Photoshop :: Possible To Only Resize Transparent Pixels?
May 8, 2008
I'm interested in doing a non-destructive "resize" of an image, whereby only the transparent pixels are traditionally resized and the non-transparent pixels remain the same size (but move relative to the resized transparent pixels).
As an example, imagine an image that consists of stars on a transparent background. If we were to do the above to this image, the physical space taken up by the image would increase (due to the resizing of the transparent pixels), but the stars themselves would remain the same size. In short, the stars would appear to spread out to fill more space.
I can only resize an image using pixels as unit when resampling is turned on. Why can't I resize in pixels without resampling? I can use all the other units but not pixels...
background - parting linies - head lines - information
Each layer has many many elements (like parting lines: parting line 1, parting line 2, parting line 3 > each transformed 50 pixels).
The final result is a list that looks like this: _______________ Head Line - info - info _______________ Head Line - info - info _______________
Now I want to resize the distances between each list element to 60 pixels. And it's a list that consists out of about 200 boxes. Is it possible to transform all elements the same way? Or do I have to mark each element on each layer that I want to transform? And which tool do I have to use? I have three columns and it would really take a long time to do this for each box and each element of the box.
How do I resize a photo using pixels as a unit in Elements 12? I do not wish to print it, just to resize to the size it will be viewed online.I have "upgraded" from Elements 8 where this was no problem but I cannot find the option in Elements 12.
I would like to delete all transparent pixels surrounding my irregularly (non-square) image. The TRIM function only trims to a square surrounding my image and not to the edge of the image itself. Is there a work around for this? I have looked around and haven't found one.
I have an image (of many layers) which are mostly solid color, and solid alpha. 1-bit alpha, which is exactly what i want. However there are a few parts which are slightly blurry, where it fades between the two. this is causing problems in my end result. The common white halo effect.
I've tried the flaming pear solidify plugin, and it's totally not what i want. I don't want to fill everything with color or mess around with alpha channels. What i want to do is convert all partially transparent pixels, to either 0 or 100% alpha, with no middle ground. Essentially to make the image alpha binary. ideally some plugin which lets me define the threshold for one or the other would be nice.
I'm saving as an 8bit png, which naturally forces everything to be 1-bit alpha. this is unfortunately resulting in a lot of pixels being white where they show as almost invisible in Photoshop, so i want to fix this problem author side. Ideally i'd like to not mess around with alpha channels - i like the png format because i never have to touch the alpha channel, it's all done for me which speeds up workflow dramatically. I know i can save as a 24 bit png which will store the partial alpha information, but this also increases file size which is not an acceptable compromise for my purpose.
I have two layers and some of the pixels in each layer are the same. I want the difference on the pixels that are not the same and for the pixels that are the same I want the result to be a transparent pixel. I have tried using the difference blend node but for pixels that are the same it returns black as the result not transparent.
I'm making graphics for a 2D game. One important part of performance optimisation is getting rid of unnecessary alpha sorting. I don't want partial alpha unless i have a specific need for it. I'm wanting all my images to be 1-bit alpha only. That is, a pixel is either completely opaque, or completely transparent - never inbetween.
I'm not sure how to do this easily in photoshop. So far i've come up with a workflow that fills my needs pretty well, except that it seems unnecessarily long and complicated:
1. Place the graphic on a transparent background
2. Use the magic wand to select transparent space, using the tolerance value to define the cutoff point
3. Select Refine Edge, and shift the contrast up to 100. For some reason if I don't do this, photoshop will often "partially select" a pixel, giving undesired results.
4. Hit delete. This removes all pixels that were selected (transparent enough to fall within the tolerance)
5. Select > Inverse. Shifts my selection to everything except what i just delete, ie the actual graphic.
6. Using the color picker, i pick an appropriate neutral color from the main body of the graphic. or i use black.
7. Using the pencil tool, with the draw mode set to Behind, and with a colossal radius, i draw on the graphic. This fills in all the remaining partially transparent pixels with the solid colour that i chose
This workflow does the job perfectly for me, and exactly achieves my intended results. but i can only do it on one layer at a time (i usually need to do this on about 18 layers, for a character graphic).
So what i'm looking for here, is some way to automate this process. The overall point of it really is forcing all pixels to "take a side", becoming either fully transparent, or fully opaque. Is there any better built in way to accomplish this? or a plugin? I'm open to commercial tools and plugins, even.
This is probably really basic but... I'm a printer and only really use Photoshop for changing RGB to CMYK, clear cutting (with Vertus fluid mask) and things like that. However a lot of the photos I have to clear cut have been done already but not saved (or given to me to use) with the transparency still there. Is there some easy way to turn all the white pixels into transparency? Obviously you'd have to mask any white pixels in the image you wanted to keep. A sample is attached.
I have an image that has corners with partially transparent pixels. I would like to make a mouse-over image of this with the same same dimensions, only a different color.
How can I color this image a different color and still have the partially transparent pixels around the edge. Obviously, I would like the semi-transparent pixels to be semi-transparent of the new color.
I'm still not allowed to use the link vb, but let's try this: ....
I have an image that has corners with partially transparent pixels. I would like to make a mouse-over image of this with the same same dimensions, only a different color.
How can I color this image a different color and still have the partially transparent pixels around the edge. Obviously, I would like the semi-transparent pixels to be semi-transparent of the new color.
How can I select ALL non-transparent pixels from a layer?
- Ctrl + click doesn't work (it seems to have a 50% tolerance - my opacity 7% and 40% pixels don't get selected);
- The magic wand tool with 0% tolerance does work (when I click on a transparent pixel it selects ALL opacity 0% pixels around it and then I just have to inverse selection), BUT my image has a lot of holes with transparent areas and selecting them one by one would be tedious.
I'm not sure why, but Photoshop just started doing this after I converted my files to a smart object. When I try to zoom in past 50%, it displays these transparent pixel squares over the entire image. I tried rasterizing and flatting the image back, but that didn't do anything. It seems to keep doing this.
I'm interested in doing a non-destructive "resize" of an image, whereby only the transparent pixels are traditionally resized and the non-transparent pixels remain the same size (but move relative to the resized transparent pixels).
As an example, imagine an image that consists of stars on a transparent background. If we were to do the above to this image, the physical space taken up by the image would increase (due to the resizing of the transparent pixels), but the stars themselves would remain the same size. In short, the stars would appear to spread out to fill more space.
What I need is to be able to resize normal sized images taken from a standard camera down to 504x104 pixels. All I've managed so far is to distort the images is there a way I can avoid this?
I have a client who has a set of .ai files they need resized to meet various epub layout specs. Apple and Kindle both are different sizes so I need to know the best way to resize an .ai file 11.25 x 5.65 inches into a smaller dimension in pixels while retaining the layers.
i tried to make dimensions 16x16 and keeping the bg transparent with ImageReady, and did save optimized as > then saved it png but the resolution was very bad, and the background was white not transparent.
I need to know how to make these pixels in the picture go away. I tried making the background black but It makes it look ugly. Is thier an eaiser way to do this?
How can I control the color of transparent pixels?
I save my image in RGBA (Windows bitmap). The RGB value of the completely transparent areas always turn completely black, and I want to control the color myself.
It may seem pointless, as the color is transparent, but I am using the image elsewhere (for creating mipmaps) where there is interpolation between pixels. Interpolating between 0 and 1 in opacity gives 50%, but the RGB part of the interpolation will mix with black, giving a dark halo.
I have this file and its a transparent GIF. If I resize the image and export it as a transparnet GIF and then I look at it there is a small white line all around the border. But if i save it as the orginal size and export it the same way there is no line ....
