Jump to content

  •  

CNers have asked about a donation box for Cloudy Nights over the years, so here you go. Donation is not required by any means, so please enjoy your stay.

Photo

Color balance issues

  • Please log in to reply
18 replies to this topic

#1 descott12

descott12

    Apollo

  • -----
  • topic starter
  • Posts: 1,316
  • Joined: 28 Sep 2018
  • Loc: Charlotte, NC

Posted 10 November 2019 - 11:11 AM

Hello all,

I was hoping for some insight on my color balance as it seems way off.

This is with a 294 MC PRO on my Evo 8 + Hyperstar. This  is with an IDAS LP P2 filter in place.

You will notice that my red peak is always way to the left of the others. This occurs when I have the color balance set to AUTO or manually adjusted all the way up as is shown in the screenshot.

Any thoughts? I have been struggling with this for awhile (I think with or without the filter) and I can often mitigate it with the vertical RGB sliders in the stacking window but it seems like something is not quite right.

Thanks
Dave

Attached Thumbnails

  • colorbalance2.png


#2 Forward Scatter

Forward Scatter

    Explorer 1

  • -----
  • Posts: 97
  • Joined: 22 Jul 2018
  • Loc: Wandering the PNW

Posted 10 November 2019 - 11:41 AM

Hi David

Robin does explain the difference between the two sets of sliders in the SharpCap Forum:

https://forums.sharp...topic.php?t=573



#3 pyrasanth

pyrasanth

    Surveyor 1

  • *****
  • Posts: 1,983
  • Joined: 08 Jan 2016

Posted 10 November 2019 - 12:39 PM

I use one of the earlier generation IDAS filters with the QHY168C camera and find it a really good filter. I did get the issue you describe where channels seemed to be well out of what I would consider the normal range with an Atik OSC camera which I no longer own. However I don't use your software only PI and find once I've run background neutralisation and colour calibration the channels are pretty much very well aligned.

 

Sadly some sensor & filter combinations play better together and you might have found a combination that does not but PI is very good at correcting these types of issues. I'm not by any means advocating that you switch software as we all need to work issues within our chosen work flow however giving PI a whirl would not hurt and may well be a way forward to resolve your calibration issue.


Edited by pyrasanth, 10 November 2019 - 12:42 PM.


#4 descott12

descott12

    Apollo

  • -----
  • topic starter
  • Posts: 1,316
  • Joined: 28 Sep 2018
  • Loc: Charlotte, NC

Posted 10 November 2019 - 12:58 PM

I use one of the earlier generation IDAS filters with the QHY168C camera and find it a really good filter. I did get the issue you describe where channels seemed to be well out of what I would consider the normal range with an Atik OSC camera which I no longer own. However I don't use your software only PI and find once I've run background neutralisation and colour calibration the channels are pretty much very well aligned.

 

Sadly some sensor & filter combinations play better together and you might have found a combination that does not but PI is very good at correcting these types of issues. I'm not by any means advocating that you switch software as we all need to work issues within our chosen work flow however giving PI a whirl would not hurt and may well be a way forward to resolve your calibration issue.

Thanks for the suggestion. But unfortunately, I only do EAA. All live with no post-processing.  I think I need to test this again without the filter and see what it looks like.


  • mikenoname likes this

#5 DSO_Viewer

DSO_Viewer

    Viking 1

  • -----
  • Posts: 955
  • Joined: 03 Jan 2016

Posted 10 November 2019 - 01:12 PM

Hello all,

I was hoping for some insight on my color balance as it seems way off.

This is with a 294 MC PRO on my Evo 8 + Hyperstar. This  is with an IDAS LP P2 filter in place.

You will notice that my red peak is always way to the left of the others. This occurs when I have the color balance set to AUTO or manually adjusted all the way up as is shown in the screenshot.

Any thoughts? I have been struggling with this for awhile (I think with or without the filter) and I can often mitigate it with the vertical RGB sliders in the stacking window but it seems like something is not quite right.

Thanks
Dave

David what I see the guys doing on NSN.com is with their 294 cameras and sharpcap, they capture one image with no stacking using the same exposure, gain brightness/offset etc. and look at the histogram RGB numbers. They then adjust the red and blue sliders until all three numbers are close. Once achieved they start the live stacking and then hit the auto stretch on the live histogram followed by the auto color balance stretch. It seems to work out very good and their results are fantastic. Go and watch them on NSN, I have learn't so much from watching over there.

 

Steve


  • saguaro likes this

#6 saguaro

saguaro

    Apollo

  • *****
  • Posts: 1,029
  • Joined: 19 Apr 2012
  • Loc: Southern Arizona

Posted 10 November 2019 - 01:16 PM

Did you use a master dark during live stacking? Even if you didn’t, this topic I started in the SC forums may be helpful: https://forums.sharp...opic.php?t=2050


  • DSO_Viewer likes this

#7 descott12

descott12

    Apollo

  • -----
  • topic starter
  • Posts: 1,316
  • Joined: 28 Sep 2018
  • Loc: Charlotte, NC

Posted 10 November 2019 - 01:43 PM

Did you use a master dark during live stacking? Even if you didn’t, this topic I started in the SC forums may be helpful: https://forums.sharp...opic.php?t=2050

I did not. I have been lazy about that stuff lately. But actually I use flats occasionally but have never used darks and I don't recall having this problem, at least not as bad, in the past. Thanks for the link



#8 descott12

descott12

    Apollo

  • -----
  • topic starter
  • Posts: 1,316
  • Joined: 28 Sep 2018
  • Loc: Charlotte, NC

Posted 10 November 2019 - 01:44 PM

David what I see the guys doing on NSN.com is with their 294 cameras and sharpcap, they capture one image with no stacking using the same exposure, gain brightness/offset etc. and look at the histogram RGB numbers. They then adjust the red and blue sliders until all three numbers are close. Once achieved they start the live stacking and then hit the auto stretch on the live histogram followed by the auto color balance stretch. It seems to work out very good and their results are fantastic. Go and watch them on NSN, I have learn't so much from watching over there.

 

Steve

Hey Steve,

Yes, I will check that out. I haven't been able to view NSN in the past due to the need for some special browser and I have just put it off but that is a very good idea.


  • DSO_Viewer likes this

#9 DSO_Viewer

DSO_Viewer

    Viking 1

  • -----
  • Posts: 955
  • Joined: 03 Jan 2016

Posted 10 November 2019 - 09:40 PM

Hey Steve,

Yes, I will check that out. I haven't been able to view NSN in the past due to the need for some special browser and I have just put it off but that is a very good idea.

Dave, I think from what I have installed is Chromium Portable for both to hear the people signed in talk and to communicate back to them. Here is the link

 

https://sourceforge....cts/crportable/

 

Have a good day

 

Steve



#10 descott12

descott12

    Apollo

  • -----
  • topic starter
  • Posts: 1,316
  • Joined: 28 Sep 2018
  • Loc: Charlotte, NC

Posted 10 November 2019 - 09:47 PM

Dave, I think from what I have installed is Chromium Portable for both to hear the people signed in talk and to communicate back to them. Here is the link

 

https://sourceforge....cts/crportable/

 

Have a good day

 

Steve

Thanks. I was going to search for that as I couldn't remember the name of the browser.


  • DSO_Viewer likes this

#11 GaryShaw

GaryShaw

    Ranger 4

  • *****
  • Posts: 357
  • Joined: 28 Nov 2017
  • Loc: Boston

Posted 10 November 2019 - 11:38 PM

David what I see the guys doing on NSN.com is with their 294 cameras and sharpcap, they capture one image with no stacking using the same exposure, gain brightness/offset etc. and look at the histogram RGB numbers. They then adjust the red and blue sliders until all three numbers are close. Once achieved they start the live stacking and then hit the auto stretch on the live histogram followed by the auto color balance stretch. It seems to work out very good and their results are fantastic. Go and watch them on NSN, I have learn't so much from watching over there.

 

Steve

Hi Steve

I have similar color balance problems as Dave’s and I don’t quite follow the procedure you outlined above. Could you please clarify it a bit. There are 3 histograms so which one do you first refer to? Also, there are 4 sliders on the live stack histogram panel and 2 sliders in the image controls panel on the right so which red and blue sliders are they adjusting and which 3 numbers are they trying to get to be close? Sorry if I’m being thick but, no matter how many times I read it, I just can’t follow what you’re suggesting. Thanks 

Gary


Edited by GaryShaw, 10 November 2019 - 11:41 PM.


#12 DSO_Viewer

DSO_Viewer

    Viking 1

  • -----
  • Posts: 955
  • Joined: 03 Jan 2016

Posted 11 November 2019 - 10:30 PM

Hi Steve

I have similar color balance problems as Dave’s and I don’t quite follow the procedure you outlined above. Could you please clarify it a bit. There are 3 histograms so which one do you first refer to? Also, there are 4 sliders on the live stack histogram panel and 2 sliders in the image controls panel on the right so which red and blue sliders are they adjusting and which 3 numbers are they trying to get to be close? Sorry if I’m being thick but, no matter how many times I read it, I just can’t follow what you’re suggesting. Thanks 

Gary

Gary, I wrote "they capture one image with no stacking using the same exposure, gain brightness/offset etc", so this eliminates the live stacking histogram and 4 color sliders. Only the single image histogram has RGB median number counts and to get this you select the green histogram tab at the top middle right of Sharpcap. Adjust the only two color sliders red & blue in the image control until the numbers are close to each other.

 

Steve



#13 descott12

descott12

    Apollo

  • -----
  • topic starter
  • Posts: 1,316
  • Joined: 28 Sep 2018
  • Loc: Charlotte, NC

Posted 15 November 2019 - 09:05 PM

I just read a blurb about the IDAS LPS D1. 

 

"New revised LPS-D1 design improves on the popular LPS-P2 filter by reducing reddish background cast, improving the color rendition for one-shot-color CCDs..."  from Starizona's web page

 

so possibly my red problems ARE coming from the filter. Although I haven't had a chance to test it.

 

Very interesting. The D1 was on sale when I bought the P2. Not sure why I chose the way I did...Oh well.


Edited by descott12, 16 November 2019 - 09:48 AM.


#14 Clouzot

Clouzot

    Explorer 1

  • -----
  • Posts: 92
  • Joined: 09 Jul 2018
  • Loc: South of France

Posted 16 November 2019 - 02:42 AM

With the L-Enhance, as well as the LPS-D2, I see the same issue with the red "bump" being put to the left of the display histogram (small histogram on the right hand side of Sharpcap's window.

 

 

The method I (and a lot of others) use for color balancing is to look at the small display histogram (on the right), see how RGB curves differ from each other, and try to align them using the LiveStack histogram sliders. Auto balancing almost never works in this case.



#15 selfo

selfo

    Vostok 1

  • *****
  • Posts: 198
  • Joined: 18 Oct 2015
  • Loc: Boulder CO

Posted 16 November 2019 - 06:25 PM

Hi Dave:

 

I never touch the Display Histogram in the lower right but I use the color sliders to the left to get the red,blue,green Histogram to overlap as best I can.  When I do that the color sliders are much lower down than I see in your screenshot and my color balance is usually adjusted properly unless I am using a narrowband filter which may require additional manual adjustment. . After this I usually have to tweak the Mid level and Black level to get a good exposure. Last step... Save exactly as seen in SC.   Your color slider seems a bit high.  Maybe try increasing your total exposure time to bring out the nebulosity and drop the color slider down that may help with the color balance as well

 

Cheers

 

Stewart


  • descott12 likes this

#16 tmaestro

tmaestro

    Messenger

  • *****
  • Posts: 483
  • Joined: 22 Jul 2015
  • Loc: West of Houston, Texas

Posted 19 November 2019 - 10:50 PM

I had an issue with my C6/Hyperstar, ASI224MC, and Astronomik CLS CCD giving me a color cast that wouldn't adjust out.  It seemed to happen as I was experimenting more broadly with this SharpCap histogram and that, and varying exposure lengths (and new darks taken to match), so my problem is likely having screwy color sliders while taking darks...  

 

but it got me thinking, shouldn't there be an objectively *correct* color adjustment?  If you're looking at a red object, your histogram peaks *shouldn't* be aligned, right?

 

I haven't done this yet, but I was considering getting a large gray photographic white balance card and taking a shot at various settings to see what the histogram looks like when the data is absolutely neutral and the histogram peaks SHOULD be aligned.  Then I can make note of the adjustments I need to make to compensate for the filter and such and just apply those...  seems like that would be objectively correct no matter what.

 

plenty of experimenting to do


  • mikenoname and GaryShaw like this

#17 GaryShaw

GaryShaw

    Ranger 4

  • *****
  • Posts: 357
  • Joined: 28 Nov 2017
  • Loc: Boston

Posted 20 November 2019 - 09:29 AM

Great idea - I’m all for experimenting and sharing whatever emerges. Given the importance of accurate image hues and saturations, I’ve wondered why the SC Display Histogram isn’t larger and easier to view for aligning the peaks. I often struggle to even see where the 4 curves are. It’s crossed my mind before and your post has nudged me to raise the question with Robin Glover in the SC User Forum. 
cheers

Gary



#18 mikenoname

mikenoname

    Viking 1

  • *****
  • Posts: 570
  • Joined: 29 Jan 2015
  • Loc: Death Valley Region (Bortle 2 - SQM 21.9+)

Posted 20 November 2019 - 01:09 PM

I haven't done this yet, but I was considering getting a large gray photographic white balance card and taking a shot at various settings to see what the histogram looks like when the data is absolutely neutral and the histogram peaks SHOULD be aligned.  Then I can make note of the adjustments I need to make to compensate for the filter and such and just apply those...  seems like that would be objectively correct no matter what.

 

Will be interested to see what results you come up with!


  • tmaestro likes this

#19 Howie1

Howie1

    Ranger 4

  • *****
  • Posts: 309
  • Joined: 22 May 2013

Posted 20 November 2019 - 09:21 PM

 

 

I haven't done this yet, but I was considering getting a large gray photographic white balance card and taking a shot at various settings to see what the histogram looks like when the data is absolutely neutral and the histogram peaks SHOULD be aligned.  Then I can make note of the adjustments I need to make to compensate for the filter and such and just apply those...  seems like that would be objectively correct no matter what.

 

Jerry Lodriguss (award winning photographer as well as award winning astrophotographer) says you should set color balance both with and without filters during the day using a digital grey card. IE develop Custom white balance / color balance settings for no filter, camera + filter A, camera + filter B, etc. At night set WB to the custom one applicable for whatever is in the image train .... camera + no filter, camera + A, etc. Then away you go, start imaging. NOTE you will still have to slightly tweak the WB as you move for object to object. The CWB is a GREAT starting point and greatly speeds up getting a good EAA experience, but obviously in different parts of the sky with different LP, dust, colors of the object you will still have to make slight tweaking. 

 

You only have to develop the table once. You don't do this every day  you decide to head out. 

 

I'll put how to do it with DSLR and with ZWO astro cam. First the DSLR ...

 

His advice works perfectly for my DSLR's. Set the ISO you are going to use at night, then set AV and shoot the grey card. Set that image taken as a Custom White Balance (CWB) within the camera. Put on filter A and repeat ending up with an image stored on your SD card for use as a CWB within the camera for no filter, filter A, filter B and so on. At night just select which image for that particular filter applies and set it as the CWB to use. Done!

 

Now two methods with ZWO's ...

 

For my ZWO's, set the typical gain you are going to use at night, point at the grey card and select Auto for the whitebalance in sharpcap and then note where the RGB (or RB) sliders values automatically adjust to. NOTE this is NOT stacked frames (no stars to stack!) but just the single shot image preview window and its histogram. Then add filter A and so on resulting in a table of RGB / RB values for camera + filter A, and so on. At night, set the RGB / RB sliders so the value in sharpcap is as per what you noted down in the digital greyscale tests and start imaging. Add filter A, set those values and so on. 

 

But, I have found it not so good for my ZWO's. Why? I think the auto WB in SharpCap may not quite get it right when it see's a solid grey image on the view window? Or maybe its my old version of SC? However, it may work for you so I suggest trying your ZWO first with the method above. If it works for you then away you go. If, like me it doesnt work so well for the ZWO, then try my method two down below. 

 

For my ZWO's (again you dont have to do this everytime you go out ... just pick a day when you are bored and shoot a whole bunch of CWB settings and note them down). I setup in the daytime around an hour before sunset, and point the scope and camera at a typical imaging gain at a distant scene in the North or South (IE away from setting Sun bright West and away from highly illuminated stuff over in the East). The scene should preferably contain bits of blue sky, green grass, red roof of a house or if no red maybe a brown tree trunk, brown bricks .... etc ... get the drift .... something on the horizon way down the street etc with some B G and Red/brown. Without stacking - using just the single preview screen in sharpcap - click on the Auto WB checkbox in SC.. If the Auto setting looks fine then note down the RGB / RB values and use them. If the auto looks bad, then adjust RGB / RB until the blue sky, green stuff, red / brown stuff looks about as natural color balanced as you can see with eyeballs out the window/down the street. Then note down the RGB numbers for that setting. Then I stick on filter A and do the same. 

 

As earlier, then at night, set the RGB / RB sliders so the value in sharpcap is as per what you noted down in the down-the-road-scenery tests and start imaging. Add filter A, set those values and so on.

 

Again, they are very good and quick starting points, you will still have to make slight tweaks as you slew all over the sky and to different objects.

 

Cheers


Edited by Howie1, 20 November 2019 - 09:26 PM.

  • tmaestro likes this


CNers have asked about a donation box for Cloudy Nights over the years, so here you go. Donation is not required by any means, so please enjoy your stay.


Recent Topics






Cloudy Nights LLC
Cloudy Nights Sponsor: Astronomics