jdavee

OK Rick, I'll bite - what is the reason for linking to a single tolerance level across all tools? My assumption was that this was just an internal programming decision that could be enhanced to provide more functionality.

I'm glad this isn't something already discussed and dismissed, and that others see some value to deglobalizing the tolerance setting. Is there a procedure to submit enhancement suggestions? Thanks for the layering suggestion - I'll give it a try.

I just hit a real frustration problem with the tolerance setting. It's global, and the value as set will affect all tools that use it. I searched the forums to see if there was any reference to this issue and didn't find one, so I'm posting for discussion (maybe I'm not looking in the right place) or enhancement if it doesn't exist. Problem: I have a PNG image that has a transparent border. I want to select the pixels in the image that are not transparent (it's a photo-type object with non-geometric edges) then replace the object with a patterned fill. Process: 1. Select Magic Wand and click on the transparent portion of the image (I have all but what I want selected) 2. Invert the selection (I have what I want selected) 3. Select the Fill bucket and fill the selection with the desired patterned fill (a checkerboard, let's say). Problem: 1. To properly select the transparent portion of the image without the object and retain a sharp edge, I need to set a low tolerance level. 2. To properly fill the object I need to set a high tolerance level. The photo object has typical color variety - setting a high tolerance will fill the entire thing to the selection border and not miss any pixels. 3. The Tolerance level is global and its setting is used by all tools that understand it. Resulting Process: 1. Set Tolerance to low 2. Select Magic Wand and click on transparent portion 3. Invert selection 4. Set Tolerance to high 5. Select Fill bucket and fill the selection Forgetting to set the tolerance to the desired level based on the tool in use produces unwanted results. Proposed Solution: I'm a programmer, so I envision the new Tolerance bar as an array of values instead of a single value. The array saves tolerance levels individually for each tool, and when I switch tools the tolerance level changes as well. When I have the Magic Wand and change the tolerance, that new value is stored for Magic Wand and all other settings are unaffected. When the app opens the values are set to some default setting (or user defined settings if there is such a startup setup process, or remembered settings from the last session if state values are retained). Am I missing this ability for Tolerance to be set individually by tool? Are others having this type of problems as well? Thanks, Jim

More things to try! I really appreciate the suggestions! It's good to know there are so may options with the features already in PDN.

Rick & ssaamm, thanks for the additional details. I'll look into the PDN format for this project. The final output is PNG with the neutral background being the transparent color. That's required for the visual display features later on the project. I'll have to see if it's easier to stay with PDN while editing then bulk convert to PNG (and set the transparent color...) later on.

I'll revisit the 15 degree issue. I'm sure you're right - I'm just not using it correctly. Thanks!

The project is to create a faux 360 degree rotation of specific object (mostly computer hardware). 40 still images are taken from different angles, and when the object is shown in a neutral field the user can use up/down/left/right arrows or mouse actions to effectively rotate to view all sides. The effort will build a database of these image sets. At 40 images per object, the number of still images builds quickly. It's not the cutout work itself, it's the volume. Initial efforts used a green screen background, but the chroma key wasn't providing a clean edge, especially with computer cards that have similar green hues. We could mess with the tolerance on the magic wand and get close, but never as clean as polygon selection and cutout editing. And we don't want to mess with the tolerance for each image, and polygon selection proved to be faster and cleaner. So with cutout editing the remaining method, we were able to capture the images without concern for the background. Since we aren't doing anything besides cutouts, we were using the SnagIt Image Editor and it's polygon selection tool. But SnagIt (at least v8) lacks an inverse selection, which was needed for some additional variations on images already cut out. That lead me to Paint.net, which has a simple inverse selection (magic wand the cutout outside the image, then inverse the selection to get just the object. But Paint.net doesn't have polygon selection. It's a much better editor than SnagIt, but difficult to perform this cutout task without a polygon selection.

Sorry if the post duplicated - site didn't report the comment as updated and I tried a 2nd time.

But certainly it knows what the currently opened image is, doesn't it? If I open a 256 color PNG 640x480 file, it should know this. And if I save it still at 256 colors and 640x480, it should know that nothing was changed. In that scenario, how does the Save dialog help me (other than providing options that I'm not going to change)?

OK, but the line/curve tool is limited to 15 degrees of straight. That's not fine enough. My typical project involves taking a series of images of an object then cutting out the background so the item is suspended in a neutral field. Often I can't capture the image in a green screen or other setup to control the background. A 15 degree straight line limitation is not enough to follow an object's outline. I've also got to learn how to better connect lines with the tool. When I tried the line tool, I was unable to reliably connect the end of one line to the start of the next by doing a "click to start / click to end / click to start line 2". The second click doesn't start a new line but instead activates the end of the first line so I can move it. Not what I want to do... I really appreciate the suggestions, however. They're able to do a faux polygon selection, but none match the real deal.

Yeah, that's the layering trick I found earlier. Functional, but still not quite "it". Mostly due to the lack of straight line locking. My hand-drawn lines aren't very straight. That's really the "Part B" of my request - some ability to control the line drawing. Freehand is one thing, fixed objects another. PDN has all of that. But nothing in between, the freehand polygon if you will.

New to Paint.net, and like others I noted the lack of a polygon select tool right off the bat. I found some polygon draw plug ins, and a description of a layering trick to draw a pencil line on a new layer to effectively select the desired area, the remove the pencil layer to get the selection from the base image. Functional, but not as simple as a basic polygon selection. The lasso has about 90% of what's needed already - auto completion to the start point, visible coverage of the selected area, etc. But it's too freehand, picking up every shake and jumble when I move the mouse any decent distance. If the lasso had a "forced straight line" feature that would help a bit, although the requirement drag the mouse instead of clicking then moving still makes it more difficult than it needs to be. What it's missing is straight lines and easier mouse movement. Other tools (open source and purchased) have normally had some method to force straight line movement. I've seen it mostly done by holding <ctrl> or <shift> while drawing. I know there is a 15 degree straight line feature like that, but that still leaves lots of angles not able to be drawn. My preference? A polygon select tool that I left click once, setting an anchor point, then move from that anchor to a new point (not click and drag, mind you), click to set another anchor point, move to a new point, etc. Double click to close the polygon and get the selected area. Does such a polygon selector plug in exist?

I found the prior discussion on the Save Dialog in this forum. Unfortunately it devolved into some bitter commentary. However, the main concept is worth revisiting. I have hundreds of images on which I'm performing very basic editing. The mandatory Save Dialog is providing me with nothing I need to change. It only causes delay in my work effort. I can understand opening the save dialog if I change formats or do a "Save As", but here I'm opening a PNG, editing it, and saving the results. No change to bit depth, format, etc. Why would a mandatory Save Dialog be needed here? I know it's only "hit enter", but after a few dozen times one does question the value of the mandate. Not to mention the efficiency of moving the mouse from the upper left "Save" button to the lower middle "OK" button on the dialog, or removing my hand from the mouse to hit "enter" then moving back to the mouse again. Or releasing my Jolt Cola with my non-mouse hand to hit "enter" . A few times is nothing. Dozens or hundreds is not. I'm not suggesting it be eliminated, only more selective when it appears, or user selectable via toggle. I will concede that there may be reasons to keep it for which I am unaware. The prior thread was not able to get to that level in the discussion. If there are such reasons, I'd certainly like to understand them better. One might be this: I noted a comment from a developer in a different thread that Paint.net will never default to something that could introduce data loss (this in regards to setting JPG as the default save format). Laudable, but I don't think that logic is applicable to my scenario.