MXI

I'm just pointing at an inconsistency. If this mistake slipped into the post and test images - chances are, the plugin itself might be doing not what it says it does.

S/L gradient (of HSL) in the test pattern actually looks like S/V gradient (of HSV). Triggered. I'd expect S/L gradient to look like this:

Woo-hoo! PDN 5.0 finally brings adjustable transparency background. One slider but it will do.

My take: https://imgbox.com/zjgE42Y1 1. Resize 400% (working in higher resolution might help in multiple ways. The sample was too small); 2. Sharpen max (to make text thicker); 3. Sharpen max (one might be enough, but I found this to be a bit better); 4. Pencil Sketch (tip size max, range min); 5. Resize 50% (still 200% from the source, helps with the next step a bit); 6. Levels (adjust input white to remove as much remaining noise while retaining required details, may also adjust output middle point).

I remember using a multistep processing for this: - convert to grayscale; - invert image; - isolate the background levels in a separate layer (remove high-contrast stuff such as text with Median filter); - substract it from the image; - invert back; - clean up remaining noise with Levels filter (histogram there is also helpful to evaluate the quality of cleanup). I later found that there is actually one existing filter that can achieve about the same effect much easier and faster: Use Effects > Artistic > Pencil Sketch!

Been there, did that. It is definitely a UX issue, not specific to new users. It can cost me some time even if I know that it can happen. I (as a user) can be so deep in my thoughts about the object of my work so I can miss something about the tool. I would definitely love to have some sanity/safety check and save me some time. Although sometimes it can also happen on a layer that is not hidden by checkbox, just completely overpainted by a higher layer. This one might not have any nice solution.

Is there a way to change the color of the grid that is shown under transparent parts of an image? If no then this is a feature request. And I hope a pretty easy one. Sometimes I need to work with small sprites / decals. If I need it to be a white detail on a transparent background - I run into problem - I can't see anything on my image. For example, this render is quite misleading about what is actually drawn: There are workarounds: use different color for a time being and then change it with filters, such as "Invert colors"; add a dark layer underneath. But I find it rather annoying to have to set this up and don't forget to remove it and save properly. I would love to see a color selector(s) for the transparency grid in the app settings. If I can select non-gray colors - I will be able to avoid collisions with any colors I use (or 99% at least).

I agree about the icon - it doesn't need to be something as stressful as the red cross. But I don't agree about text change. Two "Don't"s will introduce equal amount of confusion. "Cancel" is the standard name for this behavior across all and every applications out there.

I had this issue. The few times when I needed to print something from Paint.NET, I used Levels adjustment to make image darker so it will look fine on paper. I have HP LaserJet P1005 if that matters. Paint.NET provides access to Color Management options for printing, where color profile can be set up for printers, like for displays. Maybe the defaults used in Paint.NET not good for everyone. And apps that have no such setting - get better defaults. If that's the case, then there better be a way to bypass this settings for average people who can't care less about color management and just want the damn thing to print good enough. And a comprehensible instruction how to use it for the rest of people.

Similar request:

For something as simple as Invert (or even a bit more complex) there are batch convert tools in XnView/XnViewMP and probably other similar apps. Depending on the problem you're trying to solve, ImageMagick might be an option too. I doubt the batch processing has a high priority for a paint application. Although it won't hurt to be able to leverage the power of existing plugins in this way, of course. Reuse the manually created routine in an automated manner, without the need to look for filter analogues. Question to people close to Paint.NET inner workings: How hard will it be to make a CLI application being able to run ScriptLab presets?

What can be done in plugin now: You select a color and a plugin modifies existing layer to leave only pixels with this color or pixels with any color except this. Of the plugins I have installed, - Unblend by ArgusMagnus can do the latter. - Color Accent by Kris Vandermotten sort of does the former (just desaturates the rest of image). - Color To Alpha (can't see the author) sort of does the former (fiddly controls). Maybe there are more. As for file type plugin, it might be something like existing ZIP plugin, with custom layer splitting logic: Speaking of custom logic: It might be viable to have just an external script using ImageMagick to modify the layers inside archive produced by existing ZIP file format plugin. I think this topic is somewhat related to as an industry-specific task that can benefit from better plugin API.

I agree it's tedious. I should've accented this better, or rather dropped the second part of my comment completely. It was a digression from what I wanted to post here. I think this problem, among others, can have a better solution than it is now - through improvements in layers system. Rather than investing developer time into industry-specific system, provide a way for extensions to solve tasks like this in an efficient manner. If extensions will be allowed to produce new layers and blend them back in custom ways, - it will allow to create an extension to split image into color components with a single step and not being limited to RGB. (I'd love to have HSL for example.)

Something similar to layers but separate from layers - looks excessive to me. There should be some way to mate this with layers... Recently I had few situations when I needed a way to recombine color layers. I'd love to see the layers tree, and new APIs that will allow custom blending modes and allow plugins to add new layers. Currently it's possible to extract channels one by one with "Extract channel" plugin from Ed Harvey effects pack. Clone the image to three equal layers, extract R, G, B on each layer (non-grayscale to keep color), make additive blend mode for layers, voila. No alpha though (another extension and some manual work will be needed to add alpha back), and editing might be a pain (need some accuracy with colors).

But it isn't, as I'm thinking one step further already and all of you saying step one is enough. Hearing this from Rick instead of "Nope" would've been reassuring enough for me. But all I got is the same "Clipboard is enough for everyone" story. I consider myself a developer too. I understand there are time and effort requirements in implementing any feature. I can see some of possible technical difficulties involved, and that there are things like priorities and a road map. But nothing bad in discussing possible features and what paths they open for users. Let people think in terms of possibilities and evaluate their uses for new features. By the time the feature gets into development, there better be a good understanding of how and what for this feature will be used. The answer cutting any creative input and further exploration won't be beneficial for anyone. It might end up in a situation when formally implemented feature won't or can't be used by those who were seeking for it. If ever implemented. It would be really shortsighted not to see people will start sharing what is worth to be shared and possible to do so. Local thing? I'm not familiar with this abbreviation.

You probably don't understand what I'm talking about. Clipboard is a workaround. A common interface you rely on when you don't have a better solution. Selections better be first class citizens in the app. With names. With the ability to keep and share them as a part of the project. Like layers, they are a cross-section of the project, can think of it as an orthogonal one to layers. In-app selections management will let you to provide a refined workflow for sprite sheets, UI kits, etc and spare artists time to figure out their own. Let alone saved time to switch contexts and get the right selection in place every time . Can't check the new Windows clipboard yet, but I bet it's pretty bare-bones, no chance to name things.

I feel like I'm on some Linux users forum and all the best tools were already invented - nothing will ever surpass console and text editor (or even better, console text editor). Why should I build a complicated workflow for the things that I feel can easily be just under one click of a mouse? Granted, for now it's the only way to achieve the goal. But while new versions of Paint.NET are still made, we can shape it to be a better tool. So better look at it with open mind. I don't think this feature will be like fifth wheel or something. It will be a natural extension of what we got so far.

The main difference is that Saturation and Lightness of HSL are more intuitive and easy to handle than Saturation and Value of HSV. "The HSL model attempts to resemble more perceptual color models ..." (as Wikipedia says) I have a pretty good intuition which slider and where should I drag in case of HSL. Lightness for amount of light, Saturation for how "clean" the color is. Normally I need either make color more clean or grayed while keeping brightness, or adjust brightness up or down in wide range without affecting "cleanliness". HSV has these features entangled - no way to achieve the result without twiddling with both sliders.

HSV is so damn useless. Aaaargh! I'd throw it away without hesitation. I'd suggest to redesign Colors widget and make it pluggable, so people can design their sliders for whatever color model they want.

To support the fifth request: I'd love to have Selections list (F9 for example) alongside with Layers list (F7), and have it stored in the PDN file itself. This will allow to use it as a sprite sheet - a single file with all the art pieces for a project. Less context switching, more consistency within artwork when it all in front of your eyes... A simple way to export all stored selections to files would also be helpful (point out to a folder and it will use selection names as file names). I can adapt https://hluk.github.io/CopyQ/ to manage selections, but it will always be less convenient than it should be.

This question keep reappearing. And I came to this forum to leave the very same question, because I couldn't figure it out myself, without someone showing it to me. Similar questions: (these boxes taking too much space, distracting from the essence of the comment. I'd love them to be optional) Maybe there is more. This means there is a serious usability issue. Some solution should be found to improve discoverability of this feature. Also, I'm not quite happy with two-step process when exporting to file. In some cases it makes sense to have PDN and flat file to be opened alongside, in other cases it might be simpler/faster to do a single-step export. I always envisioned this as "Export to file ..." and "Export to clipboard" items in "File" menu.

I could probably make it in a way similar to aspect ratio shapes. But I should know if it is really useful. I know there are some good svgs out there if you need a very detailed Smith chart.

I prefer it to Shape Maker. Partly due to specifics of shapes I've made. And I created converter script to automate geometry grafting labor. Check my topic for details.

What's new 2017.04.03 - added pyramid to photo shapes, updated aspect ratio shapes slightly 2016.06.02 - added superellipses Photo grids I had some photo-editing to do recently and I thought it will be handy to have cropping masks to help me with it. I found that I can implement them as custom shapes, so I decided to make a set. After little research in places like this and that I came up with this set of shapes: 3x3 grid and Phi 3x3 (Golden Ratio) grid are ultimately all that I was looking for by the moment, but it's hard to stop when you started, so I've implemented everything that looks half-reasonable at least (trying not to overload shapes with details). How to use: make a new layer over your photograph; create a grid shape there and move it to find a proper composition; the simplest way to convert a shape to selection I know is to deselect it, do a Magic Wand outside and invert selection; crop image to selection; delete this utility layer; you're done. Some shapes are sensitive to proportion. Don't forget to hold a Shift button when using them. If you need differently oriented spiral, for example, you can drag one side over other side while holding Shift to keep proportions. I by no means an experienced photographer, so if you have a better idea on what should be in the set - I would like to hear your ideas and motivation. Measurements, dimensions, other grids As I said, it was hard to stop, so I started to think what other kinds of useful shapes can be implemented. And here they are, various measurement assisting tools: And some more miscellaneous grids: Some of them may have no practical use (Smith chart in raster graphical editor, anyone?), some of them may have non-optimal parameters. This is just an experiment. Let me know if you have some use for them or something like them. Superellipses Superellipses lie in between of regular ellipses and rectangles. In some designs superellipse can be more appealing option than rounded rectangle. There are few variants in the pack, with different level of "roundness": 2.5 is a power factor of choice of Piet Hein, popularizator of superellipses; superellipse with factor 4 is called squircle; factor 5 is used in iOS icon designs; factor 3 is here just to fill the gap between other values. Download mxi_paint.net_shapes_20170403.zip Or visit my page and grab it there. Ideas for future (shapes) Parametric shape generator. Your suggestions. Converter (I'd rather prefer to move this to a different topic in Shapes subforum. Not sure how to deal with forum rules.) While making the shapes, I found it's tedious and error-prone to do conversion from Xaml pages to Paint.NET shapes manually. So I've spent some time to write a converter. It's a PowerShell script which does the simple thing - takes source Xaml files from one place and puts shape files to other place. (Source files are the files created in Kaxaml according to this instruction.) (Update: Kaxaml site is dead, I changed the link to GitHub repo if anyone willing to build it from sources. As of 2019, XAML Studio might be a good enough substitution.) Script can be obtained on GitHub. The simplest way of using it is to call it from bat file. Check for a sample bat file in the same place on GitHub and customize it for your needs. PowerShell is included in Windows since Windows 7, and there should be no issues in using it on modern versions of Windows. But I will not teach you how to use it in ways other than with a bat file. Quirks Folder "C:\Users\<username>\AppData\Local\VirtualStore\Program Files\paint.net\Shapes" is not working for Paint.NET. It's quite annoying to confirm file operations in Program Files every time. Paint.NET developers should move user manageable content to user profile. (Update: since 4.0.18, it's now possible to move shapes to C:\Users\<username>\Documents\paint.net App Files\Shapes folder. Note that it's different from "paint.net User Files" folder, though.) Shapes should be fillable. If your shape consists only of PathFigures, you need at least one figure which is closed and filled. (Update: fixed in paint.net 4.0.10) Shapes will be shown in dropdown menu and will have moving grips located according to fillable area. I had to do a hair-thin trick on measurement_mark shapes to make them work properly. (Update: fixed in paint.net 4.0.10) Default FillRule is EvenOdd. Placing one fillable geometry inside other switches filling state. You may need to set FillRule="NonZero" if you want to keep all the shape filled. To be able to quickly iterate creation process and fix all of this is a good reason for the script to exist. It allows you to test your shapes, fix source files and have proper shapes immediately, instead of having manually corrected shapes which doesn't match their sources. Ideas for future (converter) Warnings for improper shapes? (Unfillable paths, what else?) Back-conversion. Not required if you follow a good practice of editing source files only. Your suggestions.