Reptillian

No. They usually go under testing if you want to have these codes into the g'mic filter database. They only go in outside of the filter once they seem to be "finished". I would argue some of the testing filter are already finished though. If they don't seem to be great or eh, there's really no issue about them being in the testing. Some are buggy, and that's a reason for some of them under testing. Well, I think I can demonstrate a example of converting PDN HSL Posterization to G'MIC filter with even more color space and control at some point as that is something I"ll be making soon.

Hi, I'm in the process of remaking existing PDN plugins/filters into G'MIC with credits applied and giving them more features, and I believe I have almost learned enough of G'MIC scripting to do what I want (I still need to know how to work with rotating using mathematics or something like that to learn enough though, so...). So I will be making G'MIC varients of certain filters here. And, I am posting here because I'd believe I'd have more testers of codes in here. To download G'MIC filters that aren't found in the official updates, you have to type in %appdata%/gmic into explorer, and copy and paste code into somewhere you'd like. Kate text editor works, and I'm not too sure about Notepad++,bu that can work too. Sometimes, you may have to make the tabs and things like that before using that code or just test it within code[local] filter. Report issues related to my G'MIC filter here - https://github.com/Reptorian1125/gmic-community Reptorian's GMIC code - https://github.com/Reptorian1125/gmic-community/blob/master/include/reptorian.gmic

Oh I see it now. It just that I use g'mic in different softwares, and had thought it would put to another place. And I was using a different version on PDN. So, in that case, gmic that are on the same versions in different software would share them.

Is there a way to edit PDN .gmic file? I mean, the PDN .gmic has to be somewhere. Without access to it, you can't code new filters by editing .gmic file.

There's ways to make filters for G'MIC. You can edit .g'mic file with Kate text editor and see codes for every single filters on G'MIC. You can even reuse part of filters as part of a new filter. Reference : https://gmic.eu/reference.shtml There's also a PDF file with code examples. Mind you, it is over 400 pages.

Memory issues as it seem 32-bit at the moment, I guess. G'MIC has 64-bit support.

Surprised it can work with multiple layers in that way. However, it can't output multiple layers, right?

There's an additional issue of Paint.NET filters cannot work on multiple layers at once. So, multi-layer filters or configuration will not work as expected with G'MIC on other software like GIMP or Krita.

I think at this point, there's an alternative to C# when developing filters for Paint.NET, and writing filters for this also enables other software to use the very same filters while it is arguably more maintainable as internal changes in Paint.NET probably wouldn't affect filters inside of G'MIC. I do have experience coding filters for G'MIC, and I would love to make a post describing a brief tutorial on making filter for G'MIC filter for Paint.NET along with another tutorial submitting codes to resources where G'MIC developers would push those filters. Plugin Developer seem to be limited to C# based filters, but I feel that's probably where's the right thread for it. There's the issue that a tutorial wouldn't necessarily make you a master of G'MIC filter development as you still need to understand 183+ pages documentation of G'MIC, and a lot of months of practices to develop filters, but of course I'll add that to the resource section on the tutorial. The filters I have developed for G'MIC are Vibrato Texture, and Mod. Both are very simple to generate and understand. I could give a write-out on how they work to get people started on developing those filter. Vibrato Texture is inspired by MadJik Vibrato plugin though I wish I knew how to work with multiple channels when coding for G'MIC. There are also some other filters that are inspired by PDN filters developer like Layer Cake, and Cascading Self-Glitch Filter. For people who code filters, and is interested into developing filter for G'MIC, what else am I missing out on?

The main difference between HSL and HSV is that HSL enables more control over tinted areas. That's pretty much it. That to show how they compare.

You forgot to mention .ora format. But, it's less complete than .psd. For what it's worth, there's practically zero difference in feature interchangeability or something on the line of that between pdn exported psd, and ora as paint.net lacks non-destructive editing. If PDN support layered TIFF, that can work too, but not as well as psd or ora for this purpose.

Are we really going to have a fight over what software to use? While there are cases where it makes more sense to use one software over the other for certain purposes i.e see logo and vector software or CAD blueprint and vector software (There's no real good reason to use raster for logo or presentation when vectors are generally far more suitable for those task, and some softwares just ain't compatible with raster like CAD softwares), a lot of time it's pointless to debate which software is better than the other since it ultimately comes down to the individual. To one person, PDN is a marvelous invention, and to another, it's bad. To one person, GIMP is a marvelous invention, and to another, it's bad. To one person, Photoshop is great, to another, it's bad.... I could go on.

Wouldn't libvips be more suitable here? I have created big images with Photoshop, but even there it struggles with humougous images.

I think you found your fastest solution here. A slower solution is to assign alpha using auto-leveled applied image as the base for creating alpha for a white layer, and then paste your image below, and do the necessary edits to get to what you need.

Affinity Designer is not free, but Inkscape and Scribus are free. In your case, Inkscape tracing would be perfect with the use of pixel snap extension. Of course, you have to clean up with something like Paint. NET before converting to vector form. It'll take more time, but your work will be transferable at any scale.

Are those shadows in pixel form? Personally, I would almost never use raster graphic softwares to make PDFs. I almost always use a combination of Scribus, Inkscape, and Affinity Designer to make my PDFs, and for very good reasons. The only time when raster graphics are ever acceptable for print if that you plan to print it right away without elements that are just better off with vector softwares, or you plan to mix actual vectors with raster graphics.

You might want to use a 3d program for this kind of task. If you have a height map, use that as a displacement map, and use your flag as a diffuse map. Then adjust perspective, and then render your flag over a mountain. With 2D programs, you're pretty much forced to guess how to wrap around.

Actually, I guess this is something I can code now to try to understand more about coding in general to improve on open source projects. C# is the programming language for Paint.NET right? I'm more accustomed to C++.

I'm more interested into selecting one of those option, and then the hypothetical plugins leads me to one of those result.Tried the Black and White+, but the sliders aren't exactly what I want to be seeing. There is an application where I use pixel sort to figure out gradient mapping of a image which involves applying grayscale operation to figure out how it was mapped. Some people use luminosity to map, and other use lightness, and sometimes people use one of the channel to map.

Looking for quick searches, I am not looking at any plugins that will allow me to convert a color to black and white to the following : Lightness based on the HSL model (Can't be done with gradient map reliably. Close sometimes, but not always) Luminosity (Default Black and White) (Looks close to ITU-R BT.709) Value(Not sure) Average of Channel (Not possible) Min Channel (Not possible) Max Channel (Not possible) The issue with the current one is that I need Lightness based on the HSL model, and Paint.NET has yet to have a plugin like that. And this raises another question. Why is luminosity on the default black and white different than luminosity in plugins? In plugins, it's using the HSL *L.

So, dimensions are apparently 16-bit restricted? Am I correct? That means maximum dimension are 65,335 x 65335 pixels.

There is no such thing as that in Paint.NET, and the author of Paint.NET will not be support CMYK for Paint.NET. So, you have to convert your PDN file to a CMYK program compatible software, and if you're looking into editing without paying a cent, there's only Krita. Scribus only offers a straight conversion. Plus, .ora is one way of transferring information from PDN to Krita. PSD is another. .dll is just a installation file, put it in plugin folder to install. The reason why CMYK to RGB results in less file size because 4 channels require a more information than 3 channels in general. Haven't tested LAB vs RGB vs CMYK file size to confirm. I know LAB requires more info than RGB just by the nature. LAB covers the entire human range spectrum, and that means colors RGB can't be used to display.

Krita has convert Image Color Space option, and convert layer color space option. You can use 8-bit integer, 16-bit integer, 16-bit float, and 32-bit float. You have CMYK, RGB, LAB, XYZ, and Grayscale color space as an option in Krita. To use Paint.NET exclusively, and have CMYK document option with free programs. You need to download .ora plugin for paint.net. It'll allow you to retain layers in paint.net, and use it as a file layer in Krita, so you can keep editing in Paint.NET while it can be automatically converted to CMYK in Krita. That's your best solution for CMYK while using Paint.NET for about everything except CMYK end result. .PSD is fine too, but .ORA is easily more interchangeable here.

The problem here is that pixels just ain't scalable like that. He probably used non-destructive indexing filter with a vector layer.

The only difference is that gray/white bar.