Reptillian

Members
  • Content Count

    109
  • Joined

  • Last visited

Everything posted by Reptillian

  1. Well, not sure if I should update this thread as frequently seeing as PDN G'MIC no longer updates. Made a filter to reproduce this - https://m.imgur.com/gallery/gP0lecI --- Example using a image generated from thispersondoesnotexist.com I really wish I can use apng here.
  2. Hmm, another idea I can think of. G'MIC does support updating offline. I wonder if it would be possible to use that to get around this issue. One would have to disable weekly update, and use another updater with using the latest g'mic. You would have to rename updatexxxx where xxxx is the latest version, and yyyy is the current PDN version. This isn't ideal, I know, but it's the only thing I can think of. Maybe another filter just to do this?
  3. Hmm, I refreshed into PDN and not seeing anything. Krita and GIMP does have the palettes support using latest beta G'MIC refreshed. The palettes should be in Testing-> Reptorian -> Premade Palette. To reduce colors using those palettes, they should be under transfer color reduced, and you can change the reference layer option to palette to reduce using palettes. If they're not there, then at this point, any changes I made will only be seen in G'MIC 2.5 PDN which won't be arriving in a long while. @null54 Would it be possible to provide beta version of G'MIC? Krita and GIMP have the benefit of using beta G'MIC while PDN just does not have that benefit.
  4. Okay, those changes are merged through master, and now can be seen via refreshed. You should see premade palette filter. The emboss cli is however here as well as changes through the other two filters related to it. Here's the palette list extracted from the g'mic code
  5. Premade Palette filter has been done. Blend with image option is for PDN support is to avoid creating new layer as PDN doesn't support creating new layer via filters. In Pre-Made Palette Filter, there is a info box where it'll tell you the command, and their id. pal 129 = pal zx. So, in code[local], typing in 'pal zx rv' would get you the same result as 'pal 129 rv', and the result of the zx spectrum palette. The total amount of palettes that is support by pal command - 137 Just for wait for the changes to get into G'MIC.
  6. Made some new changes that will be coming. Also, I think @nycos62 would love that I added 128 palettes into G'MIC. Change #1) Added 128 palettes. Some of them are based off Windows, others are based off older systems, and many of them are palettes to support pixel artists. You can type in 'pal db32 rv' in code[local] to have db32 palette directly into PDN after you apply. It's hard to get the full list, but you can get the list here - https://github.com/dtschump/gmic-community/pull/155/commits/b309ec158a0f55f6bc85a9a7217b365b5c812ebe edg = endesga g8 = generic-8 sepac = amiga 2600 sepac, i should have done amiga2600sepac, but sepac is shorter and less memorable than ntsc and pal,so sepac it is. Change #2) Transfer Color[Reduced Colors] includes palette options. You do not need to copy and paste palette to reduce colors based on those palettes. Change #3) Improved Emboss cli command. Now you only need to do this at the minimum 'em 140' to emboss at angle 140. You only need to assign 3 variable at most. You can also do 'em 140,2' to emboss at angle 140 at 200% depth. 200% depth is the maximum depth. em is eq. to emboss_image emboss_image $1,$2,$3 $1 = Angle $2 = Depth $3 = Smoothness Change #4) Two of my filter were updated to be consistent with the changes to emboss cli.
  7. Pixel art palette command has been added to gmic. See cli pal for more info.
  8. Emboss is now is it own cli command. You can test it via code[local]. No changes other than this for now, but this makes it easier to do filters, and a bit more faster now since it uses less lines of code. You may want to update every once in a while until construction material texture or Butterworth Bandpass isn't broken. I'll give it tomorrow. Emboss/Relief uses em command instead of multiple lines of unnecessary codes. em $1,$2,$3,$4 $1 = Smoothness $2 = Angle $3 = Depth $4 = Depth Boost Note: em = emboss_image
  9. XOR can be used for creating abstract art or glitch art. Actually, it works better with other binary modes if one were to create abstract art. Also, nice find, could be turned into a filter, but that's not necessary when there are other similar result filter.
  10. Joan Rake is back, so the changes that I applied to Cascading self glitch filter will be there. She has merged the changes. I also do collaborate with her from time to time. However, I will be taking a break, but as always, you are welcome to leave issues or feature requests in here or in here or by discuss.pixls via PM -https://github.com/Reptorian1125/gmic-community How do I enable issues on github? P.S I consider changes to default parameters as well. I have considered trying to add existing palette to transfer color reduced as well as chroma edit by ab channels into construction material, so doing those will be done after the break.
  11. Actually, most quantify their rendering to 8BPC which means they can produce 256^4 (RGBA) color value combination. Affinity Designer can render at 16 BPC. But my statemwnt still stand that you should use vector graphics if you are going over 2000 dpi. Those apps don't care about pixels in a way which adds information. Let me do a little math here. 2000 dpi 8.5*11. 1*3 at 2000 DPI = 6000 * 2000 px 8.5*11 at 2000 dpi = 17000 * 22000 px That's too much info that can be reduced via vector route.
  12. Honestly, you are just better off with a vector graphic software if you need 2400 dpi. Takes way less memory there. Why do you need 2400 dpi with a raster graphic software?
  13. With that update, only XOR didn't get loaded from Krita ora. I could try to add those to the cpp file whenever I get the chance.
  14. Paint.NET G'MIC QT can convert image to a palette, and I have already made a filter a while ago which allows you to convert pictures to reduced colors picture using a reference picture, but that appears to be not what you want. This below should really be turned into a filter, rather than a command G'MIC QT -> code local -> colormap Reference - https://gmic.eu/tutorial/_colormap.shtml
  15. It's a unsolved Windows bug. This happen with other softwares too, not just PDN. Linux clipboard works just fine and Mac as well.
  16. Here's the name info inside stack.xml in .ora, and it's not they made the patch. I made the patch actually. <image w="256" version="0.0.1" h="256" xres="300" yres="300"> <stack visibility="visible" isolation="isolate" opacity="1" name="root" composite-op="svg:src-over"> <layer src="data/layer4.png" visibility="visible" opacity="1" selected="true" name="Layer 4" composite-op="krita:xor"/> <layer src="data/layer3.png" visibility="visible" opacity="1" name="Layer 3" composite-op="krita:negation"/> <layer src="data/layer2.png" visibility="visible" opacity="1" name="Copy of Layer 2" composite-op="krita:reflect"/> <layer src="data/layer1.png" visibility="visible" opacity="1" name="Layer 2" composite-op="krita:glow"/> <layer src="data/layer0.png" visibility="visible" opacity="1" name="Layer 1" composite-op="svg:src-over"/> </stack> </image>
  17. That because it been commited yesterday and can be tested via Krita NEXT - https://phabricator.kde.org/D15584 So, should I export ora using Krita NEXT?
  18. Is there a way to add a interface for changing blend modes compatibility? Right now, I'm interested in adding a checkmark tool to allow compatibility for Krita. The reason? It is the first notable program to support every PDN blend modes as of today and gmic existence.
  19. Transfer Colors [Reduced Colors] is now on G'MIC. Update via G'MIC QT.
  20. Solved the LAB editing issue on transfer color[reduced colors] I found out that the contrast effect goes over the range of [0,255] so, I had to apply cut 0,255 command. Now, it's ready for use though there's still a issue with preview, but however, that wouldn't affect productivity so much.
  21. In relations to my last few post, I have implemented triangle wave in cascading glitch filter which would be E option. However, I still want to wait on that as the original author isn't back on developing filter. I'll give it 2-3 months for E, and if she had not returned, I might have to put my own version on reptorian.gmic if i am not granted permission to edit joan_rake.gmic file. It turns out I still need to do work on reduced color filters, so I'll figure out something sooner or later, but it's not a filter I would prioritize on since pixel art isn't really my thing. I'm working on finally implementing modulo operations cli into a filter. The main thing I need to figure out is how to apply those operations per every two layers and by channels separately and then blend them.
  22. Since at the moment, I'm not really ready to update my gmic as I still want to do a few more filters before I upload to g'mic community, I decided to prematurely release the g'mic code for transfer color [reduced colors]. This code will be deleted once it gets to G'MIC of course like I did with the Vibrato code. 1. Type in %appdata%/gmic in Windows Explorer 2. Copy and paste this inside updatexxxx.gmic somewhere between the end of preview of a filter, and before #@gui, and then go to paint.net -> advanced -> g'mic QT, type in Reduced Colors and this should show up. #@gui Transfer Colors [Reduced Colors] : tcrc, tcrc_preview(0) #@gui : sep = separator(), note = note ("<b>Transfer Color Setting<b>") #@gui : note = note ("<i>If using all layers as input layer, the layer is the reference layer. If using active and above/below, the layer is the target layer.</i>") #@gui : Layer to Use as Reference/Target Image = choice(0,"Top Layer","Bottom Layer") #@gui : 1. Is reference image smaller than target image? = bool(0) #@gui : 2. Color Count = int(16,2,4096) #@gui : 3. Color Dithering (%) = float(50,0,100) #@gui : 4. Color Indexing Method = choice(1,"Median Cut","Median Cut and K-Means") #@gui : 5. Alpha Value Count = int (2,2,255) #@gui : 6. Alpha Dithering (%) = float(50,0,100) #@gui : 7. Alpha Indexing Method = choice(1,"Median","K-Means") #@gui : sep = separator(), note = note ("<b>Local Contrast<b>") #@gui : 7. Spatial Radius = float(80,2.5,200) #@gui : 8. Amount = float(0.5,0,5) #@gui : 9. Darkness Level = float(1,0,4) #@gui : 10. Lightness Level = float(1,0,4) #@gui : sep = separator(), note = note ("<b>RGB Adjustments<b>") #@gui : 11. Brightness (%) = float(0,-100,100) #@gui : 12. Contrast (%) = float(0,-100,100) #@gui : 13. Gamma (%) = float(0,-100,100) #@gui : 14. Hue Shift (%) = float(0,-180,180) #@gui : 15. Saturation (%) = float(0,-100,100) #@gui : sep = separator(), note = note ("<b>LAB Adjustments<b>") #@gui : 16. Lightness Brightness (%) = float(0,-100,100) #@gui : 17. Lightness Contrast (%) = float(0,-100,100) #@gui : 18. Red Chroma (%) = float(0,-100,100) #@gui : 19. Green Chroma (%) = float(0,-100,100) #@gui : 20. Blue Chroma (%) = float(0,-100,100) #@gui : 21. Yellow Chroma (%) = float(0,-100,100) #@gui : sep = separator(), Preview type = choice("Full","Forward horizontal","Forward vertical","Backward horizontal","Backward vertical","Duplicate top","Duplicate left","Duplicate bottom","Duplicate right","Duplicate horizontal","Duplicate vertical","Checkered","Checkered inverse") #@gui : sep = separator(), note = note("<small>Author : <i>Reptorian</i> Latest update: <i>2019/1/31</i>.</small>") tcrc : if {$1==1} rv fi ac[^0] "tcrc_adjustment ${9-23}",rgb rv[0,{$!-1}] repeat {$!-1} l[{$!-1},$<] tcrc_colormapping[0,1] ${2-8} endl done rv[0,{$!-1}] if {$1==0} rv fi tcrc_colormapping : rv split_opacity[1] to_rgba[1] +colormap[0] {$1?$2+1:$2},$4,1 index[1] [3],{$3/100},1 l[2] skip {ia==255} if {ia!=255} autoindex $5,{$6/100},{$7} n 0,255 fi endl rm[3] to_rgb[1] a[1,2] c rv tcrc_adjustment : fx_LCE ${1-4},1 adjust_colors ${5-9},0,255 rgb2lab s c l[0] adjust_colors $10,$11 endl l[1] +adjust_colors 0,$12 blend lighten +adjust_colors 0,$13 blend darken endl l[2] +adjust_colors 0,$14 blend darken +adjust_colors 0,$15 blend lighten endl a c lab2rgb tcrc_preview : tcrc $* Reference Layer - http://amiga.lychesis.net/resolution/Lowres/Corruption_17_Casino.html Target Layer - https://www.pexels.com/photo/eiffel-tower-paris-1267533/ Result of Reference Layer to Target Layer
  23. Well, it seem like I decided to continue making new G'MIC filter as my main computer is overheating and I stepped out of doing filters for a while. I'm using my secondary computer. My options - A ) Finish Transfer [Reduced Colors] filter, this filter basically transfer colors from reference image to a target image using a palette of colors from the reference image, and the target result contains only the palette colors. From 2 colors to 4096 colors. It works well with PDN limitations, but I also have to make it work as I want it to work with 3+ layers the way I want to on GIMP and Krita as well, so that is why it is not on G'MIC yet. @nycos62 , does this benefit pixel artists? If it does, should I release this regardless of the bug? B ) Pixel-art scaling C ) Wood texture D) Improve my Construction filter, any suggestion welcome. I thought of laplacian normalization processing. And also I can add color processing using LAB channels and adding red,green,blue, yellow sliders. Are these the only thing that needs to be added? Is there a technique I can implement akin to Butterworth Bandpass processing? Should I add description? E) Implement Modulo Blending Modes to self-cascading glitch filter. Regular modulo already exist, so this would be about continuous regular modulo, divisive modulo, continuous divisive modulo, modulo addition, continuous modulo addition. F) Bubble filter Besides those long filters, I might update my modulos operation using modf and modo while being able to perform them on separate channels using a number or image for calculation.
  24. I tested colormap command with alpha, and it does indeed work. Do you want the alpha values and the color values to be separated. Duplicate layer If so, in code[local] split_opacity colormap [desired palette size],1,1 rv Remove rv for one of those layers and you'll get both of them separated. Mind you, there are 255 colors possible in alpha in PDN.