Jump to content

Reptillian

Members
  • Posts

    1,236
  • Joined

  • Last visited

  • Days Won

    18

Everything posted by Reptillian

  1. @Ego Eram Reputo @lynxster4 Thank you for the feedback. So far, the gradient bars is nearly done. For proof (I got duotone color to work): TODO: 1. Add Duotone option for spaces. 2. Add animation support. For such a simple idea, that's a lot of work. Maybe for just spaces filling, I could just make a separate command for that, and then combine both as a gui filter. Side note: Lyupanov Fractal is under work
  2. @MJW On the response to you, I would like to demonstrate that it is feasible as a proof of concept. Here's my test of that idea, but this time, I had not used a loop per all the color in the image.

     

    This is what it looks like:

     

    ( Left - Original : Middle - Threshold Distance : Right - Looped Average Distance )

     

    KlkuHUn.png

    Not sure if this helps to gather insight, but just to throw this out there anyway.

     

    EDIT: Another image with more colormap value.

     

    01a5VDv.png

  3. There is another way to do this, but it's pretty slow. You would have to loop using threshold value, then normalize. However, I found that it leads to dark image if using a image with large number of transparent pixels rather than just the edge. EDIT: You can index colors to specified numbers of colors and then loop per threshold images with distance finder, and then you'll get what you want. I did posted a picture in my profile to demonstrate proof of concept. It is quite ok in performance.
  4. More has been done for gradient bars. Now, featuring random gradient mode. TODO: Symmetry Mode Color Input
  5. Decided not gonna release that filter because there is a problem with current implementation of quadrilateral distort. Particularly when the quad is so distorted that nan values start showing up. So, I went to work into a random bars filter. One that will eventually be more flexible than @pyrochild gradient bars. None of his/her code has been used, it's entirely a new code.
  6. I did wanted to completely convert gmic thorn fractal as a gmic-pdn obl plugin. Particularly the problem of dynamic gui would be solved with tabs. I'm not going to stop you, but there are others that see a use in this. Does anyone else have a suggestion to this problem? That filter use dynamic gui, and it is apparent when using custom formula. The current pdn thorn fractal lacks it because of inherent limitations of static gui.
  7. @Ego Eram Reputo Source code gone. Plugin gone.
  8. Some of you may notice that some of my filters has broke. That's because @G'MIC has updated gmic internals to have functions like fill(), and repeat(), and I had made a commit to replace for loop with repeat(). It's done because they are to be optimized. So, one should wait until next g'mic update to use them.
  9. There's actually a easy way to do this, and I think I remember @Ego Eram Reputo have a plugin that does import color curve map. I have made a filter called "Reverse Engineer Gradient Map" in G'MIC which basically creates a color curve profile by comparing via 2 images. This is what it looks like to get a match: If one can import that 255xH image and then plug it in as a color curve map, then one has a quick step of doing this.
  10. On your skew thing, I made G'MIC Skew a while ago, and I added more features to G'MIC Skew recently. This update will enable enlarging Skew. Mind you, be careful of the angle, and enlarge option.

    1. Show previous comments  2 more
    2. Reptillian

      Reptillian

      That is the plugin, it has a internal updater as well.

    3. mszlazak

      mszlazak

      Got it. Thank you.

    4. Reptillian

      Reptillian

      As for the warning question, large angle means that the dimension to fit in the skewed dimension is quite large.

       

      To demonstrate:

       

      tan(0°)=0

      tan(45°)=1

      tan(60°)≈1.73

      tan(80°)≈5.61

      tan(89.9°)≈593

       

      So, if you're skewing a 1000x1500 on x-axis with angle set to 89.9°, that would mean the resulting dimension is about 593000x1500.

  11. Can you show a example of the effect that is to be achieved?
  12. That's not possible with PDN. Also skewing with a large angle means that the enlarged picture would very big easily. If using 80 degree, the enlarged picture would add 5k to both side for 1k picture.
  13. I'm wondering if there can be a find/replace function similar to the one implemented in KDE Kate. This is more for faster editing. I would just copy and paste to KDE Kate, and then back to CodeLab.
  14. I'm actually working on porting my gmic skew filter into skew plugin using @BoltBait original code as a starting point. That filter has angle input. EDIT: I got something here though I don't know how to work with this too well. Maybe the author will improve this code.
  15. Added new filter! Skew! Thanks to @BoltBait for inspiration, and @mszlazak for the idea of using angle. It should be here in a hour or two. Probably here by the time you're reading this comment. So check by. Commit: https://github.com/dtschump/gmic-community/commit/9ae1f91f611840cafc893b8c85e08aada62d39b5
  16. She want to define the skew based on angle (I think that is what she wants). You can actually find the skew displacement at the end points based on angle by this formula (for horizontal skew): #hh is Height - 1 or end y indice# const dist=hh/2*tan(ang);
  17. Gmic 2.9.7 been released. Also, new notable filter:
  18. @lynxster4 Is there any special rules for plugins that are essentially multiple plugins in one? There is g'mic and that other plugin whose name escape me, but has multiple filters, and psplugin.
  19. Let me get this straight. You want something like this? If it does, I will push it since it was 12 lines of codes. Though in this case, you would need to install the gmic plugin. If you do need to find the original position, I can add that too.
×
×
  • Create New...