Jump to content

Gradient correction


dav

Recommended Posts

Not really sure how to go about normalizing the color on this from corner to corner, but thought I'd throw it out there. I actually don't even know where to start with Photoshop or even know if it's the right tool...

AndromedaFinal.jpg

Link to comment
Share on other sites

ZWO ASI1600mm Pro, but I have to backpedal on that previous statement. The xisf was ever so slightly cropped to remove a band along the very bottom that had no data. Here is the original Ha channel output as jpg.Light_M31_480.0s_Bin1_H_gain132_20210719-041658_-10C_0009_1_c_lps_r.thumb.jpg.7db85e2ec8292a4a9963bc34ba0c6748.jpg

Edited by dav
Link to comment
Share on other sites

8 minutes ago, Damien Symonds said:

I'm confused.  Where did the colour go?

It's a mono camera. The single images are stacked for each channel(LRGB) and cropped to exclude sections with missing data and then combined into colored version. There's no gradient in the mono images. I posted the mono to show the cropped area.

 

 

Edited by dav
Link to comment
Share on other sites

The color should be more consistent throughout with the upper right. I believe the bottom left has discoloration from smoke lower on the horizon over some of the nights I was collecting the data. 

 

Link to comment
Share on other sites

The first image is the only color image. You can't make a color image from a single channel. Turn off Red and Green channels on an image in PS and you have a single Blue channel that is monochrome. I take many separate images in each channel, stack each channel, then map them to their respective channels to create the first image I posted. 

Link to comment
Share on other sites

I'll try and explain in a little better detail...

Multiple images are shot through a Red, Green, Blue or Luminance filter for RGB images. Narrowband images are shot through filters that only expose the sensor to narrow frequencies of light corresponding to the emission band they are capturing: Hydrogen alpha, Sulphur II and Oxygen III. This is what Hubble does quite frequently, creating a SHO image mapping SII to Red, Ha to Green and  OIII to Blue.

A single exposure is generally 5-12 minutes in length and I generally shoot 30 to as many as 100+ exposures through each filter. Each exposure by itself is virtually worthless and signal to noise ratio is rather poor. Stacking these images increases the ratio and creates a single monochromatic image for that specific channel. There are slight changes in framing of a few pixels from one frame to the next to eliminate "walking noise" in a process called Dithering and that results in the first dozen or so pixels around the border being devoid of usable data which has to be cropped before processing. I posted the mono image to show this effect after you asked about cropping.

Once each separate channel is processed in this way, they are mapped to their respective color channels to create a RGB image. It is this image that has color calibration and gradient issues and that is the what I need help with. I use specialized software called PixInsight to do all processing, but it is extremely complex and I'd like to use PhotoShop for some of the final processing if possible. 

Sorry for the confusion.

Here is an example of a well processed image of M31:

https://thumbor.forbes.com/thumbor/960x0/https%3A%2F%2Fspecials-images.forbesimg.com%2Fimageserve%2F612f383e9ab18a007b719776%2FAndromeda-Galaxy%2F960x0.jpg%3Ffit%3Dscale

 

Edited by dav
Link to comment
Share on other sites

Sorry, there's no reason to keep it in this case and a couple really good reasons not to. In simplifying the description of processing, I failed to mention a dynamic crop is applied to each channel before integration. The actual color image has not been cropped. You can even see the dead pixels on the bottom of the color image. This reduces processing time and the lost data is just not necessary for astro. I read the link you posted about cropping several times and understand the content, but it just doesn't apply to astro. In this instance, M31 fills the entire uncropped frame, so there is no usable data missing from the image.

Link to comment
Share on other sites

The latter. I'd just like more even color distribution that is more consistent with the top right.  The yellow/brown dust lanes are about right until they drop into bottom right frame and get overwhelmed by the yellow green.

 

Link to comment
Share on other sites

Ideally, I should probably just collect new data when the smoke clears, but this represents nearly 50  hours of exposure time and it would be great to salvage it.

 

Link to comment
Share on other sites

That looks pretty damn good. I shot new flats tonight and think the root cause might be there. There is a hard gradient running diagonally  corner to about center frame

 

Link to comment
Share on other sites

That's fine. It's not relevant to the course. Flats neutralize sensor and optical train anomalies such as dust specks and vignette. 

This is an older master luminance flat I used to calibrate the L channel of the image you fixed, but the new flat I shot last night looks drastically different. The black extends in a gradient from one corner to mid frame. Flats change over time and need to be redone, but it's great that the image can be salvaged to some degree in PS. I assume the 'how' of that process will become apparent to me as I continue with your classes.

Flat.thumb.JPG.39aea70674dd2a64500c3a3a1f58f9ff.JPG

Link to comment
Share on other sites

Thanks for taking a look and sharing the solution.  I'm still working, but will review calibration, get my monitors squared away, go again at WB and then see what I can do with your solution. Thanks again.

 

Link to comment
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now
×
×
  • Create New...