Page 2 of 4 FirstFirst 1234 LastLast
Results 11 to 20 of 40
Like Tree5Likes

Thread: Step-by-step processing example (NGC 891)

  1. #11
    Fuzz's Avatar
    Fuzz is offline Bright Giants
    Points: 11,818, Level: 75
    Level completed: 23%, Points required for next Level: 232
    Overall activity: 0%
    Achievements:
    10 Days registered200+ Posts Achievement!First 1000 Experience Points365 Days+ Registered Achievement!750 Days+ Registered Achievement!
    Join Date
    Apr 2009
    Location
    Silverthorne, CO
    Posts
    469
    Points
    11,818
    Level
    75
    Thanks
    68
    Thanked 450x 165 Posts

    Default



    Quote Originally Posted by mplanet62 View Post
    LR limits the capabilities, but makes it simpler and more "visual" at a lower level of astro photography expertise - which I honestly do belong to. Also, it's simpler to see the frame's development potential before proceeding to PS - which can be started from Lightroom - must be quite useful for DSLR phography.
    Very good point. Photoshop can be very overwhelming -- it just opens up and doesn't give any help about what to do. With LR, one can easily follow each of the panels in the Develop tab from top to bottom in sequence and get a good result -- it is limited, but structured and very helpful. One example are Curves. In PS, first you have to know what Curves are to dig them up from the many menus. Then it gives you a dialog box with a complicated chart and no hint about how to use it. In LR on the other hand, it's the fourth panel in the Develop tab, so its right there. And if the curve graph doesn't make sense, the four sliders at the bottom quickly give an intuitive idea of what it does (plus you can see the behavior of the curve as a result).

    And the same Curves function is a good example also of how PS excels over LR. In LR, if I move a point on the curve, the rest of the curve moves to make a smooth gradient. In PS, I can set numerous points to fix any point on the curve, and only the portions I want to move will move. I can create a way more complicated and custom curve adjustment. For instance, LR won't really allow the Curve function that I applied in my workflow above -- the curve has a "hard" angle and isn't smooth.

    I'll admit, I've been using Photoshop for several years now (for regular/non-astro photography), so I've become conversant with it, have developed a certain workflow, and know where to go to do something. LR makes it so simple though, I've abandoned PS for my regular processing (except for really problematic ones) - I do almost everything in LR. The astro-images are the main exception as I mentioned.

    At the moment GIMP is my editor of choice - it's free and works with both Windows and Ubuntu on my double-system comp.
    I recall reading somewhere on this forum that GIMP cannot handle 16-bit files? Is that right?
    Visual: Celestron NexStar 8SE
    Imaging: Astro-Tech AT6RC, Sirius EQ-G mount, Orion autoguider, Canon 40D
    Location: Colorado Rockies

  2. The Following User Says Thank You to Fuzz For This Useful Post:

    mplanet62 (09-18-2010)

  3. #12
    gremzor's Avatar
    gremzor is offline Main Sequence
    Points: 4,706, Level: 46
    Level completed: 78%, Points required for next Level: 44
    Overall activity: 0%
    Achievements:
    5+ Referrals Achievement!First 1000 Experience Points365 Days+ Registered Achievement!750 Days+ Registered Achievement!2 Posts Achievement
    Join Date
    Aug 2010
    Location
    SE Michigan
    Posts
    69
    Points
    4,706
    Level
    46
    Thanks
    33
    Thanked 30x 15 Posts

    Default

    Thank you for the workflow, definitely will give total newbs like me some direction in processing our images .

    I recall reading somewhere on this forum that GIMP cannot handle 16-bit files? Is that right?
    This is correct. I am "stuck" with gimp for general editing for now. I want to eventually try to learn Iris but it is very specialized and adjusting curves and levels seems like a major task in that program. I believe the next major version of gimp will add support for 16 bit files though. I am guessing a large amount of subtlety is lost when converting to 8 bit, although I have no direct comparison.

    Direct quote from gimp's faq:

    When can we see 16-bit per channel support (or better)?

    For some industries, especially photography, 24-bit colour depths (8 bits per channel) are a real barrier to entry. Once again, it's GEGL to the rescue. Work on integrating GEGL into GIMP began after 2.4 was released, and will span across several stable releases. This work will be completed in GIMP 3.0, which will have full support for high bit depths. If you need such support now and can't wait, cinepaint and krita support 16 bits per channel now.

    It should be noted that for publishing to the web, the current GIMP release is good enough.
    Also, I found a plugin for gimp that adds some extra processing features to gimp:

    http://gmic.sourceforge.net/
    Last edited by gremzor; 09-18-2010 at 02:36 PM. Reason: Added more gimp information.
    Celestron NexStar 127 SLT
    6.4mm,9mm,12mm,25mm,32mm EPs
    Canon T1i DSLR

  4. The Following 3 Users Say Thank You to gremzor For This Useful Post:

    Fuzz (09-18-2010),mplanet62 (09-18-2010),mrschmitz (01-04-2012)

  5. #13
    mplanet62's Avatar
    mplanet62 is offline SUPER GIANT
    Points: 7,498, Level: 60
    Level completed: 74%, Points required for next Level: 52
    Overall activity: 0%
    Achievements:
    Ghost Achievement! Averaging 5+ posts a day!200+ Posts Achievement!Visitor Messenger Achievement!Got three FriendsFirst 1000 Experience Points
    Join Date
    Aug 2010
    Location
    @
    Posts
    1,223
    Points
    7,498
    Level
    60
    Thanks
    638
    Thanked 493x 367 Posts

    Default

    On 16-bit issue in GIMP GIMP - Documentation Also here 8.5.*Use GEGL
    Some interesting info on color depth here Canon Professional Network - 8 or 16 bit
    Except Most ink-jet printers work in 8-bit mode, and 8-bit files give excellent prints. 12-bit mode is mostly useful when you intend to do a lot of processing work on the file after it has been shot.
    JPG files are 8 bit per channel anyway, so with JPG as final result and a possibility of the internal processing being done in 32bit floating point linear light RGBA, I am quite happy to use GIMP for free. If I will ever be able to publish my photos professionally, sure thing I will need PS.
    Clear skies,
    Michael

  6. The Following User Says Thank You to mplanet62 For This Useful Post:

    Fuzz (09-18-2010)

  7. #14
    Fuzz's Avatar
    Fuzz is offline Bright Giants
    Points: 11,818, Level: 75
    Level completed: 23%, Points required for next Level: 232
    Overall activity: 0%
    Achievements:
    10 Days registered200+ Posts Achievement!First 1000 Experience Points365 Days+ Registered Achievement!750 Days+ Registered Achievement!
    Join Date
    Apr 2009
    Location
    Silverthorne, CO
    Posts
    469
    Points
    11,818
    Level
    75
    Thanks
    68
    Thanked 450x 165 Posts

    Default

    Quote Originally Posted by gremzor View Post
    This is correct. I am "stuck" with gimp for general editing for now. I want to eventually try to learn Iris but it is very specialized and adjusting curves and levels seems like a major task in that program. I believe the next major version of gimp will add support for 16 bit files though. I am guessing a large amount of subtlety is lost when converting to 8 bit, although I have no direct comparison.
    There's a lot of technical reading online about bit depth. But at a very simple level, bit depth has to do with levels of brightness.

    Each image is made up of 3 channels of color, Red, Green, Blue. In an 8-bit image, you get 256 levels of brightness in each color. In a 16-bit image, you get 65,536 levels of brightness. This means that a 16-bit image has a lot more subtle shades available -- so you can really distinguish the very fine shades in an astro-image. An 8-bit image doesn't have as many shades available, so it has to compress some of the subtler shades into a single shade thus losing the faint distinction.

    Here's a simplified example:


    With only 4 levels of brightness, there is not much distinction. But with 8 levels of brightness, you can see a lot more variation. If we take an image with many subtle shades, you can see the difference in the two levels of shades. The image below is exactly the same, except in 4 levels, the 8 levels of brightness are compressed into the available 4 levels:


    For regular daytime photography, 256 shades of brightness may be adequate. But for astro-images, which have very faint details, the greater levels of brightness obviously help distinguish the subtle details like nebulosity, spiral arms, etc.
    Visual: Celestron NexStar 8SE
    Imaging: Astro-Tech AT6RC, Sirius EQ-G mount, Orion autoguider, Canon 40D
    Location: Colorado Rockies

  8. The Following 5 Users Say Thank You to Fuzz For This Useful Post:

    dynamo (04-09-2012),gremzor (09-18-2010),Mark W (06-17-2011),mplanet62 (09-18-2010),ozzmozizz (08-16-2011)

  9. #15
    Fuzz's Avatar
    Fuzz is offline Bright Giants
    Points: 11,818, Level: 75
    Level completed: 23%, Points required for next Level: 232
    Overall activity: 0%
    Achievements:
    10 Days registered200+ Posts Achievement!First 1000 Experience Points365 Days+ Registered Achievement!750 Days+ Registered Achievement!
    Join Date
    Apr 2009
    Location
    Silverthorne, CO
    Posts
    469
    Points
    11,818
    Level
    75
    Thanks
    68
    Thanked 450x 165 Posts

    Default

    Quote Originally Posted by mplanet62 View Post
    JPG files are 8 bit per channel anyway, so with JPG as final result and a possibility of the internal processing being done in 32bit floating point linear light RGBA, I am quite happy to use GIMP for free.
    While final JPEGs are 8-bit, it does make a difference in whether you use 16 bit or 8 bit to EDIT the raw image.

    With greater shades of brightness, you can "spread" out the faint details into distinct shades. I don't have the time right now to go into a lot of detail, but when you stretch your astro-image, you can spread out similar values of brightness into separate brightness levels to make them distinct. Then, when you convert into 8 bits, all those now-distinct shades are mapped to the 256 brightness containers - yes, you'll lose some detail, but you've still managed to separate out the subtle shades before compressing to lesser number of shades.

    But when you open a 16-bit image in 8 bits, all those similar shades are compressed into a single shade in 8 bits -- so before you even begin to do anything, you've lost all that subtle distinction -- no matter how much you stretch, you only have that single level.
    Visual: Celestron NexStar 8SE
    Imaging: Astro-Tech AT6RC, Sirius EQ-G mount, Orion autoguider, Canon 40D
    Location: Colorado Rockies

  10. The Following User Says Thank You to Fuzz For This Useful Post:

    mplanet62 (09-18-2010)

  11. #16
    Fuzz's Avatar
    Fuzz is offline Bright Giants
    Points: 11,818, Level: 75
    Level completed: 23%, Points required for next Level: 232
    Overall activity: 0%
    Achievements:
    10 Days registered200+ Posts Achievement!First 1000 Experience Points365 Days+ Registered Achievement!750 Days+ Registered Achievement!
    Join Date
    Apr 2009
    Location
    Silverthorne, CO
    Posts
    469
    Points
    11,818
    Level
    75
    Thanks
    68
    Thanked 450x 165 Posts

    Default

    Example...

    Here's an original image captured in 3 bits (8 levels of brightness). You can see there are several subtle details captured in 8 distinct shades. They're all compressed towards the black end, but they're there.


    Now, I open up that raw image and stretch it to make the distinctions more prominent. Still in 8 distinct levels.


    Now I convert into the lower 2-bit (4 levels) format, so some of the levels get compressed. Yes, there's a loss in detail, but at least there's some gradient.


    But what happens if we open that 3-bit image in 2 bits right up front? All those dark shades get compressed into a single shade of brightness, zero (since that's the closest match). So my starting image contains only 2 distinct shades. No matter how much I stretch, all those subtle shades are lost.
    Visual: Celestron NexStar 8SE
    Imaging: Astro-Tech AT6RC, Sirius EQ-G mount, Orion autoguider, Canon 40D
    Location: Colorado Rockies

  12. The Following 4 Users Say Thank You to Fuzz For This Useful Post:

    dynamo (04-09-2012),ozzmozizz (08-16-2011),P-Dub (03-30-2012),stefski (02-27-2012)

  13. #17
    mplanet62's Avatar
    mplanet62 is offline SUPER GIANT
    Points: 7,498, Level: 60
    Level completed: 74%, Points required for next Level: 52
    Overall activity: 0%
    Achievements:
    Ghost Achievement! Averaging 5+ posts a day!200+ Posts Achievement!Visitor Messenger Achievement!Got three FriendsFirst 1000 Experience Points
    Join Date
    Aug 2010
    Location
    @
    Posts
    1,223
    Points
    7,498
    Level
    60
    Thanks
    638
    Thanked 493x 367 Posts

    Default

    Quote Originally Posted by Fuzz View Post
    While final JPEGs are 8-bit, it does make a difference in whether you use 16 bit or 8 bit to EDIT the raw image.

    With greater shades of brightness, you can "spread" out the faint details into distinct shades. I don't have the time right now to go into a lot of detail, but when you stretch your astro-image, you can spread out similar values of brightness into separate brightness levels to make them distinct. Then, when you convert into 8 bits, all those now-distinct shades are mapped to the 256 brightness containers - yes, you'll lose some detail, but you've still managed to separate out the subtle shades before compressing to lesser number of shades.

    But when you open a 16-bit image in 8 bits, all those similar shades are compressed into a single shade in 8 bits -- so before you even begin to do anything, you've lost all that subtle distinction -- no matter how much you stretch, you only have that single level.
    Very good point. But GIMP itself does not handle RAW http://www.rockynook.com/samples/20/Chapter_2.pdf
    Except:UFRaw can be used in three different ways. If used as a GIMP plug-in,when you open a RAW file in the GIMP, the UFRaw preview window will
    automatically open. You can set corrective options for color and brightness
    values in the preview window.
    You can simply click OK on an open image to load it into the GIMP.
    Then you can use the GIMP’s tools for corrections—a legitimate practice.
    Since UFRaw supports a color depth of 16 bits per color channel, this
    method allows you to make detailed adjustments while using the GIMP’s
    familiar tools. (Without UFRaw, you’d be limited to 8 bits.)

    An example on how it works here:Recovering highlights with gimp and ufraw
    I absolutely agree - GIMP is not Photoshop, but a lot can be done in it.
    Cheers,
    Michael

  14. The Following User Says Thank You to mplanet62 For This Useful Post:

    Fuzz (09-18-2010)

  15. #18
    Fuzz's Avatar
    Fuzz is offline Bright Giants
    Points: 11,818, Level: 75
    Level completed: 23%, Points required for next Level: 232
    Overall activity: 0%
    Achievements:
    10 Days registered200+ Posts Achievement!First 1000 Experience Points365 Days+ Registered Achievement!750 Days+ Registered Achievement!
    Join Date
    Apr 2009
    Location
    Silverthorne, CO
    Posts
    469
    Points
    11,818
    Level
    75
    Thanks
    68
    Thanked 450x 165 Posts

    Default

    Quote Originally Posted by mplanet62 View Post
    Very good point. But GIMP itself does not handle RAW
    Since UFRaw supports a color depth of 16 bits per color channel, this
    method allows you to make detailed adjustments while using the GIMP’s
    familiar tools. (Without UFRaw, you’d be limited to 8 bits.)
    So sounds like there is a work-around using UFRaw to really edit in 16 bits? My only point is that for astro-processing, working in 16 bits will retain more detail than editing in 8 bits. The exact software or method used is not important.
    Visual: Celestron NexStar 8SE
    Imaging: Astro-Tech AT6RC, Sirius EQ-G mount, Orion autoguider, Canon 40D
    Location: Colorado Rockies

  16. The Following User Says Thank You to Fuzz For This Useful Post:

    mplanet62 (09-18-2010)

  17. #19
    gremzor's Avatar
    gremzor is offline Main Sequence
    Points: 4,706, Level: 46
    Level completed: 78%, Points required for next Level: 44
    Overall activity: 0%
    Achievements:
    5+ Referrals Achievement!First 1000 Experience Points365 Days+ Registered Achievement!750 Days+ Registered Achievement!2 Posts Achievement
    Join Date
    Aug 2010
    Location
    SE Michigan
    Posts
    69
    Points
    4,706
    Level
    46
    Thanks
    33
    Thanked 30x 15 Posts

    Default

    Here's an original image captured in 3 bits (8 levels of brightness). You can see there are several subtle details captured in 8 distinct shades. They're all compressed towards the black end, but they're there....
    Yeah, that makes perfect sense. My uncertainty was just how the different channels work together to create the image. In essence it's actually 24-bit depth when you add all the colors together. 8 bit for each channel. Logically I know there will be diminishing returns the higher the bit depth goes. Now thinking about it even more, my current target M31 appears (in my image at least) to be almost completely black and white. So I am guessing the total dynamic bandwidth of color gradient doesn't lie much over 8-bit to begin with. So that would essentially mean a good deal of data is lost in the conversion to 8 bit from 16 bit. Sort of thinking out loud here...lol. At any rate I think I am going to fiddle around with installing ubuntu on a virtual machine if possible and trying cinepaint for something different. Thanks again for the elaboration.
    Celestron NexStar 127 SLT
    6.4mm,9mm,12mm,25mm,32mm EPs
    Canon T1i DSLR

  18. #20
    Fuzz's Avatar
    Fuzz is offline Bright Giants
    Points: 11,818, Level: 75
    Level completed: 23%, Points required for next Level: 232
    Overall activity: 0%
    Achievements:
    10 Days registered200+ Posts Achievement!First 1000 Experience Points365 Days+ Registered Achievement!750 Days+ Registered Achievement!
    Join Date
    Apr 2009
    Location
    Silverthorne, CO
    Posts
    469
    Points
    11,818
    Level
    75
    Thanks
    68
    Thanked 450x 165 Posts

    Default

    Quote Originally Posted by gremzor View Post
    In essence it's actually 24-bit depth when you add all the colors together. 8 bit for each channel.
    Correct. 8 bits per channel = 256 shades per color (black, dark red, red, light red, white...black, dark green, green, light green, white...and so on) = total 8 x 3 = 24 bits of color.

    I also think that to take advantage of 16-bit editing, the processing must be subtle too. If I make big adjustments in 16 bits which compress the shades drastically (to 8 bits), then there's not much advantage in using 16 bits - I could just as well use 8 bits, since my processing isn't taking advantage of the smoother gradients.

    Also, some images are probably more well suited than others, depending on how much subtle detail they have. Something like M13 or M104 maybe would not yield much more detail in 16 bits. Something like M42 or M51 on the other hand, would have a lot more hidden detail that we can tease out of 16 bits.
    Visual: Celestron NexStar 8SE
    Imaging: Astro-Tech AT6RC, Sirius EQ-G mount, Orion autoguider, Canon 40D
    Location: Colorado Rockies

  19. The Following 3 Users Say Thank You to Fuzz For This Useful Post:

    gremzor (09-18-2010),mplanet62 (09-18-2010),stefski (02-27-2012)

 

 
Page 2 of 4 FirstFirst 1234 LastLast

Similar Threads

  1. Intro to image stacking: Step-by-step
    By A.S.I.G.N. Observatory in forum Astrophotography Image Processing Forum
    Replies: 23
    Last Post: 05-10-2014, 12:06 AM
  2. Next step
    By harbinjer in forum Binoculars Forum
    Replies: 12
    Last Post: 12-31-2008, 07:54 PM
  3. Taking the Step Up
    By John Nichols in forum Amateur Astronomy Forum
    Replies: 4
    Last Post: 11-08-2007, 01:37 AM
  4. Replies: 4
    Last Post: 03-27-2007, 02:34 AM

Tags for this Thread

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •  
Powered by vBulletin® Version 4.2.0
Powered by vBulletin®
All times are GMT. The time now is 08:45 AM.