Jump to content

Banner.jpg.b89429c566825f6ab32bcafbada449c9.jpg

Which software/techniques for RGB colour balance?


Recommended Posts

I've been taking DSO pictures with my astro modified Canon 1100D for a few years now.  I think I've got a reasonable  workflow for image processing using DSS and Photoshop.  It's by no means perfect, but I'm reasonably happy with the images I produce.

However, I've recently bought an Atik 460EX mono CCD camera so that I can do RGB imaging.  At the moment I'm testing out the kit and how to use it.  I'm not yet attempting to produce finished images.  Over three nights I accumulated 3 x 10 x 300s exposures of M101 using red, green and blue filters.  The amount of high cloud and mist varied over the three nights, so the degree of contrast in the images for each filter differed partly due to these changes in weather patterns.

Last night I tried a quick and dirty stack of each set of images (no flats taken) and then imported the images into Photoshop.  I did some level and curves adjustment on each image to bring out some detail and tried to make sure the black point of each image was set to 40 and the white points were set to roughly similar values.  I then assigned each image to the relevant channel and was disappointed to see the resultant image had a strong red (not orange) cast.  This cast was prevalent in shadows, midtones and highlights.

I presume I should be aiming to process the uncombined images so that the colour balance in the combined image is already pretty well balanced right off the bat.  Is this correct?

If yes, what techniques should I be using to ensure this is the case?  Are there any photoshop techniques/actions or separate processing software that will help achieve this?

Thanks

 

Link to comment
Share on other sites

There is a great piece of software for this. It's called PixInsight. :icon_biggrin:

Combine the 3 images and process as colour image. There is a linear fit process to 'colour-align' the channels.

Link to comment
Share on other sites

I stack in AstroArt 5 then combine in the RGB Trichromy section. You can do a colour balance in that or use the separate tools for colour balance and contrast. But that's mainly because I've been too cheapskate to stump up for Pixinsight, an oversight I intend to remedy soon.

Link to comment
Share on other sites

I would calibrate the images with any Darks and Flats etc then stack the images. Next I would combine the RGB then once combined you can 'calibrate' the colour with as suggested above Pix-Insight or similar. Then you can adjust levels and curves and other processes.

 

Regards

Link to comment
Share on other sites

I've looked at Pixinsight already, but the total cost of getting it looks prohibitively expensive.  The 230 Euros + VAT for the software is just about doable.  However, reading the blurb about the specification of a machine recommended to run it seems to indicate that trying to run it on my Pentium dual core desktop with 2GB RAM running Windows 7 64 bit is just not going to be up to the job.  It also looks as though they won't be supporting Windows 7 for much longer.

https://pixinsight.com/sysreq/index.html

Buying a new computer that even comes close to meeting their recommended supported spec is going to push the total cost up to around £700 - too rich for me. :(

Link to comment
Share on other sites

1 hour ago, Stub Mandrel said:

 

Try running Gradient Exterminator on coarse and let it balance the image for you?

I already tried this with some success.  It balanced the background, but left the highlights and midtones still with a distinct colour cast.

Link to comment
Share on other sites

33 minutes ago, michaelmorris said:

I already tried this with some success.  It balanced the background, but left the highlights and midtones still with a distinct colour cast.

Shame, it works really well on my one shot RGB data, even if I don't have a gradient I often use it as a first step in making my colour channel.

Link to comment
Share on other sites

One standard approach is

1. Find a star with spectral type G2V that is at an altitude of 60 degree or more (to minimise differential atmospheric extinction per colour channel). I attach a list of relatively bright examples.

2. Collect exposures of the same duration in each of RGB, ensuring that the star is not saturated in any of them (stacking is fine)

3. Calibrate (darks etc) as normal, separately for each of RGB

4. Using the calibrated channels, measure the brightness values in the pixels at the centre of the G2V stellar image (averaging a few pixels at the centre is ok)

This gives you ADUs for each of RGB. Selecting one (say the highest) as a reference, compute the ratios with the other two colour channels. When imaging, you can either expose the channels with the least bright values for longer (i.e. multiply the exposure duration by the ratio of brightest channel to this channel), or use them in software to weight the contribution of the different channels.

This only needs to be done once for each sensor/filter combination :icon_biggrin: 

(description summarised from Charles Bracken's The Deep Sky Imaging Primer)

Another more approximate approach that I've used in the past is to take a shot of an open cluster with a known range of (B-V) colour values and attempt to match them. 

Cheers

Martin

 

Screen Shot 2017-03-29 at 20.00.19.png

Link to comment
Share on other sites

Hi michaelmorris.

I think it was a mistake to use curves on your individual rg and b images before combining.

This will have introduced changes to your mid-grey colour balance.

But in any case, why not go back into curves and adjust rg and b again for your desired colour balance?

Michael

Link to comment
Share on other sites

Archived

This topic is now archived and is closed to further replies.

  • Recently Browsing   0 members

    • No registered users viewing this page.
×
×
  • Create New...

Important Information

We have placed cookies on your device to help make this website better. You can adjust your cookie settings, otherwise we'll assume you're okay to continue. By using this site, you agree to our Terms of Use.