Return to the Cloudy Nights Telescope Reviews home pageAstronomics discounts for Cloudy Nights members
· Get a Cloudy Nights T-Shirt · Submit a Review / Article

Click here if you are having trouble logging into the forums

Privacy Policy | Please read our Terms of Service | Signup and Troubleshooting FAQ | Problems? PM a Red or a Green Gu… uh, User

Astrophotography and Sketching >> DSLR & Digital Camera Astro Imaging & Processing

Pages: 1 | 2 | 3 | 4 | 5 | (show all)
Mike Unsold
Vendor, MLUnsold - ImagesPlus
*****

Reged: 05/21/09

Re: ImagesPlus+Photoshop Astro Image Processing new [Re: mmalik]
      #5458269 - 10/06/12 06:09 PM

Mike

Apply digital development at the same point as shown in your high level workflow. Just apply it so that

1) The faint background is shown and not clipped

2) Bright areas of the Bubble are not too bright

3) Star have some color


Post Extras: Print Post   Remind Me!   Notify Moderator  
mmalik
Postmaster
*****

Reged: 01/13/12

Loc: USA
Re: ImagesPlus+Photoshop Astro Image Processing new [Re: Mike Unsold]
      #5458784 - 10/07/12 04:14 AM

Mike, I have updated the document...; please review.

Note: There seems to be a bit of confusion about DD; dark background image you saw in the doc was from the previous step of 'Auto' button click in DD, and before actual DD steps were performed on it. I have added a new pic to reflect post-DD look of the image. Please let me know if all the steps now make sense in light of this clarification?


Post Extras: Print Post   Remind Me!   Notify Moderator  
Mike Unsold
Vendor, MLUnsold - ImagesPlus
*****

Reged: 05/21/09

Re: ImagesPlus+Photoshop Astro Image Processing new [Re: mmalik]
      #5458939 - 10/07/12 08:51 AM

Mike

The updated text at step 37 looks good.

The auto DDP button is at best a rough guess.

If the image above the updated step 37 is after DDP then it still has way too dark background.

If the image shown above step 38 is the after DDP image then it is better and has some star color but I bet there is still a lot of nebula in the background that is not shown.

Post the 16 or 32 bit stacked and otherwise unprocessed image and I will set the DDP parameters for you.

The image shown above step 38 - Multi Point Flatten shows no sign of a gradient and does not need to have its background smoothed so no I would not use the steps that follow.

Gradients and background smoothing become more important the more you stretch the background to show faint detail. After the best DDP is applied you can the see if the next steps are required.


Post Extras: Print Post   Remind Me!   Notify Moderator  
mmalik
Postmaster
*****

Reged: 01/13/12

Loc: USA
Re: ImagesPlus+Photoshop Astro Image Processing new [Re: Mike Unsold]
      #5459580 - 10/07/12 05:01 PM

Quote:

If the image above the updated step 37 is after DDP then it still has way too dark background.

If the image shown above step 38 is the after DDP image then it is better and has some star color but I bet there is still a lot of nebula in the background that is not shown.

Post the 16 or 32 bit stacked and otherwise unprocessed image and I will set the DDP parameters for you.

The image shown above step 38 - Multi Point Flatten shows no sign of a gradient and does not need to have its background smoothed so no I would not use the steps that follow.




Mike, image above step 38 is "post-DDP".

Combined/unprocessed image is here...; please see if you can describe proper settings for DDP, and beyond. Thx


Post Extras: Print Post   Remind Me!   Notify Moderator  
mmalik
Postmaster
*****

Reged: 01/13/12

Loc: USA
Re: ImagesPlus+Photoshop Astro Image Processing new [Re: Mike Unsold]
      #5461387 - 10/08/12 10:40 PM

Quote:

With ImagesPlus 5 you would use a star mask with the selective HSL color balance tool to adjust star color and saturation and overall color in 3 steps.

http://www.mlunsold.com/Temp/IrisReproc.jpg




Mike, this is from the 'Iris Nebula' thread; thought I'll ask here to keep all documentation in one place. Would you be able to expand upon the selective HSL color balance tool and provide some some granular details how to use it and where it fits in our workflow? Thx

Note: I have uploaded the combined/unprocessed FIT of ‘Iris Nebula’ here...


Post Extras: Print Post   Remind Me!   Notify Moderator  
srosenfraz
Post Laureate
*****

Reged: 03/06/11

Loc: United States
Re: ImagesPlus+Photoshop Astro Image Processing new [Re: mmalik]
      #5461517 - 10/09/12 01:16 AM

From a different thread (Mike's Flaming Star Nebula Reprocessing):

Quote:

Thanks Scott!

On a side note, what are your thoughts about the new workflow process? It was put together in consultation with Mike Unsold as you know; not sure if you had a chance to review it? Your feedback will be appreciated.




Hi Mike -

Sorry, but I've been a bit busier than normal lately, so I haven't been able to keep up on this thread. I haven't had a chance to review the entire document, but I took a quick look at the first section. You have as your first steps Raw conversion, normalizing, aligning, stacking, etc. I see that you're doing all of these steps "manually".

I've always used IP's Image Set Operations | Automatic DSLR, OSC CCD, CCD Image Set Processing, as its completely automated and works extremely well. I don't know if there's any advantage to using your manual method, but it feels like too much work for my taste. The automated image set processing ends up doing the raw conversion, normalizes the images, calibrates them, generates master dark, flat, and bias frames, aligns, stacks, and combines the image without hardly any user intervention.

Mike Unsold may be able to elaborate as to whether there are reasons to use the automated versus manual processes. As I said, for my taste, the automated processing works and is easy.

The basic steps I use for automated processing are:

Go to the Light Frames tab and select all of your light frames. Do the same on the Dark Frames tab (obviously selecting your darks), Flat Frame tab, Flat Darks tab (if you have them) and Bias Frames tab. Then, go back to the Process tab. Select a folder to save your calibrated files (I usually create one named "Calibrated Frames").

Next, in the Perform after calibration and color conversion section, check "minimize hard drive use" (this deletes the interim files and leaves you only with converted, calibrated files - if you want the interim files, do not check this box). In the same section, check the normalize, align, and combine boxes. Personally, I do NOT check the grade box. The grade box analyzes the images and decides which ones work best. The grading DOES work and is generally good to use. However, I find
that it throws out more of my subs than I'm willing to sacrifice. For my purposes, I just review my raw light frames BEFORE I put them into IP (Picasa's Photo viewer works well for viewing RAW CR2s) and look for subs that have obvious guiding errors, major airplane trails and other big issues. Its not as mathematical as using the grading function in IP, but I'm able to utilize more of my subs and haven't seen any resulting image degradation.

Leave all settings at their defaults - you can go nuts trying to optimize the process to eventually achieve some trivial improvement. The only settings I occassionally use are in the "Light Frame Processing parameters" box. If you're combining image that were taken on either side of the meridian, you'll need to check the Meridian flip box (so that IP knows to try to flip the images for alignment). Every now and then, when I have perfect conditions and really tight focus, I have to check the "Short FL - small stars" box to stack my images (IP trys to stack the image, fails and then issues a recommendation to try this setting).

After you have these handful of boxes filled in, click the Process button and take a nap. IP will take care of all the ugly details for you and leave you with a nice and pretty calibrated and stacked combined image.

I'll try to look at the rest of your workflow in the next few days. It looks like you've made some great progress with your images - keep up the good work.


Post Extras: Print Post   Remind Me!   Notify Moderator  
mmalik
Postmaster
*****

Reged: 01/13/12

Loc: USA
Re: ImagesPlus+Photoshop Astro Image Processing new [Re: srosenfraz]
      #5462221 - 10/09/12 02:21 PM

Thanks Scott; I see you use the automated approach. I'll see if I can get that incorporated into the doc. I'll wait to hear from you as you review the rest of the doc.

On a side note, since I don't take flats, can it be compensated in Photoshop; if yes, how is it done in Photoshop? I think there is a way to do it, but not much familiar yet with PS. Scope I use is huge and taking flats everytime is not as convenient as it may seem. Thx


Post Extras: Print Post   Remind Me!   Notify Moderator  
McConkey
sage


Reged: 05/11/12

Loc: Sherwood Park, Alberta
Re: ImagesPlus+Photoshop Astro Image Processing new [Re: mmalik]
      #5462357 - 10/09/12 03:34 PM

I use this for my artificial flats

http://www.corius.net/cont/artic/artiflat.html


Post Extras: Print Post   Remind Me!   Notify Moderator  
srosenfraz
Post Laureate
*****

Reged: 03/06/11

Loc: United States
Re: ImagesPlus+Photoshop Astro Image Processing new [Re: mmalik]
      #5463144 - 10/10/12 12:00 AM

Quote:


On a side note, since I don't take flats, can it be compensated in Photoshop; if yes, how is it done in Photoshop? I think there is a way to do it, but not much familiar yet with PS. Scope I use is huge and taking flats everytime is not as convenient as it may seem. Thx





Hi Mike - its possible to do some compensation in Photoshop, but its not near as effective as taking flats. The issue with flats is that it'll correct for vignetting as well as dust motes on your sensor. If you look at what flat fielding does, you can see why other forms of compensation don't do the same.

Imagine a handful of pixels on your sensor that are towards the edge of the chip and in the vignetted area of the scope. Each of these pixels is going to receive less light than the pixels in the center of the FOV. Let's say that these pixels receive 2/3 of the light that the center pixels receive. Your flats are used in calibration to increase the values recorded by each of those pixels by 3/2 (the inverse of 2/3). In this way the values for these pixels are adjusted to make these pixels more accurately reflect what they would have received if your optics were not vignetted. So, stars that are dimmer than they should be will be brighter, and background sky that is darker than it should be will be brighter as well.

I'm not aware of any flat fielding techniques in Photoshop that do the same thing - multiply pixel values to bring them up to the correct adjusted values. Most flat fielding techniques will take the background and even out the associated gradient so that the background doesn't show obvious vignetting. However, these techniques don't do anything to brighten the non-background areas (stars, nebulosity, etc.). By the same token, they don't do anything to correct for darkened pixels due to dust on the sensor.

As such, I would encourage you to take flats - even if it isn't easy. You'll have much better results than trying to retroactively correct using Photoshop's (or other software's) flat fielding techniques.

That having been said, one of the best techniques in Photoshop for dealing with uneven illumination is using Russell Croman's Gradient Xterminator. Now, Russell will tell you explicitly that Gradient Xterminator is not designed for dealing with vignetting (for the reasons I describe above), but it is quite effective in reducing/eliminating gradients caused by vignetting. If you use Gradient Xterminator, be sure to first read the Gradient Xterminator Tutorial. The method described on this page is what I use to correct gradients in my images (i.e., those gradients not caused by vignetting - I use flats for that purpose).

The other common method in Photoshop for dealing with gradients and non-flat fields is to:

1) Duplicate the image

2) Apply a large radius Gaussian Blur (or use a combination of Dust and Scratch filters and Gaussian Blurs). Sometimes this is done after attempting to clone out any of the primary object (so that the target more or less matches the surrounding background). The idea here is that you're trying to model the background illumination.

3) Subtract the duplicated image from the original. Usually when you do this you'll want to apply an offset (about 30 usually works). What this does is to subtract the averaged background across the image and then add 30 to each of those values. In that way you ensure that you haven't clipped your black point.

The problem with this basic technique is that on many images (particularly those covered in nebulosity), you'll tend to subtract the faint nebulosity that you're trying to enhance. So, it'll flatten the field, but there may be a price for this. For this reason, I definitely prefer Gradient Xterminator for dealing with gradients (and/or flattening background).

Hope this helps.


Post Extras: Print Post   Remind Me!   Notify Moderator  
mmalik
Postmaster
*****

Reged: 01/13/12

Loc: USA
Re: ImagesPlus+Photoshop Astro Image Processing new [Re: srosenfraz]
      #5463254 - 10/10/12 02:01 AM

Thanks Karl/Scott; both of your Photoshop techniques for flats seem similar, appreciate the feedback.

Scott, if you can explain briefly, how to take a flat on a fairly large scope during the night; my most sessions start in darkness and end in darkness. What are optimal ISO/shutter settings for a flat? Also, do I have to take a flat for a given session or can I use one for other day’s session/s?


Post Extras: Print Post   Remind Me!   Notify Moderator  
Mike Unsold
Vendor, MLUnsold - ImagesPlus
*****

Reged: 05/21/09

Re: ImagesPlus+Photoshop Astro Image Processing new [Re: McConkey]
      #5463661 - 10/10/12 11:25 AM

Exact same steps

1) Make a duplicate layer of the original light image
2) Apply a Median filter to this copy
3) Remove any left over nebulousity and star images from this image
4) Blur the image
5) Subtract the artificial flat field from the light

for one way of using using ImagesPlus to make a synthetic flat. Mike M. you should really consider using a real master flat since it measures real vignetting.

Mike


Post Extras: Print Post   Remind Me!   Notify Moderator  
Mike Unsold
Vendor, MLUnsold - ImagesPlus
*****

Reged: 05/21/09

Re: ImagesPlus+Photoshop Astro Image Processing new [Re: Mike Unsold]
      #5463733 - 10/10/12 12:04 PM

For ImagesPlus 5.0 x32 or x64 image processing a cookbook that shows how to make and modify LRGB luminance,
custom luminance, selected HSL range, star, edge, area, threshold, painted, and inverse masks of all these masks is posted at

http://www.mlunsold.com/process/IP5/PH-BAFIME/PH-BAFIME.html

The masks are used with either the Process History layers or Combine Images & Mosaic layers.

Basic steps for using Process History filter layers with blend modes, opacity, and masks are shown in this tutorial

http://www.mlunsold.com/process/IP5/PH-BAFIM/PH-BAFIM.html


Basic steps for using Combine Images & Mosaic layers with filters, blend modes, opacity, and masks are shown in this tutorial

http://www.mlunsold.com/process/IP5/CIQG-AF/CIQG-AF.html


The main IP image processing tutorial page index by function type

http://www.mlunsold.com/ILProcessing.html

has been updated to include the new mask creation cookbook and basic steps for using Process History filter layers and Combine Images & Mosaic layers with masks.


Post Extras: Print Post   Remind Me!   Notify Moderator  
Mike Unsold
Vendor, MLUnsold - ImagesPlus
*****

Reged: 05/21/09

Re: ImagesPlus+Photoshop Astro Image Processing new [Re: srosenfraz]
      #5463788 - 10/10/12 12:45 PM

Quote:

From a different thread (Mike's Flaming Star Nebula Reprocessing):

Quote:

Thanks Scott!

On a side note, what are your thoughts about the new workflow process? It was put together in consultation with Mike Unsold as you know; not sure if you had a chance to review it? Your feedback will be appreciated.




Hi Mike -

Sorry, but I've been a bit busier than normal lately, so I haven't been able to keep up on this thread. I haven't had a chance to review the entire document, but I took a quick look at the first section. You have as your first steps Raw conversion, normalizing, aligning, stacking, etc. I see that you're doing all of these steps "manually".

I've always used IP's Image Set Operations | Automatic DSLR, OSC CCD, CCD Image Set Processing, as its completely automated and works extremely well. I don't know if there's any advantage to using your manual method, but it feels like too much work for my taste. The automated image set processing ends up doing the raw conversion, normalizes the images, calibrates them, generates master dark, flat, and bias frames, aligns, stacks, and combines the image without hardly any user intervention.

Mike Unsold may be able to elaborate as to whether there are reasons to use the automated versus manual processes. As I said, for my taste, the automated processing works and is easy.

The basic steps I use for automated processing are:

Go to the Light Frames tab and select all of your light frames. Do the same on the Dark Frames tab (obviously selecting your darks), Flat Frame tab, Flat Darks tab (if you have them) and Bias Frames tab. Then, go back to the Process tab. Select a folder to save your calibrated files (I usually create one named "Calibrated Frames").

Next, in the Perform after calibration and color conversion section, check "minimize hard drive use" (this deletes the interim files and leaves you only with converted, calibrated files - if you want the interim files, do not check this box). In the same section, check the normalize, align, and combine boxes. Personally, I do NOT check the grade box. The grade box analyzes the images and decides which ones work best. The grading DOES work and is generally good to use. However, I find
that it throws out more of my subs than I'm willing to sacrifice. For my purposes, I just review my raw light frames BEFORE I put them into IP (Picasa's Photo viewer works well for viewing RAW CR2s) and look for subs that have obvious guiding errors, major airplane trails and other big issues. Its not as mathematical as using the grading function in IP, but I'm able to utilize more of my subs and haven't seen any resulting image degradation.

Leave all settings at their defaults - you can go nuts trying to optimize the process to eventually achieve some trivial improvement. The only settings I occassionally use are in the "Light Frame Processing parameters" box. If you're combining image that were taken on either side of the meridian, you'll need to check the Meridian flip box (so that IP knows to try to flip the images for alignment). Every now and then, when I have perfect conditions and really tight focus, I have to check the "Short FL - small stars" box to stack my images (IP trys to stack the image, fails and then issues a recommendation to try this setting).

After you have these handful of boxes filled in, click the Process button and take a nap. IP will take care of all the ugly details for you and leave you with a nice and pretty calibrated and stacked combined image.

I'll try to look at the rest of your workflow in the next few days. It looks like you've made some great progress with your images - keep up the good work.




Using the manual tools in ImagesPlus to convert CR2, NEF,... and calibrate, convert to color, align, and stack does teach you what each step of the auto image set processing is doing. As a next step towards automation I would select all the CR2 light, flat, bias, and dark frames in the auto image set process tool and check only Normalize. Auto image set will convert all CR2 to Bayer format, build calibration frames, calibrate, convert calibrated frames to color, and normalize the color frames. Now do the same pick a star or two align followed by combine. After auto image set process finishes normalize press the Align TSR button to setup the align tool with all of the normalized light frames.


Post Extras: Print Post   Remind Me!   Notify Moderator  
Mike Unsold
Vendor, MLUnsold - ImagesPlus
*****

Reged: 05/21/09

Re: ImagesPlus+Photoshop Astro Image Processing new [Re: Mike Unsold]
      #5463877 - 10/10/12 01:45 PM

Quote:

Exact same steps

1) Make a duplicate layer of the original light image
2) Apply a Median filter to this copy
3) Remove any left over nebulousity and star images from this image
4) Blur the image
5) Subtract the artificial flat field from the light

for one way of using using ImagesPlus to make a synthetic flat. Mike M. you should really consider using a real master flat since it measures real vignetting.

Mike




Flat calibration uses division and not subtraction to apply the flat frame. Problem with subtraction is that it can destroy data.

Mike


Post Extras: Print Post   Remind Me!   Notify Moderator  
srosenfraz
Post Laureate
*****

Reged: 03/06/11

Loc: United States
Re: ImagesPlus+Photoshop Astro Image Processing new [Re: Mike Unsold]
      #5464753 - 10/11/12 01:24 AM

Quote:

Thanks Karl/Scott; both of your Photoshop techniques for flats seem similar, appreciate the feedback.

Scott, if you can explain briefly, how to take a flat on a fairly large scope during the night; my most sessions start in darkness and end in darkness. What are optimal ISO/shutter settings for a flat? Also, do I have to take a flat for a given session or can I use one for other day’s session/s?




Hi Mike -

For my purposes, I take sky flats during twilight just after sunset or just before morning sunrise. The biggest downside to this is that I can't rotate my camera during the night or my flats will be invalid (unless I do another set of dawn sky flats). For me, its not a big deal because I plan my sessions accordingly and I typically shoot only one or at most two objects in a night. Sky flats are easy, but they're best done at twilight when sky illumination is relatively even.

You can take daytime sky flats as well, so don't rule that out. If you're shooting twilight flats, or even daytime flats, its best to keep the scope moving while you're shooting them so that any uneven illumination (i.e., bright stars at twilight) are spread and averaged throughout the flat frame.

If you take flats during the night, then you'll need an even source of illumination. There are light boxes that are designed for this (e.g. Spike-a Flat Fielder), or you can use a laptop that has a blank screen illuminated on it, or any other device that can create an evenly illuminated field. Most people stretch a white T-shirt over the OTA in order to add additional diffusion to the illumination source and ensure a more truly flat illumination (this is also good practice when shooting sky flats).

Once you have your diffused, flat illumination source (be it light box, daytime sky, twilight sky, or whatever) - Images Plus' Camera Control module makes it very easy to take your flats. (Bear with me, because I'm doing this from memory - the camera's on the scope so I can't look at the IP screens right now). All you do is set the dial on your camera to Av mode. This causes the camera to calculate the correct shutter speed to properly expose the image. Then, in IP, you go to the connect tab and choose the directory where you want to save your flats. Then, on the Capture tab, click on the Add button. Choose the number of flats you want (I usually do 30), set your ISO, set any startup delay you want, any delay between images (not essential) and then choose the Flat-Av Mode as the image type. This sets IP so that it knows the camera is in Av mode and will calculate the correct exposure for each flat frame.

As far as ISO - I shoot my flats at the same ISO as my bias frames (which are the same ISO as my lights). Mike Unsold can correct me here if I'm wrong, but I believe the calibration for the Flats is that it Corrected Flat = Flat Frame - Bias Frames (to remove the bias from the Flat). I don't know if IP can do some form of scaling if the bias frames are at a different ISO, but I just play it safe and use the same ISO as my Bias for my Flats.

Also, if you're shooting Flats with, say, a narrowband filter in place, you may find that your exposure time for the flats is fairly long (many seconds). At some point, the thermal signal in a long flat frame becomes a fairly significant component. As such, if you're shooting long exposure flat frames, you should shoot Flat Dark frames. These are simply dark frames (scope with cap in place) that are shot at the same ISO as the Flats and the same exposure time as the Flats. I use as a rule of thumb that if my Flats are longer than about 10 seconds, then I shoot Flat Darks. It generally won't hurt to shoot flat darks (even for shorter exposures), but for most exposure times its overkill.

And - the most important aspect of Flats is that you don't want to change your optical train after shooting your flats. If you rotate the camera, pull it off the scope and put it back on, or any other changes, you're revising the positioning of the sensor relative to the optics and your flats are no longer valid. Most folks will tell you that even if you change your focus you've invalidated your flats. For my purposes, I try to take my flats with what should be a good focus. Throughout the course of the night, my focus might change. My choices are to correct focus and (to some extent) invalidate my flats, or to have good flats and have the image out of focus. Obviously, focus is more important than flats, so I adjust my focus. The big key here is that I do reshoot flats if my initial focus wasn't close to where it should have been. I don't reshoot flats if I start with a good focus and have to make minor changes to my focus throughout the night.

Concerning reusing flats - if you haven't removed the camera from the scope from one night to the next, then you can reuse the flat. You have some slight risk that some new dust may have settled on the sensor or moved, but this isn't too common while the camera is connected to a closed optical system. On occasion, I miss my twilight window for taking flats, and I'll reuse the flats from the night before (I usually don't move the scope or camera from night to night). Reasonably good flats are definitely preferable to no flats.

Hope this helps.


Post Extras: Print Post   Remind Me!   Notify Moderator  
mmalik
Postmaster
*****

Reged: 01/13/12

Loc: USA
Re: ImagesPlus+Photoshop Astro Image Processing new [Re: Mike Unsold]
      #5464843 - 10/11/12 04:33 AM

Quote:

One way of using ImagesPlus to make a synthetic flat. ...you should really consider using a real master flat since it measures real vignetting.

Flat calibration uses division and not subtraction to apply the flat frame. Problem with subtraction is that it can destroy data.




Yes, real flats are the way to go, no doubt. Thx

And Scott, thanks for your flats elaboration, greatly appreciated!


Post Extras: Print Post   Remind Me!   Notify Moderator  
mmalik
Postmaster
*****

Reged: 01/13/12

Loc: USA
Re: ImagesPlus+Photoshop Astro Image Processing new [Re: mmalik]
      #5464845 - 10/11/12 04:43 AM

Latest versions of very easy to follow, click-by-click, ImagesPlus instructions that I have been creating for my own learning sake can be found here...; will keep updating as I learn various methods/tools within ImagesPlus. Thanks to Mike U. for his great support.

If folks needed combined/unprocessed FIT files to practice on, they can be found here.... Note: If you use one of these FIT files, start with step # 34 in the document.


Post Extras: Print Post   Remind Me!   Notify Moderator  
mmalik
Postmaster
*****

Reged: 01/13/12

Loc: USA
Re: ImagesPlus+Photoshop Astro Image Processing new [Re: Mike Unsold]
      #5470557 - 10/14/12 09:01 PM

In effort to keep ImagesPlus related items/workflows/techniques in one area, I am posting following from NGC 7023 - Iris Nebula [Re-processed]... thread regarding managing star Color without a mask.

Quote:

Quote:

Or you could use Photoshops "Match Colour" tool without having to do any fancy masking/layers work. The tool is nominally designed to aid you to colour match one image to another but the largley unknown feature is you can use this on a single image to do highly controlled colour saturation work. You can add masks as well if you need to add further control.

I also prefer the OPs version. The additional colour saturation example done in the follow up is nice but unfortunatly has blown the highlights around the central star (on my monitor at least). This could have been controlled with a mask or you could have done a HDR blend back on the original to preserve the core - just a suggestion.




You do not need to use masks but a star mask is easy to make and use with a color adjustment tool in ImagesPlus 5.

The original posted image that was used to add color to stars does have the center of the Iris Nebula a little too bright. I made an adjustment to step 5) shown below that increased red in the Iris nebula. No mask was used I just reduce the feathered luminance range where the red increase is applied so that the center is maintained at the same brightness as the original posted version. Here is the result

http://www.mlunsold.com/temp/IrisReproc.jpg

Steps used with IP 5 :

1) Initial posted image with little star color and bright center star.

http://www.mlunsold.com/temp/Iris0.jpg


2) Blue star color is increase using a star mask. The mask selects only stars. The hue, saturation, and luminance range set on the Selective HSL Color Balance tool changes only blue to magenta stars.

http://www.mlunsold.com/temp/Iris1.jpg


3) Red star color is increase using the same star mask. The mask selects only stars. The hue, saturation, and luminance range set on the Selective HSL Color Balance tool changes only red stars.

http://www.mlunsold.com/temp/Iris3.jpg


4) Increase blue in Iris Nebula. No mask is used. But the luminance range is set to [17733, 65535] with a feahter radius of 10000. Saturation range is set to [0.0, 1.0].

http://www.mlunsold.com/temp/Iris4.jpg


5) Increase red in Iris Nebula. No mask is used but the luminance range is set to [24929, 36751] with a feahter radius of 13622. Saturation range is set to [0.0, 1.0].

This was the step where the center star became a little too bright since the luminance range was [24929, 65535]. Lowering the max L upper bound to 36751 with feather radius 13622 correctly blends the red increase without making the center star brighter.

http://www.mlunsold.com/temp/Iris5.jpg


6) Increase overall saturation. No mask is used.

http://www.mlunsold.com/temp/Iris6.jpg




Post Extras: Print Post   Remind Me!   Notify Moderator  
mmalik
Postmaster
*****

Reged: 01/13/12

Loc: USA
Re: ImagesPlus+Photoshop Astro Image Processing new [Re: Mike Unsold]
      #5470573 - 10/14/12 09:12 PM

In effort to keep ImagesPlus related items/workflows/techniques in one area, I am posting following from NGC 7635 - Bubble Nebula (Revisited)... thread regarding maintaining star color from the start rather than at the end.

Quote:

Mike M.

I used your stacked Bubble Nebula download and process using the steps shown below with IP 5 so that you can see an example of how you workflow can be used to maintain star color from the start. Rather than try to fix star color at the end maintain star color at each step so very little needs to be done at the end of processing.

Here are the steps

1) Initial stacked but otherwise unprocessed image

http://www.mlunsold.com/Temp/BN1.jpg


2) Digital Development with no color emphasis. Normal blend mode, opacity = 100%, no mask.

Note that stars have color and the gradient from left to right with the right edge darker.

http://www.mlunsold.com/Temp/BN2.jpg


3) Fix gradient using the Mulit-Point Flatten tool. Normal blend mode, opacity = 100%, no mask.

http://www.mlunsold.com/Temp/BN3.jpg

Star color is not changed.


4) Smooth background. Star color is not changed. Normal blend mode, opacity = 100%, no mask.

http://www.mlunsold.com/Temp/BN4.jpg


5) Apply smooth background again but with a larger feather radius. Star color is not changed. Normal blend mode, opacity = 100%, no mask.

http://www.mlunsold.com/Temp/BN5.jpg


6) Use Levels to set black point. Normal blend mode, opacity = 100%, no mask.

http://www.mlunsold.com/Temp/BN6.jpg

Stars have obvious color.


7) Neutralize background and set color balance.

http://www.mlunsold.com/Temp/BN7.jpg

Stars have obvious color.


At this point you can do a lot of different operations subject to your own taste.

- Increase saturation. Normal blend mode, opacity = 100%, no mask

http://www.mlunsold.com/Temp/BN8.jpg

- Perhaps smooth the background more using an inverse luminance mask. Blend mode normal and opacity 100%

http://www.mlunsold.com/Temp/BN9.jpg

or same mask smooth background and increase contrast with Soft light blend mode and opacity = 0.540 to blend a little with the previous layer at step 7.

http://www.mlunsold.com/Temp/BN10.jpg

Note that special steps to fix star color are not needed. Star color was shown from the first step and
maintained or enhanced a little at each step.

Based on your taste you can sharpen, boost color, .... at this point




Quote:

Hi Scott,

Quote:

Hi Mike Unsold -

Out of curiosity, I see that in steps 4 and 5 you apply smoothing twice with the second step identical except the larger feathering range. Does this two step process have a different result than just starting with the larger feather? Do you mind clarifying why you chose this method?





Smoothing is applied first in step 4 with luminance range [0, 20766] and feather range 4198 so that only the darker background is smoothed with no noticeable effect on stars. It seemed that a little more background smoothing was needed so I pressed the Set View button on the top horizontal toolbar in ImagesPlus to save the image result from the first smoothing and allow ImagesPlus to apply the same Multiresolution smoothing filter again for twice the effect. The second smoothing used the same luminance range as the first but I increased the feather range to smooth a little brighter detail slightly. You can clearly see the change in the background of the image if you setup step 3, 4, and 5

http://www.mlunsold.com/Temp/BN3.jpg

http://www.mlunsold.com/Temp/BN4.jpg

http://www.mlunsold.com/Temp/BN5.jpg

in different tabs of your browser then click to blink them.

The use of the Set View toolbar button is important. If I had not pressed the Set View button after step 4 then ImagesPlus would think I was changing the parameters of the Multiresolution tool that was applied in step 4 and I would not get the second application of smoothing. If you want to apply the same filter twice in a row then you need to use the Set View button to tell ImagesPlus that this is a new application of the filter and not a change of parameters.


Quote:


Also, I've usually used Standard Smoothing and NR with very pleasing results. I see you used a very large window as a last step for smoothing - would you mind elaborating on some of the rationale you use for deciding how to do smoothing/NR?

Thanks loads.




Steps 8, 9, and 10 are intended as examples of things that you might want to do subject to your taste. Steps 9 and 10 are the same Gaussian blur with large 21 x 21 blur window. I wanted to apply a strong smoothing to the dark background without effecting dim stars. So I made a star mask that included the dim stars and then inverted the star mask to get the inverse star mask shown with the Gaussian blur tool in step 9

http://www.mlunsold.com/Temp/BN9.jpg

The inverse star mask does it job well and protects the dim stars from background smoothing.

Step 10 used the same inverse star mask and 21 x 21 Gaussian blur but also uses Soft light blend mode to increase contrast and opacity = 0.540 to soften the Gaussian filter result and blend it with the previous layer which is shown in step 8.

In general the mask, blend mode, and opacity option extends control of every stretch, smooth, sharpen, deconvolution, and color adjustment tool. Smooth, sharpen, and deconvolution tools all have the luminance range and feather control built in for additional control. The color adjustment tools also have feathered luminance range control but also have feather hue and saturation controls to specif an exact hue, luminance, and saturation range where the tool is applied. A selective color mask really allows you to customize the hue, saturation, and luminance range.

Does this all makes or should I post some examples?




Post Extras: Print Post   Remind Me!   Notify Moderator  
mmalik
Postmaster
*****

Reged: 01/13/12

Loc: USA
Re: ImagesPlus+Photoshop Astro Image Processing new [Re: Mike Unsold]
      #5470586 - 10/14/12 09:23 PM

In effort to keep ImagesPlus related items/workflows/techniques in one area, I am posting following from NGC 772 - Spiral Galaxy In Aries... thread regarding correcting un-even frame (vignetting) with ImagesPlus 5.0.

Quote:

...the stacked image does have the left and right vertical edge darker than the center vertical section. How are you taking flats. I use sky flats which are easy to do with a DSLR and give me the best result.

Here is how I would correct the NGC 772 uneven frame with ImagesPlus 5.0.

NOTE: Steps 1 - 7 all use normal blend mode and opacity = 100%. Step 8 use Soft Light blend mode and opacity = 0.573.

1) Initial stacked image

http://www.mlunsold.com/Temp/NGC772-1.jpg


2) Stretch | Digital Development

http://www.mlunsold.com/Temp/NGC772-2a.jpg

same stretch parameters as above but now with color emphasis enabled

http://www.mlunsold.com/Temp/NGC772-2b.jpg


3) Special Functions | Multi Point Flatten used to correct uneven frame

http://www.mlunsold.com/Temp/NGC772-3a.jpg

Left click on each of the + marks around the edge and increase the level then select the two center + marks near NGC 772 and decrease the level.


The Line Profile button on the top toolbar can be used to check flatness along several different lines

http://www.mlunsold.com/Temp/NGC772-3b.jpg


4) Control point mask is used to control Stretch | Micro curves so that only the edges become a little brighter

http://www.mlunsold.com/Temp/NGC772-4.jpg


5) Smoothed inverse luminance mask is used to smooth the background. Smooth Stretch | Smoothing tool is used

http://www.mlunsold.com/Temp/NGC772-5.jpg


6) Color | Color Balance tool is used to neutralize the background color. Check the Neutralize Background box on Color Balance then left click on the background in the center but away from a star and NGC 772.

http://www.mlunsold.com/Temp/NGC772-6.jpg


7) Smoothed luminance mask is used to increase NGC 772 and stars. Stretch | Micro Curve tool is used.

http://www.mlunsold.com/Temp/NGC772-7.jpg


8) Inverse luminance mask used in step 5 was loaded and used with Stretch | Micro Curve to reduce background brighness with little effect on NGC 772. Soft Light blend mode used to increase contrast and opacity = 0.573 to blend result into previous layer at step 7.

http://www.mlunsold.com/Temp/NGC772-8.jpg


At this point NGC 772 appears as

http://www.mlunsold.com/Temp/NGC772.jpg

and it would be a good time to adjust color or maybe sharpen a little all based on your personnel preference.




Post Extras: Print Post   Remind Me!   Notify Moderator  
Pages: 1 | 2 | 3 | 4 | 5 | (show all)


Extra information
3 registered and 19 anonymous users are browsing this forum.

Moderator:  Dave M, WOBentley, tecmage 

Print Thread

Forum Permissions
      You cannot start new topics
      You cannot reply to topics
      HTML is disabled
      UBBCode is enabled


Thread views: 6473

Jump to

CN Forums Home


Cloudy Nights LLC
Cloudy Nights Sponsor: Astronomics