MJW

I looked at the same file on my system, and noticed the Properties has an additional line below the Attributes line. It says: "Security: This file came from anther computer and might be blocked to help protect this computer." It has an Unblock button. This is, I'm pretty sure, what you're taking about. I can use the plugin under PDN without doing anything with this, but if you have a similar line on your Properties window, you ought to click the Unblock button. I sort of doubt you do, though, since I think you would have noticed it when you got the screenshot of the file's Properties. EDIT: Check the second Properties tab, Security, and make sure the file has "Read & execute" permission. My theory is that, being a computer from your work, it's set up to prevent newly installed files from being executable, at least by default. So unless you can change the permissions, it can't execute the DLLs. That may be why you don't get the Unblock button: no need to Unblock a non-executable file.

I don't know anything about Unlock. If no one answers soon, I'll try to figure out what that's about. I hope someone can provide a quick answer. I do know that Flip.dll is supposed to add two items, Flip Horizontal and Flip Vertical, to the the Effects menu.

That isn't a plugin; it's plugins. So you won't see an effect called BoltBaitPack or something like that. What you should see if you unzipped the file in the Program Files\paint.net\Effects folder, is a number of new effects, such as Effects>Objects>Feather Object.

Which plugin is it? Is it perhaps under one of the Effects submenus, like Objects?

I'm really really new to this so how do I install the plug in? I already have the dll file Close Paint.Net if it's running, then put the DLL file in the "Program Files\paint.net\Effects" folder. That's all you need to do to install it. Next time you run PDN, the plugin will be included in the list of Effects (or Adjustments, if it were an adjustment).

I mentioned this a while ago. I find it quite confusing. I often toggle a layer off and back on to make sure it's the layer I think it is.

dylanjosh, set all the destination pixels, but choose the pixel color based of its position. I'm not certain what you intend to do, but for the sake of demonstration, I'll make a white background with black lines in each direction spaced 50 apart (49 pixels in between). If I wanted the grid superimposed on the original image, instead of CurrentPixel = ColorBgra.White, I'd use CurrentPixel = src[x, y]. Each call to the render code isn't writing the whole destination window at one time; it's writing a portion, referred to as a Rectangle of Interest. Notice the loop bounds for y aren't 0 to dst.Height, they're rect.Top to rect.Bottom. So you can't step by 50, since the starting point can be anywhere within the destination window. Read BoltBait's CodeLab help pages carefully. They describe most everything you'll need to know to write CodeLab plugins. I'll also mention that, ignoring the other problems, what you're doing would only put pixel dots every 50 pixels rather than pixel lines. Assuming you want lines, every y row of pixels contains some grid pixels, so you can't skip over 50 y rows. (To change my code to draw dots instead of lines, replace the || OR with && AND.) (If I were actually writing code to do this, I might do it slightly differently, but this version is straightforward.) ColorBgra CurrentPixel; for (int y = rect.Top; y < rect.Bottom; y++) { if (IsCancelRequested) return; for (int x = rect.Left; x < rect.Right; x++) { // See if it's a grid pixel. if ((x % 50 == 0) || (y % 50 == 0)) CurrentPixel = ColorBgra.Black; // Grid pixel. else CurrentPixel = ColorBgra.White; // Not a grid pixel. dst[x, y] = CurrentPixel; } }

To copy the selection boundaries, you can just copy the selection from the first image, paste as a new layer into the second image, then immediately delete the new layer. The selection will remain on the original layer of the second image.

Though I think his chance of getting it is nil, I don't think it's quite that clear that what Gunnar is asking for is forbidden. If operate means modify, than he's not asking to operate on pixels outside the active layer's selection. Operate surely can't mean "use in any manner," since plugins can obviously use pixels outside the current selection to affect the results within the selection. They can also use external data stored in the clipboard.

Though I, too, wish plugins had more (read) access to other layers, and probably agree that it would be desirable for plugins to be able to temporarily -- only while the plugin is active -- change the visibility of other layers, I think your chance of getting this feature are very nearly zero. I don't believe it's a matter of implementation difficulty. Access to other layers seems to be deemed contrary to PDN plugin philosophy.

This is a sort of similar effect to the bag that might be modified to be useful. I produced it as follows: BoltBait's Gradient plugin to produce a square gradient in an 800x800 canvas. Layer Rotate/Zoom with Tiling to duplicate the pattern 5 times in each direction. Gaussian Blur of 12 to round the corners. Dents with Scale 150, Refraction 5.5, Roughness 0, Tension 11, to distort the squares. New layer with Clouds with Scale 250, Roughness 0. Blend Mode Normal, Opacity 127. This further distorts the squares. Flatten image. Posterize with 15 levels to get bands. Use my own HSV Scrambler with Hue from Value, Saturation and Value set to One. Hue Offset about 0.33, Hue Scale about 4.25. The last step was just to make the demo more colorful, and because I want to find uses for HSV Scrambler. I forgot to write down some of the plugin settings, so I repeated the procedure to determine them. Therefore, some may be slightly off, but I think they're close.

I've noticed the various suggestions have produced quite disparate results, so I wonder what Eli had in mind. Eli, is there some particular use you have for the square dents? Do you want the squares to have the same orientation, or random orientations? Do you want more-or-less true squares, or just generally straighter sides? I think more details might help. (I also think it might be difficult to fit together randomly sized and orientated squares. I think there'd probably need to be some non-squares to fill in the gaps)

One way you can achieve a regular pattern of rectangular dents (not the random dents that seem to be what's wanted) is to apply the Distort>Ripple plugin a few time, alternating vertical and horizontal, with different size and phase settings, and with the highlight set to 0. Dents can be used afterward, with large-sized dents, to add some randomness. With apologies to the lovely Ms. Lawrence:

That's quite an imaginative use of the HSV Scrambler, Maximilian. The plugin can also be used to produce effects that resemble Solarization and similar sort of metallic looks. They're kind of interesting, though it seems to be quite sensitive to JPEG noise. One of these days I hope to find something really useful it can do.

I see the processed version, but not the original, before the background was removed. I'm not sure why the Magic Wand would have left so much of a black edge. If you provide the original, with the black background (assuming you haven't already, and I just missed it), I think we can probably help you do a better job of removing the background. I don't think feathering or antialiasing will work nearly as well if you have a black edge surrounding the image. A feathered, antialiased black edge is still black, and will likely show up against the gradient background you want to add.

I'd be curious to see the original photo. I suspect you could do a better job of removing the background without leaving an edge. That, to me, seems like a better approach then removing the edge after the fact.

Yet another version, with the mountains enhanced using the Laplacian Pyramid plugin. Any anomalies, especially where the layers meet, are likely due to my haste.

Here's a very quickly done version using the multiply layer. To get the cloud layer, I pretty much just used the Magic Wand to select the clouds from my previous version, which actually seemed to work reasonably well. The multiply-layer opaqueness was set to 100. Higher opaqueness seemed to make the image a little dark. I do think the colors are somewhat improved. Using auto-level or something like that might improve the mountains further.

I may try to do it later. Unfortunately, I foolishly didn't think to save the separate layers, just the merged version. After seeing Cc's versions, I realized I probably should have punched up the mountains a bit more. There's a fine line between naturalistic and washed out.

Thanks, Rick and BoltBait. That sounds very nice.

Sounds like Cc4FuzzyHuggles and I were working along the same lines. Cc seems to have gone for deeper contrast in the mountain area. It's hard to know what the original poster wanted in that regard.

This is quickly done, so it's far from perfect, but illustrates how I suggest it be done: I separated the foreground and clouds into separate layers. The foreground was easy, just a click of the Magic Wand and a small tolerance adjustment. The clouds were more difficult, since they are sometimes hard to differentiate from the mountains. I copied the original image into a layer, used the Lasso to erase most of the area below the clouds, then used the Eraser to erase around the cloud edge. Once I had the foreground and sky in separate (higher, of course) layers. I adjusted the color and sharpness of each layer separately. I applied heavy feathering to the clouds, and lesser feathering to the foreground before flattening the image.

The selection tool you'll probably want to use is Rectangle Select. If your selection isn't quite right, you can select the Move Selection (not Move Selected Pixels) tool to adjust it prior to cropping.

Would using a polarized filter help?

How do you merge layers Besides Merge Layer Down already mentioned, you can also merge all the layers into one with Image>Flatten.