless a master dark is used along with a master flat. When I first got my 2600MC I used master bias only to calibrate both lights and flats, but soon moved to master bias for the flats to generate a master flat, and that master flat and a master dark, to calibrate lights. Dark-flats are not very useful for me as I normally shoot sky flats so the exposure times vary widely, in the range of 0.2 to 40 seconds per filter.

For the OP, you should be able to create master dark and master bias frames and re-use those for months, you don't need to shoot bias frames to go with each set of lights at the time you shoot them. So long as you use the same gain and sensor temp master bias and master dark frames are good to reuse for months with these cameras. Flats are the ones you don't want to disturb the image train for, to make sure any vignetting and dust bunnies match up with the lights.
Appreciate that advice, so sounds like i have made a mistake in saving my final stacked image to start with.
Should they be TIFF or FITS ? and then should they be 32 bit/ch integer or 32 bit/ch rational ?

All i have been doing in DeepSky Stacker is just loading all my lights in, all my Flats in, all my Bias in and then just let it stack them.

I haven't tried or looked into separate stacking like i think you are talking about?
So it sounds like you stack all your Bias frames first, this generates a master Bias.
Then you load all your lights and Flats in and the mater bias then stack
these?

Last edited by bluesilver; 04-07-2022 at 01:37 PM.
Reply With Quote
  #10  
Old 04-07-2022, 01:22 PM
bluesilver (Peter)
Registered User

bluesilver is offline
 
Join Date: Jul 2018
Location: Australia, Tasmania
Posts: 426
Quote:
Originally Posted by xelasnave View Post
Have you done a stack with no flats?

I am trying to include flats in my game but with them I am getting crazy results similar to your Cats Paw.

I dont know what I am doing incorrectly and in time will sort it out but I have the same camera ...however try a stack with no flats.

alex
Good point, no i haven't tried it without the flats.
Might even just try doing a stack with just the Lights only and see what that produces, good point there in elimination
Reply With Quote
  #11  
Old 04-07-2022, 01:45 PM
Bassnut's Avatar
Bassnut (Fred)
Narrowfield rules!

Bassnut is offline
 
Join Date: Nov 2006
Location: Torquay
Posts: 5,064
the 1st picture is exactly what happens in Startools auto stretch if you have stacking artifacts. crop 1st to eliminate these.
Reply With Quote
  #12  
Old 04-07-2022, 01:51 PM
bluesilver (Peter)
Registered User

bluesilver is offline
 
Join Date: Jul 2018
Location: Australia, Tasmania
Posts: 426
Quote:
Originally Posted by Bassnut View Post
the 1st picture is exactly what happens in Startools auto stretch if you have stacking artifacts. crop 1st to eliminate these.
Appreciate the advice.
The only issue with cropping that image was that the artifacts are all over the entire image.
But i do appreciate what you are saying there.

I just did a real quick try of just stacking lights and bias on that spiral image.
And saved as a 32 bit TIFF and a very quick process in StarTools.
Not a real lot of difference in the result, but appears to be a slightly cleaner image
I will have to come back to this though, as just shift work calls.
Appreciate all advice.
Attached Thumbnails
Click for full-size image (Spiral13, lights, darks.jpg)
195.9 KB56 views
Reply With Quote
  #13  
Old 04-07-2022, 02:14 PM
Startrek (Martin)
Registered User

Startrek is offline
 
Join Date: Dec 2017
Location: Sydney and South Coast NSW
Posts: 5,990
Hi Peter,
As you probably know I use exactly the same camera and software
Firstly what do you mean by “Light frames were done in APT at 22,000ADU” I’m assuming you mean the Flats ?
Secondly your APT histogram is the roadmap or guide for successful Flats with this camera.
You should adjust the brightness of your flat panel or light box etc..and ADU target value to achieve an ADU range on the histogram of around 8500 to 35,000 ADU or around halfway along the Histogram. Not just select 22,000ADU and hope it’s ok
I found an ADU target of 21,000 to 22,000 was the right value after adjusting the brightness on my light panel and achieving a “ADU range” on the histogram of roughly between 8500 and 35,000 ADU or up to half way

In regard to Stacking with DSS

“RGB Channel Background Calibration” is a definite no no for Startools so set this Calibration option to either “Per Channel Background Calibration”or preferably “No Channel Background Calibration” In “options” leave Calibration Method to Linear and RGB Background Calibration Method to none.

Below are a bunch of settings you should use to ensure your stacked fits file is linear and meets the requirements for loading into Startools

DSS Latest Settings for both RGB and Narrowband data ( Prerequisites for Startools )

1/ Tick “No White Balanced Processing” in Raw Files Digital Development
2/ Leave “Set the Blackpoint to 0” unticked in Bayer Matrix Transformation
3/ Select “Per Channel Background Calibration” in Stacking Parameters even though when using Startools it should be “No Calibration”. I found Per Channel Calibration no issue when using Startools.
4/ Select “Intersection Mode” in Stacking Parameters (helps reduce stacking artefacts
5/ Tick “Align RGB Channels in final image” in Stacking Parameters
6/ Digital Development Process Fits setting and Raw settings select Bilinear Interpolation
7/ In Digital Development Process Settings ensure you tick “Monochrome 16bit Fits files created by a DSLR or a colour CCD camera” when using a DSLR or OSC camera like your 2600MC.
8/ Stacking Parameters, Lights use Kappa Sigma Clipping
9/Stacking Parameters, Calibration Frames use Median
10/ Stacking Parameters, Flats Median Kappa Sigma Clipping
11/ Stacking Parameters, Bias Median Kappa Sigma Clipping
12/ Ensure you save your stacked file as a “32 bit channel integer Fits files with no adjustments applied”
13/ In Digital Development Process Settings under FITS tab , use Generic RGGB for your cameras Bayer matrix ( I have the Latest version of DSS and the 2600MC is not on the camera list)
Leave colour adjustment box as default 1.0000 for all 3 items

I performed exhaustive testing for a month on all types of data earlier this year on calibration frames with the 2600MC and found that Flats and Bias worked fine ( Darks are not needed )
Your Bias frames can be captured at an exposure of 0.000032 sec which is the shortest exposure in the ZWO specs for this camera. I tried 1 sec , 0.5 sec , 0.1 sec as well with no discernible difference. I just use 0.000032 sec

Also when capturing Calibration frames ensure there is no stray light leakage that may be entering your image train.
As I use Newts , I wrap the focuser and camera with a black cloth and fit a shower cap sprayed matt black over the back end of the OTA to eliminate stray light when capturing Flats and Bias frames during the night

The photos you posted look like a calibration issue ( either with colour balance or your background)

Hope in the above info you have picked up something but I’ll keep thinking of things that might have caused this issue

Cheers
Martin

Last edited by Startrek; 04-07-2022 at 02:47 PM.
Reply With Quote
  #14  
Old 04-07-2022, 02:17 PM
AdamJL
Registered User

AdamJL is offline
 
Join Date: Sep 2010
Posts: 2,100
Quote:
Originally Posted by bluesilver View Post
Appreciate the advice.
The only issue with cropping that image was that the artifacts are all over the entire image.
But i do appreciate what you are saying there.

I just did a real quick try of just stacking lights and bias on that spiral image.
And saved as a 32 bit TIFF and a very quick process in StarTools.
Not a real lot of difference in the result, but appears to be a slightly cleaner image
I will have to come back to this though, as just shift work calls.
Appreciate all advice.
StarTools will show you those issues, and then you need to mask them out or crop if they're on the edges. Once that's done, do another stretch and see how it goes (or upload to the link I sent you and we can have a go).
Reply With Quote
  #15  
Old 04-07-2022, 02:20 PM
AdamJL
Registered User

AdamJL is offline
 
Join Date: Sep 2010
Posts: 2,100
Quote:
Originally Posted by bluesilver View Post
Appreciate that advice, so sounds like i have made a mistake in saving my final stacked image to start with.
Should they be TIFF or FITS ? and then should they be 32 bit/ch integer or 32 bit/ch rational ?

All i have been doing in DeepSky Stacker is just loading all my lights in, all my Flats in, all my Bias in and then just let it stack them.

I haven't tried or looked into separate stacking like i think you are talking about?
So it sounds like you stack all your Bias frames first, this generates a master Bias.
Then you load all your lights and Flats in and the mater bias then stack
these?
It seems like a long way to go about it
You should be able to just chuck your lights and calibration files into DSS without targeting each stack separately. I have never done anything else when I used to use DSS.

btw, TIFF or FITs is fine. I just stick to FITs for the main astronomy programs, then change it to TIFF when I export to Photoshop for the final touch ups.
Reply With Quote
  #16  
Old 04-07-2022, 02:57 PM
Bassnut's Avatar
Bassnut (Fred)
Narrowfield rules!

Bassnut is offline
 
Join Date: Nov 2006
Location: Torquay
Posts: 5,064
Quote:
Originally Posted by bluesilver View Post
Appreciate the advice.
The only issue with cropping that image was that the artifacts are all over the entire image.
But i do appreciate what you are saying there.

I just did a real quick try of just stacking lights and bias on that spiral image.
And saved as a 32 bit TIFF and a very quick process in StarTools.
Not a real lot of difference in the result, but appears to be a slightly cleaner image
I will have to come back to this though, as just shift work calls.
Appreciate all advice.
This pic looks even more like stacking artifacts.
draw a box over the central part that has no artifacts and turn off ROI (Region of interest) in stretch. With ROI off the whole image is stretched the same as in the box, ignoring the edges.
Reply With Quote
  #17  
Old 04-07-2022, 04:22 PM
Startrek (Martin)
Registered User

Startrek is offline
 
Join Date: Dec 2017
Location: Sydney and South Coast NSW
Posts: 5,990
Peter,
If your Flats and Bias have been captured correctly in APT and stacked in DSS accordingly, you will still have issues in Startools if you don’t follow the operational flow chart on the left
Since you are using a 150mm scope at focal length of 1050mm ( assuming no reducers etc.. ) with the 2600MC , your image scale should be as follows -
3.76 uM x 206.3 / 1050 = 0.73 arc sec per pixel which is oversampled
With oversampled data you are encouraged to Bin in Startools to trade off some resolution for noise reduction. A default Bin of 50% works really well and also has 2 added bonuses , your stars will remain round ( not blocky when you zoom in ) plus the image size is more manageable to work with. Your not waiting for the computer to process the task

As Fred mentioned , cropping your image for stacking artefacts and other anomalies is so important before you apply a Wipe and then second or final stretch of the image.

So you should follow the flow chart on the left in Startools -
AutoDev ( forget Film dev ) 1st stretch to inspect image
Bin default 50%
Crop edges or areas to remove stacking artefacts
Wipe default but bump up dark anomalies to say 4 pixels
Then second or final stretch with Autodev using a Region of Interest ROI ( drawn with your mouse around the main body of the object ) Adjust “Ignore Fine detail” slider and “Outside region of interest “ slider a bit to adjust detail and brightness across image

After that you just move on to Contrast , HDR , Sharp and so on ….. to complete your processing

Attached is an image of M17 I captured 2 weeks ago under Sydney’s horrible LP using 3 minute subs ( 3 hours of data ) , dithering , Gain 100 , cooled to - 10C , Flats and Bias stacked in DSS

First image is just loaded into Startools and AutoDev for inspection
2nd image is initial stretch in AutoDev then using Wipe without Bin or Cropping
3rd image is initially stretched in AutoDev , Bin , Cropped and Wiped
4th image is initially stretched in AutoDev, Bin , Cropped , Wiped and second stretch in AutoDev using ROI and sliders ( ignore fine detail 4 pixels , outside region of interest 10% )

You can see the mess you end up with if you use Wipe without cropping out artefacts and other anomalies

Hope the above helps

Cheers
Martin
Attached Thumbnails
Click for full-size image (530CE132-FED7-4049-9278-8944CE3A5B84.jpg)
198.0 KB60 views
Click for full-size image (0FF6FDCF-F693-4DE0-B2DE-1B2F5099C525.jpg)
195.9 KB44 views
Click for full-size image (535013F3-E7E9-4CBA-8E56-CB1873B98D8A.jpg)
199.2 KB50 views
Click for full-size image (2D262A2C-BA0B-40C1-85DB-C97D21A5BB6F.jpg)
195.1 KB54 views
Reply With Quote
  #18  
Old 04-07-2022, 04:46 PM
The_bluester's Avatar
The_bluester (Paul)
Registered User

The_bluester is offline
 
Join Date: Feb 2011
Location: Kilmore, Australia
Posts: 3,342
Regards the actual output files, I use Astro Pixel Processor, right up until the final export stage it saves the stacked output as 32 bit FITS, then I export then as 16 bit TIFF to work in Photoshop. I would love to stick to 32 bit files but you are straight away limited in some functionality within Photoshop and have to convert to 16 bit anyway.
Reply With Quote
  #19  
Old 04-07-2022, 04:55 PM
Bassnut's Avatar
Bassnut (Fred)
Narrowfield rules!

Bassnut is offline
 
Join Date: Nov 2006
Location: Torquay
Posts: 5,064
Quote:
Originally Posted by The_bluester View Post
Regards the actual output files, I use Astro Pixel Processor, right up until the final export stage it saves the stacked output as 32 bit FITS, then I export then as 16 bit TIFF to work in Photoshop. I would love to stick to 32 bit files but you are straight away limited in some functionality within Photoshop and have to convert to 16 bit anyway.
I think you missed the Startools stage. You export 32bit fits to startools then 16 bit tiff from that to PS?. This is standard. Once stretched, there is no advantage to 32 bit in PS, the dynamic range is already compressed to the point where 16bit is more than adequate.
Reply With Quote
  #20  
Old 04-07-2022, 07:09 PM
The_bluester's Avatar
The_bluester (Paul)
Registered User

The_bluester is offline
 
Join Date: Feb 2011
Location: Kilmore, Australia
Posts: 3,342
With Astro Pixel Processor I don't use Startools. 32 bit FITS of the stacks saved in APP then finally saved as 16 bit TIFF to take over to Photoshop for post production.

All stacking and assembly of an RGB image (RGB assembly for the mono cam) I do in APP, saving a combined RGB FITS as an RGB TIFF, and likewise Lum and narrowband FITS files saved as mono TIFF images to layer on the RGB image in photoshop. Everything else I do in PS.
Reply With Quote
Reply

Bookmarks

Thread Tools
Rate This Thread
Rate This Thread:

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off

Forum Jump


All times are GMT +10. The time is now 01:15 AM.

Powered by vBulletin Version 3.8.7 | Copyright ©2000 - 2024, Jelsoft Enterprises Ltd.
Advertisement
Testar
Advertisement
Bintel
Advertisement