When resizing a layer with a layer mask and using any bicubic method, the layer maks scales 1 pixel differently from the other layers. Is this normal due to some component of bicubic interpolation or is it a bug?
Â
Try the following:
Create a new image in Photoshop at 800 px x 800 px
Leave the locked BG layer and create a new layer
Fill the new layer with green (or your favorite color)
Add a layer mask to the green layer and fill the mask with black (via Fill with foreground color or Alt+Del) to completely remove any visible green from your document
Now use Image-> Image Size (Opt+Cmd+I) to resize your document using any bicubic scaling method. Resize to 600 px by 600 px.
Â
Now zoom in on the document's edge at the pixel level. Is it completely white as it should be?In my case, no. However, if I use 'nearest neighbor' interpolation method when scaling, I do get a solid white edge. Even more strange, if I fill the layer mask with black by inverting it rather than filling it the problem does not happen.
i've just upgraded to photoshop CS6 and i found my image extremely sharp after resizing . then i found there're selection of bicubic/bicubic smoother/ bicubic sharper/bicubic automatic. What differences??
I have two art objects on two layers. I want to match to the pixel the two objects that otherwise could be exactly the same but one layered object was imported slightly smaller.
The scale tool has good scale handling, I just want to measure the tool spots to the pixel to gain an exact scale size factor and match sizes.
Ive been looking through various tutorials on how to create a mask that leaves the detail of the hair of the subject in
It seems all the tutorials I have looked at choose a subject with jet black hair, invert it, and hey presto, job done
Unfortunately, all my kids have got fair hair, and no matter what I try, find it almost indistinguishable from the background
How to do this, cos im losing my mind trying to paint strands of hair in and having the result look like Ive cut the image out of a magazine with a pair of left handed scissors (im right handed!)
I have a set of drum scans of medium format film images. The pixel dimensions of the scans vary slightly. But they all consist of images of the same frame size, with borders around the image (scanned film outside the image.) All of the scans were done at 4000 pixels per inch, at 100 percent.  For example, the pixels dimensions of one scan is 10492 x 6907, another is 10390 x 6968, and another is 10483 x 6976.  I want to crop all of them to 10200 x 6738 pixels. I want to delete the cropped pixels, and I do not want any scaling to take place. I want to preserve the original metadata (the scanner model, pixels per inch, etc.); in other words, I want to modify the files (or copies of the files), not create new files. The print size does not matter.I've tried several approaches, and none of them let me do what I want:  1.) Manually cropping each scan individually does not give me the control that I need for cropping to precise pixel dimensions.  2.) Creating a Crop Preset with the desired pixel dimensions doesn't work because using it results in scaling (I think.)  3.) Using Canvas Size doesn't work because the borders in the scans are not equal on all four sides. I need to be able to move and position the image inside the crop area.  4.) Using New, and Place… allows me to create a Canvas Size with the desired pixel dimensions and resolution, and move the placed images; they will fit without scaling. However, this results in a new file being created, and therefore the metadata is lost.
In Classroom in a book for CS5 this is clearly explained and community fora make mention of pixel mask, but none actually step you through it. I can not find it in the properties panel.
When I command + click a layer or mask to select the contents therein I get the 'marching ants'. After flattening + copying the selection + creating a new document with preset sizes + and pasting the selection onto the new image I get a 1 pixel border around 1 edge of the image.  I can tell I have extra space because when I change the default bkg of the 'Background' layer to black I can see the black on one side...  First image shows my selection...  Second image shows a zoom of the selection... aka: it has no extra space  Third image shows the blk background behind the copy / pasted image to show the extra space that somehow got added...
Have tried this with two images with the same result. Â 1. Load image
2. Bring up Image Size dialogue.
3. Resize image down using Bicubic Automatic or Sharpen (haven't tried others)
4. Resized image has a faded 1px border all the way round - see attached: Â I checked for some anti-aliasing or other setting but can't find one. Â This is 100% reproducible and the border is not present in the original image. I'm in XP Pro Sp3.
I am working on a tut and it calls for the use of a Pixel Mask.However, once I get through it I plan to use the resulting image at various sizes.I am creating the image w/a canvas size of 425px by 200px and the object itself is roughly 155px by 45px (give or take a pixel or so).
I then plan to scale it down to various sizes and so my question is will the Pixel Mask cause a problem with artifacts and interpolated bluriness?ie, will saving for the web at smaller sizes cause a problem with the clarity?
My subject line pretty much sums up what I want to ask.
The thing is I've learned how to do both of them, but every textbook exercise I've done regarding one I found out that I can pull off with the other too.
Are they simply two different mechanisms for doing the same thing?
When I go to the properties panel, enter the feather radius „0.5“ and hit the enter key it rounds it up to 1 (Pixels).
Entering „0.4“ and hitting enter gives 0. 4.3 gives 4 4.6 gives 5
and so on. Â So hitting enter rounds the value, although decimal values are accepted and do work. Why is that? What's the purpose of rounding the number?
When I enter „0.5“ and click somewhere it works without changing it to 1. Then it's really just 0.5 px. But hitting enter (the most natural thing after entering some value) rounds it.  Is this a feature or a flaw?
I have spent hours trying to do this simple task in Gimp without luck. What I'm looking for is the border effect (without outer shadow) seen in the three embedded screenshots in this image: [URL]
In inkscape it's a matter of adding a square on top of the image with identical dimensions and making the stroke colour semitransparent and the fill colour completely transparent, but image quality suffers badly when exporting from svg to png or jpg again, and I'd prefer doing it in Gimp.
So, take image, say 100 by 100, add 1 pixel inner transparent border - how? I have seen this effect enough to suspect it's a filter option but haven't found anything in my filters.
Level: Newbie  OS: Win7 64bit  Ia: Cs6  Once I use a Clipping Mask on a layer is the entire document covered in a mask?  And/or is all the artwork from then on inside the mask?  I've read about how to add and how to remove art from a mask. I've read about how to make and release and lots of other things.  So, then I went and found a lesson that contained the use of a Clipping Mask.  I am suppose to Select the main shape and "just" copy it and move it down (doesn't say rather to Paste in front or back but I assumed in back ~ not that it has mattered thus far)  However, while making the main shape I used Gradient Mesh and used an Offset to create a "replica" and so when I go to Select the main shape the mesh is selected and so when I Paste (in back ~ seems most logical) and then move it down the details such as the colors of the Gradient and other such things are visible in front ~ ??  So, I have been trying to abate my problem in lots of different ways.  I have tried using the Pen tool to draw behind the main shape, I've tried making a New Layer and placing it above and below and inbetween the main shape layer, I've tried adding the artwork to the Mask ... I've tried everything I can think of and I've been through many an article in the manual and FAQ's to no avail.  I suppose, what I think I need to know ... is, how ... how on earth do I draw behind this main shape? Am I not able to do it because I am not getting behind the Mask?
The problem is almost prehistoric. It dates from the very beginning as far I can tell... whatever image you like to export and whatever its format: png, gif, jpg, there is always one grey pixel border line at the edges. This happens when AA is activated (and deactivate it is not an option if you know what I mean).
I have been reading some threads and no matter what I do: Exporting with or without the background or changing its color... it's always happening.
 I Have now installed the trial version of Corel Draw X5 (sp1 included) as I do with every release in order to check out if the problem persist... but no luck for me: nobody seems to be taking care of this issue. I really don't know why Corel is bumping off some many (potentially) buyers like me and many others... why Corel is keeping this issue unresolved year after year, release after release?Â
I'm making an ebook cover with a white background and I want to put a grey pixel cell frame around it but I don't know how to do it. I'm only getting started with getpaint.net.
Why when I make a path for example and make a selection from it every time the selection is 1 pixel wider then the path. I'm used to it with Photoshop  and it's very precise even you use the lasso tool and paths. But here I can't understand the logic in that 1 pixel more. Every time I make a selection I have to reduce the size of the selection with 1 pixel and that is very uncomfortable.
Today i opened a psd file which 200 layers. Paint.net is very slow then (as all layers are flatten, pdn is missing hierarchical layers), but that is another problem.
I wanted to hide some background layers and "copy merged" a selection.
It took me 10mn to find all the backgrounds layers without this tool.
In photoshop the tool exist as a property of the "move selection" tool. Ie: you must first select the move selection tool, then select "layer" in the properties toolbar.
Any way to get that for v4 ? Will v4 be on github so we can clone and make patches ?
I can do one or another, but both at the same time? It is annoying because a mask selection turns into layer selection when you select a second layer/folder.
Is there a difference between selecting your layer and clicking Add Layer Mask Icon versus having that layer selected, hold down ctrl and select Add Layer Mask Icon?
Also regarding working in Lightroom, if I modify a jpeg, select another image from filmstrip to work on and then go back to the first image for more work, are those actions degrading image even though I never physically save and close.
I've often been annoyed by the fact that that I need to go into the layer panel each time I need to switch between editing a layer and editing it's mask. Surely there must be a way to do this with a shortcut. I found a way to disable the layer mask in the shortcut menu, but this is not something I need to do that often. Switching between layer and layer mask, however, happens all the time - at least in my workflow.
I can't figure out how I can use a text or shape layer as a mask for the layer below. See the example attached and you will know what I mean. (I created the example with simply cutting out the text outline from the white box). Â I need the text to be a mask and I need it to be editable as text. Don't know if this is possible, experimented with clipping masks but I cant make it work.
I`m trying to swich a image on a layer whit a layer mask. I whant to keep the layer mask but change the image. When I paste a copyed image it the laye it creates a hole new layer. How can I change a image but keep the layer mask?Â
Basically exactly how layer masks behave, except I would like to be able to have full non-destructive editing control over the masking layer (e.g. have it as a smart object). The problem with layer masks is that (AFAIK) only destructive editing is possible.
What I'm after is effectively the same as what a 'luma matte' does in After Effects. Is this at all possible in PS?