Replicating alu skins in graphic programmes

Discussion in 'Gallery & Designs' started by Leif Oh, Jul 10, 2005.

  1. Jim Krauzlis

    Jim Krauzlis Active Member

    Joined:
    Sep 26, 2005
    Messages:
    2,327
    Likes Received:
    0
    Gil, that's just incredible! :D

    You show a true artist's flair when it come to the blending of the various attributes to give the illusion of aluminum, deftly done and a vast improvement over the earlier scheme.

    The software and graphic applications are well beyond my comprehension, but I do appreciate the end result, and the process you've detailed is a wonderful lesson...thanks!

    Cheers!

    Jim
  2. Jim Krauzlis

    Jim Krauzlis Active Member

    Joined:
    Sep 26, 2005
    Messages:
    2,327
    Likes Received:
    0
    Gil, that's just incredible! :D

    You show a true artist's flair when it come to the blending of the various attributes to give the illusion of aluminum, deftly done and a vast improvement over the earlier scheme.

    The software and graphic applications are well beyond my comprehension, but I do appreciate the end result, and the process you've detailed is a wonderful lesson...thanks!

    Cheers!

    Jim
  3. Leif Oh

    Leif Oh Member

    Joined:
    Jan 27, 2004
    Messages:
    725
    Likes Received:
    0
    The real value of Gil's work lies in the fact that you can now (with some additional work!) scan & scale commercial models with a ready-made alu finish (such as the one we've used as an example here, or the Halinski P38 Lightning). Otherwise you would be forced to build the kit as bought (no backup, no scaling, no modifications at all).

    Second great advantage follows from this: Prospective designers may accomplish a credible alu finish without resorting to commercial and expensive printing. Gil's method would be the only alternative for electronic downloading of alu-clad models.

    The additional great value lies in the fact that the simulated alu finish may well turn out to be a superior option compared also to the commercially supplied alu-finish kits. These do not have any gradients or reflections of sky, grass and earth - which will be there in any real aircraft.

    Nor can a uniform commercial alu finish reflect the difference between individual panels - which is quite common; see any photo of a real aircraft.

    So this is indeed a very great step forward. Gil, if you feel at all up to it, and would be willing to share your technique in more detail (I know you have supplied some values in previous posts, but I for one fell the need for some more hands-on tutoring), this would be of immense value for many, I believe.

    A final refinement would be to elaborate a technique for creating slight differences between individual panels, while not disturbing the overall beautiful shine of your experiment above.

    The final and conclusive test then could be undertaken - building a model, or at least a section of it, straight out of the box, and compare it to an identical section built from parts recoloured with Gil's method.

    Respectfully, and gratefully,

    Leif
  4. Leif Oh

    Leif Oh Member

    Joined:
    Jan 27, 2004
    Messages:
    725
    Likes Received:
    0
    The real value of Gil's work lies in the fact that you can now (with some additional work!) scan & scale commercial models with a ready-made alu finish (such as the one we've used as an example here, or the Halinski P38 Lightning). Otherwise you would be forced to build the kit as bought (no backup, no scaling, no modifications at all).

    Second great advantage follows from this: Prospective designers may accomplish a credible alu finish without resorting to commercial and expensive printing. Gil's method would be the only alternative for electronic downloading of alu-clad models.

    The additional great value lies in the fact that the simulated alu finish may well turn out to be a superior option compared also to the commercially supplied alu-finish kits. These do not have any gradients or reflections of sky, grass and earth - which will be there in any real aircraft.

    Nor can a uniform commercial alu finish reflect the difference between individual panels - which is quite common; see any photo of a real aircraft.

    So this is indeed a very great step forward. Gil, if you feel at all up to it, and would be willing to share your technique in more detail (I know you have supplied some values in previous posts, but I for one fell the need for some more hands-on tutoring), this would be of immense value for many, I believe.

    A final refinement would be to elaborate a technique for creating slight differences between individual panels, while not disturbing the overall beautiful shine of your experiment above.

    The final and conclusive test then could be undertaken - building a model, or at least a section of it, straight out of the box, and compare it to an identical section built from parts recoloured with Gil's method.

    Respectfully, and gratefully,

    Leif
  5. Jim Krauzlis

    Jim Krauzlis Active Member

    Joined:
    Sep 26, 2005
    Messages:
    2,327
    Likes Received:
    0
    Well said, as always, Leif! :D

    I probably would be over my head with a tutorial such as you suggest, but it never hurts to try. :wink: I imagine Gil is up to his gills with projects :lol: , but if he should see fit to putting something together I, for one, would be immensely grateful.

    Cheers!

    Jim
  6. Jim Krauzlis

    Jim Krauzlis Active Member

    Joined:
    Sep 26, 2005
    Messages:
    2,327
    Likes Received:
    0
    Well said, as always, Leif! :D

    I probably would be over my head with a tutorial such as you suggest, but it never hurts to try. :wink: I imagine Gil is up to his gills with projects :lol: , but if he should see fit to putting something together I, for one, would be immensely grateful.

    Cheers!

    Jim
  7. Ashrunner

    Ashrunner Member

    Joined:
    Sep 26, 2005
    Messages:
    805
    Likes Received:
    0
    After a couple of weeks vacation to the east of the west where I live, I come back and find incredible work continuing on this thread. Very well done!

    And Gil, I echo Leif's words in that a detailed tutorial posted here for all to grab would be an outstanding benefit to all modelers. This is amazing work...absolutely amazing!
  8. Ashrunner

    Ashrunner Member

    Joined:
    Sep 26, 2005
    Messages:
    805
    Likes Received:
    0
    After a couple of weeks vacation to the east of the west where I live, I come back and find incredible work continuing on this thread. Very well done!

    And Gil, I echo Leif's words in that a detailed tutorial posted here for all to grab would be an outstanding benefit to all modelers. This is amazing work...absolutely amazing!
  9. Gil

    Gil Active Member

    Joined:
    Jan 25, 2004
    Messages:
    1,981
    Likes Received:
    0
    Hi All,

    Thanks for all the nice comments! Thanks Leif for fully explaining the motivation and consequence of this work. I have several designs that have been on hold pending the outcome of the pseudo-aluminum experiments. Lack of a supply of printable aluminum foil paper and Leif's need to find a solution for the Yak was the spark that instigated the study. It will take a few days to put together a credible tutorial....,

    Gil
  10. Gil

    Gil Active Member

    Joined:
    Jan 25, 2004
    Messages:
    1,981
    Likes Received:
    0
    Hi All,

    Thanks for all the nice comments! Thanks Leif for fully explaining the motivation and consequence of this work. I have several designs that have been on hold pending the outcome of the pseudo-aluminum experiments. Lack of a supply of printable aluminum foil paper and Leif's need to find a solution for the Yak was the spark that instigated the study. It will take a few days to put together a credible tutorial....,

    Gil
  11. Gil

    Gil Active Member

    Joined:
    Jan 25, 2004
    Messages:
    1,981
    Likes Received:
    0
    Update:

    Found an irresistable way to make rivets to satisfy any stray nitpicker..., Did only a panel plus a little to see what it looks like. Sorry, some of the rivet lines are slightly off...,

    Now for differential panel color and the rest is a tutorial on how it's done...,

    Hope I can remember...,

    Gil
  12. Gil

    Gil Active Member

    Joined:
    Jan 25, 2004
    Messages:
    1,981
    Likes Received:
    0
    Update:

    Found an irresistable way to make rivets to satisfy any stray nitpicker..., Did only a panel plus a little to see what it looks like. Sorry, some of the rivet lines are slightly off...,

    Now for differential panel color and the rest is a tutorial on how it's done...,

    Hope I can remember...,

    Gil
  13. Leif Oh

    Leif Oh Member

    Joined:
    Jan 27, 2004
    Messages:
    725
    Likes Received:
    0
    Gil, great rivets (beside the main effort). I do remember seeing a manual on how to get a setting for the pen an line tool to accomplish this very easily. I suspect you've found that on your own. Unfortunately I lost track of that manual (one of the links Tim supplied in another thread is down now; I know I should've made a note of the technique at the time).

    Luckily you've reinvented the technique. As an aside, and later on, it would be wonderful to get hold of it!

    You should know that you have inspired a renewed interest in the Jak-23 Flora on my part. I've already put in quite a lot of work on other modifications, and now that I know a superb alu finish is within reach, I feel better about taking up that work again.

    Also, a test model surely would be in place (and I suspect there may be a good few by others waiting in the wings, not least the two big ones you've started development on yourself)!

    Leif
  14. Leif Oh

    Leif Oh Member

    Joined:
    Jan 27, 2004
    Messages:
    725
    Likes Received:
    0
    Gil, great rivets (beside the main effort). I do remember seeing a manual on how to get a setting for the pen an line tool to accomplish this very easily. I suspect you've found that on your own. Unfortunately I lost track of that manual (one of the links Tim supplied in another thread is down now; I know I should've made a note of the technique at the time).

    Luckily you've reinvented the technique. As an aside, and later on, it would be wonderful to get hold of it!

    You should know that you have inspired a renewed interest in the Jak-23 Flora on my part. I've already put in quite a lot of work on other modifications, and now that I know a superb alu finish is within reach, I feel better about taking up that work again.

    Also, a test model surely would be in place (and I suspect there may be a good few by others waiting in the wings, not least the two big ones you've started development on yourself)!

    Leif
  15. Leif Oh

    Leif Oh Member

    Joined:
    Jan 27, 2004
    Messages:
    725
    Likes Received:
    0
    Gil, while you're working on the manual (many thanks in advance), here's a thought for a gimmick that would be of great immediate practical value for many, I believe:

    How about a big square - 2000 x 2000 pixels or something like that - treated according to your recipy in the parts bin? I think it should be in the form of a two-layer psd file with transparency preserved.

    I really want to learn how to do what you've done, not least to be able to "fiddle" with it, but I'm also looking for quick-and-dirty solutions for practical applications.

    A big square like that I could stretch or compress according to part sizes, and easily trim to fit in a layer beneath panel lines, markings, etc.

    A thought would be to make the opacity a degree less (double present percentage?), since I could easily reduce, but not increase, opacity on my own.

    I don't know if there's a smarter way to work this out (patterns etc.), but it is what I suddenly longed for.

    In the future, a library of such slightly different simulated alu finishes could be built up, some suitable for horizontal surfaces, reflecting only sky (top) or runway (bottom), others for vertical surfaces, reflecting sky, landscape and runway. Others still for rounded fuselages (lighter topside and darker bottom sides), like your present version.

    But don't let this distract you from your present work. Just thinking out loud.

    Leif
  16. Leif Oh

    Leif Oh Member

    Joined:
    Jan 27, 2004
    Messages:
    725
    Likes Received:
    0
    Gil, while you're working on the manual (many thanks in advance), here's a thought for a gimmick that would be of great immediate practical value for many, I believe:

    How about a big square - 2000 x 2000 pixels or something like that - treated according to your recipy in the parts bin? I think it should be in the form of a two-layer psd file with transparency preserved.

    I really want to learn how to do what you've done, not least to be able to "fiddle" with it, but I'm also looking for quick-and-dirty solutions for practical applications.

    A big square like that I could stretch or compress according to part sizes, and easily trim to fit in a layer beneath panel lines, markings, etc.

    A thought would be to make the opacity a degree less (double present percentage?), since I could easily reduce, but not increase, opacity on my own.

    I don't know if there's a smarter way to work this out (patterns etc.), but it is what I suddenly longed for.

    In the future, a library of such slightly different simulated alu finishes could be built up, some suitable for horizontal surfaces, reflecting only sky (top) or runway (bottom), others for vertical surfaces, reflecting sky, landscape and runway. Others still for rounded fuselages (lighter topside and darker bottom sides), like your present version.

    But don't let this distract you from your present work. Just thinking out loud.

    Leif
  17. Gil

    Gil Active Member

    Joined:
    Jan 25, 2004
    Messages:
    1,981
    Likes Received:
    0
    Hi Leif,

    Some of the earlier experiments were of the very nature you describe. I got a little overwrought and added engraved panels and rivet detail to an aluminum surface that should have been left blank. I came to my senses soon enough realizing that this was a false path. The better results in the long run is to use what's been learned in computer graphics and apply it to the problem in a sensible straight forward manner.

    One of the things you learn in rendering images in computer graphics is how frightenly slow and dullard the very fastest super computer is when it comes to getting a decently real looking scene to render correctly. Humans are still really great interpreters of what we see and it's that fact that makes hand rendering aluminlum a fascinating challenge.

    One has to take into account that every surface has a color. Yes, it may be nearly undetectable as in a mirrored surface but the very fact that it's made up of atomic matter gives it away in a telltale reflection of color. The color reflected depends on many variables but can be broken down into several categories which are commonly used in creating renders. First is the surface color of the object known as diffuse color. These are the colors that the object doesn't absorb. The next is whether the object is giving off light and if it is it's called ambient light. These are pretty tame compared to the one that gives the eye much of the critical information cues about the shape and texture of an object. This one is called specular reflection and depends upon the intensity of the light source, how shiney the object is and how rough the surface is. These common attributes are usually found on what's called the "shader panel" in any 3D rendering software. But we don't have the luxury of an Alias or Renderman for our purposes so we have to rely on our brains to make it work in a 2D paint package. Hence this string.

    Actually the plan I've derived is near to what you desire. The idea finally occurred that the aircraft is being modeled in an environment and it is that environment that's being reflected off the shiney aluminum panels. This means that a Mig in Siberia is going to have a different reflectance envelope than a B-29 on Saipan. Questions regarding whether it will it be on the ground or in the air come to mind. Will it be in towering cumulus or high cirrus? Sunny and bright or grey with snow flurries? The best way to do this is to prepare a gradient for each operation theatre and save that to a gradient library. The base aluminum layer retains it's own gradient which needn't be changed much. Differential panel shades are attained by simply adding a layer of grey or near greay with transparency control. The sides up to the horizon line will reflect the ground surround. Above this the sky is reflected and will require the choice of weather for the particular setting intended for the model to be displayed in. Complex? You bet! Beyond the capabilities of most modelers? Good question, it does take an average proficiency in Photoshop or similar 2D paint package but I think with a helping tutorial and enough initiative that the average modeler can become at ease with doing this kind of art..,

    Best regards, Gil
  18. Gil

    Gil Active Member

    Joined:
    Jan 25, 2004
    Messages:
    1,981
    Likes Received:
    0
    Hi Leif,

    Some of the earlier experiments were of the very nature you describe. I got a little overwrought and added engraved panels and rivet detail to an aluminum surface that should have been left blank. I came to my senses soon enough realizing that this was a false path. The better results in the long run is to use what's been learned in computer graphics and apply it to the problem in a sensible straight forward manner.

    One of the things you learn in rendering images in computer graphics is how frightenly slow and dullard the very fastest super computer is when it comes to getting a decently real looking scene to render correctly. Humans are still really great interpreters of what we see and it's that fact that makes hand rendering aluminlum a fascinating challenge.

    One has to take into account that every surface has a color. Yes, it may be nearly undetectable as in a mirrored surface but the very fact that it's made up of atomic matter gives it away in a telltale reflection of color. The color reflected depends on many variables but can be broken down into several categories which are commonly used in creating renders. First is the surface color of the object known as diffuse color. These are the colors that the object doesn't absorb. The next is whether the object is giving off light and if it is it's called ambient light. These are pretty tame compared to the one that gives the eye much of the critical information cues about the shape and texture of an object. This one is called specular reflection and depends upon the intensity of the light source, how shiney the object is and how rough the surface is. These common attributes are usually found on what's called the "shader panel" in any 3D rendering software. But we don't have the luxury of an Alias or Renderman for our purposes so we have to rely on our brains to make it work in a 2D paint package. Hence this string.

    Actually the plan I've derived is near to what you desire. The idea finally occurred that the aircraft is being modeled in an environment and it is that environment that's being reflected off the shiney aluminum panels. This means that a Mig in Siberia is going to have a different reflectance envelope than a B-29 on Saipan. Questions regarding whether it will it be on the ground or in the air come to mind. Will it be in towering cumulus or high cirrus? Sunny and bright or grey with snow flurries? The best way to do this is to prepare a gradient for each operation theatre and save that to a gradient library. The base aluminum layer retains it's own gradient which needn't be changed much. Differential panel shades are attained by simply adding a layer of grey or near greay with transparency control. The sides up to the horizon line will reflect the ground surround. Above this the sky is reflected and will require the choice of weather for the particular setting intended for the model to be displayed in. Complex? You bet! Beyond the capabilities of most modelers? Good question, it does take an average proficiency in Photoshop or similar 2D paint package but I think with a helping tutorial and enough initiative that the average modeler can become at ease with doing this kind of art..,

    Best regards, Gil
  19. Leif Oh

    Leif Oh Member

    Joined:
    Jan 27, 2004
    Messages:
    725
    Likes Received:
    0
    With you so far. Even to the point of individualizing panels. The same idea occurred to me, that it is relatively simple to accomplish by adding a layer of individualized grey or white panes and regulate the opacity. Not a problem.

    The basic alu shine, plus a library of "weathering" (for once, the appropriate use of the term!) gradients, is the task. How about combining the the two tasks into one library of combined alu shine + weathering patterns, along two categories:

    1) position on aircraft:
    - horizontal top
    - horizontal bottom
    - vertical
    - fuse & other rounded surfaces

    2) position of aircraft:
    - runway (perhaps in a forest landscape)
    - airborne

    That would reduce the basic library to eight different patterns, four in each of the two main groups, airborne or on runway. The last group in particular could be added to almost ad infinitum, as you say, but the above eight certainly would go a very long way.

    For starters, you might even reduce the four categories "postion on aircraft" to two: Vertical and fuselage. For horizontal surfaces you could probably use the fuselage pattern as is. And for vertical surfaces you might even get away with just using half of the fuse pattern.

    So for a bare minimum start, you are down to two patterns: airborne and on runway.

    And for the ones who would like to start experimenting, I would say the runway version would be nicest to have - most models are built with the landing gear down, right?

    So basically the one pattern you've already developed would be very nice to start playing with! I know it would get me a long way towards a test build. Please? For starters?

    Leif

    PS. And I do want to learn the trick of riveting!
  20. Leif Oh

    Leif Oh Member

    Joined:
    Jan 27, 2004
    Messages:
    725
    Likes Received:
    0
    With you so far. Even to the point of individualizing panels. The same idea occurred to me, that it is relatively simple to accomplish by adding a layer of individualized grey or white panes and regulate the opacity. Not a problem.

    The basic alu shine, plus a library of "weathering" (for once, the appropriate use of the term!) gradients, is the task. How about combining the the two tasks into one library of combined alu shine + weathering patterns, along two categories:

    1) position on aircraft:
    - horizontal top
    - horizontal bottom
    - vertical
    - fuse & other rounded surfaces

    2) position of aircraft:
    - runway (perhaps in a forest landscape)
    - airborne

    That would reduce the basic library to eight different patterns, four in each of the two main groups, airborne or on runway. The last group in particular could be added to almost ad infinitum, as you say, but the above eight certainly would go a very long way.

    For starters, you might even reduce the four categories "postion on aircraft" to two: Vertical and fuselage. For horizontal surfaces you could probably use the fuselage pattern as is. And for vertical surfaces you might even get away with just using half of the fuse pattern.

    So for a bare minimum start, you are down to two patterns: airborne and on runway.

    And for the ones who would like to start experimenting, I would say the runway version would be nicest to have - most models are built with the landing gear down, right?

    So basically the one pattern you've already developed would be very nice to start playing with! I know it would get me a long way towards a test build. Please? For starters?

    Leif

    PS. And I do want to learn the trick of riveting!