barninga

i'm not sure i've understood what you mean by "isolate the high frequency layer". are you trying to edit blemishes on it? when you apply the frequency separation filter, you get two layers. the low frequency layer mostly contains color, while the high freq layer contains most of the textures. this way, you can use any blemish removal tool on the low freq layer without the risk of flattening textures, since they are on the hig freq layer. you have non need to edit the hig freq layer, unless you want to edit textures. you have to select the low freq layer before using the blemish removal tools. if you apply an adjustment (color, white balance, or any oather of them) its effetc are applied to both of the layer if you place it on top of the layer stack; if you place it as a child of one of the two layers, it will be applied only to that layer. i hope i've clarified things a bit. if i misunderstood your question i apologize.

it looks like the mask should be refined a bit, maybe shrunk (so that it hides a little of the border of the church) and softened.

you just select fil->new stack in AP and choose the images you want to process; in a short time you'll get an image where all the stacked images are layers, grouped into a single stack group, where the stacking algorithm is indicated on the right. if you click on the algorithm symbol, you can select a different one and it will be applied "live". this way you can check what stacking options you have. once you select the stacking algorithm that suits your needs, you can uncheck all the images in the stack, except one, and apply adjustments and filter and live filters to the stack itself; when you're satisfied, just check all of the images again. working on one image only improves performances; when you're done, my advice is that you rightclick on the stack layer and rasterize it. this will create a static copy of it, where you will be able to apply subsequent filters and effects, or image insertions, etc. without worrying about performances. i assume that the mean algorithm (the default) should be good for star trails and smashing down digital noise..

under the filters->blur menu, there is a "minimum filter" and a "maximum filter". they work destructively.

in my experience, the ramp control (found in the refine selection dialog window), applied to an existing selection, does shrink or enlarge the selection. not by a dramatic ratio, but enough (again, in my experience) to have the selection edge following closer the border of the selected border.

the refine selection dialog has a ramp cursor, it ha the effect of shrinking or expanding the selection edge, did you try it?

you won't be disappointed by the 70d, it's powerful and really flexible. the 80d represents its evolution, but the big step further would be the 6d, mainly because it has a full frame sensor (while the 80d has an aps-c sensor, like the 70d).

if i understood how blend modes work, it's the same thing. the blend mode rules how the pixel in the layer interact with the pixels in the layers beneath. it doesn't matter if the layer which the blend mode is applied to is a pixel layer, an adjustment layer or a live filter layer: each blend mode has an algorithm associated with it, and this alogrithm is applied, no matter what kind of layer it is.

you have several options. if they are just spots within a rather uniform area, the inpainting brush could get rid of them easily. another option is to use the clone brush, to manually copy color and texture form similar areas. you might want to experiment with the tool's options, particularly opacity, hardness and aligned. the patch tool could also work. you can also select an area that features similar color and textures of the burned zone, copy&paste it, and try to arrange it so that it covers the white spot. based on my experience, i'd keep this third option as a last chance.

simpler and less powerful, i'd suggest "Exif Editor" from the apple store.it's very good when working on an image at a time and can handle a bunch of images, but it does not do batch afaik. nayway, it costs around 10$. if you feel geeky, you can use the super-powerful exiftool, freely downloadable from http://www.sno.phy.queensu.ca/~phil/exiftool/. with some shell or applescript scripting, it can be turned in a batch beast. i recently discovered that a graphical interface is available (at https://hvdwolf.github.io/pyExifToolGUI/) but i've never used it (yet) so i can't comment on it.

i can confirm that the raw files produced by the canon 70d are handled flawlessly by ap. i own a 70d and i'm able to process them with dpp or ap, the result is great both ways. imho dpp deals slightly better than ap with extreme noise (but i can't exclude it's because i just can't fine tune ap noise settings as needed); anyway ap's develop persona is waaaay more flexible and powerful than dpp, so i choose which one to use according to the settings i shot the raw images with.

if you're having troubles with the select/copy/paste as new workflow, you could try a different approach. instead of selecting, crop a single photo out of your image and select "save as", the undo. this way you get your whole image back and you're able to crop another photo, while having the previously cropped image saved on disk. just pay attention to the fact that the file name you used to save as becomes the new filename of your working file (you have the original png or jpeg anyway).

you could try adding a gaussian blur live filter and set the radius so that it smooths the lines as needed. if there are areas where you don't want the blur effect, you can click on the blur layer thumbnail and then paint black (with the paintbrush tool) where you don't need the image to be blurred. then if you select "merge visible" from the layer menu, you get a new layer where a new image with smoother lines; at this point the fill tool should work better. an alternative approach could be to select the coloured areas instead of the black lines, and then invert the selection and paint it black.

elisabeth_k, if you want to apply the same adjustment to other portions of the image, you can click on the adjustment thumbnail and paint white over the image with the paint brush; if you want instead return some of the brightened areas to their previous aspect, you can paint black.

in the layers panel, drag the levels adjustment layer onto the icon of the layer you want to apply it to exclusively, so that the adjustment gets nested to that layer.

if you want to clone from the background layer to a layer above it, set the clone tool option to current layer and below, and work on the upper layer. in the photo you posted, if your goal is to hide the white among the hairs, you can add a mask to the layer and paint black over what you want to hide; if you make a mistake you can paint white to get the details back.

well, apparently the dodge / burn tools have no effect on grayscale images in general, even if they have not been rasterized to mask...

geoffrey, the size of the image in inches has little to do with the actual file size, since it depends on the dpi setting, that you can modify at any time without affecting the file size itself: it has more to do with printing. what is the actual resolution of your image? i mean its size, base x height, in pixels? and what is its color depth, i mean how many bits are used to render a single color? the most effective way you can reduce the file size is saving it in a file format that features a strong compression. jpg is a format very suitable for that, but it lossy, that is the compression algorithm throws away some information to make the file smaller. you have a trade-off between quality and effectiveness: you can obtain a very small file (compared with the original) if you accept a loss in quality. in my experience, setting the quality level to 80% gives a strong compression and the resulting quality loss starts to be noticeable. you can determine what setting best suits your need with some experimenting. that give, you can obviously reduce the file size by scaling the image down. again, this is feasible if you can tolerate the decreased resolution of the image.

i've read somewhere else in the forums that dodge/burn doesn't work on masks. i use to paint setting the brush opacity as needed.

do you mean you have three black and white pictures you want to "assemble" to compose a color image? it looks like loading them into three pixel layers, applying the right recolour adjustment to each of them and (optionally) lowering the exposure of every single layer sould allow to build a colour image.

and here https://www.google.com/search?q=cracked+canvas+texture&newwindow=1&espv=2&biw=1664&bih=1289&tbm=isch&tbo=u&source=univ&sa=X&ved=0ahUKEwifj7HvhbfLAhVDEywKHYG5DBIQsAQIGw#imgrc=beDvse-qIsROPM%3A you can find some nice examples of cracked canvas textures.

have you tried png? it's a non-lossy format and features a good compression; in addition it retains transparency information.

the image looks like you already used the inpanitng brush on it: there are some visible artifacts below the girl's left elbow. i think that in this case the inpainting brush is not suitable to finish up the work: you'll have to use other healing tools, and particularly the clone tool and the patch tool. i'm afraid there's no way to have the job done automagically. since the job is complex and could require some trial and error iterations, my advice is to work non destructively: create a new pixel layer above the image and use the healing tools on the new layer, after setting their layer option to "current layer & below". even so, some blue will be still visible through the girl's dress, and i think it would be an overwhelming job trying to correct it spot by spot. here, you could try the recolour brush: since it doesn't affect black, there's a chance that you just paint over the dress, leaving alone the black and changing the blue to green, or whatever. the recolour brush is destructive, so you might want to apply it as the last touch to the image, after consolidating the work already done (use snapshots, or a merge visible command before you start using the brush).

place the distortion layer (i.e.: live perspective filter) above the layer you want to distort together with its mask: the ndistortion will be applied to all the underlying layers. if you don't want to distort the layers beneath, just group the distortion layer and the layer to be distorted.

you don't need to paste anything actually, just select the layer you want to mask with its own red channel and select the menu item like attached. EDIT: just like JimmiJack showed a couple of minutes ago :)