I'd do this in a page layout application though so I can't imagine which of the 20 ways to do this in Photoshop would be most ideal.
1) Double-click the background layer (white circle on black background) and name it layer 0. Then double click to the right of the name (layer 0). Up should pop the blending dialog. Pull the black "blend if" slider at the bottom of the dialog under "this layer" to somewhere near the middle of the slider bar (you should see the black bg vanish). Now with the alt key pressed, break the black slider apart the two sections apart. Where they end up is not critical, just make them separate. Hit ok. Then, use the key command ctrl-j to dupe the layer. Then hit ctrl-e to merge it down. You should now have your white circle (with a black fringe) over transparent checkerboard.
2) Make sure View menu... snap is checked. Now draw guides that bound the circle (top bottom left right) and put two intersecting in the center of the circle. These guides should "snap" into place.
3) With the polygon lasso tool, you should now be able to draw triangles that are precisely 1/8 of the circle. Draw one triangle, then with ctrl-alt-shift held down, click on the layer icon in the layers palette (you should see an x in a box next to your cursor as you hover over the icon). This should clip your selection. Now just open the pic you like, select all copy and go back to your new circle doc and edit... paste into. Repeat for all 8 pie pieces.
4) Make a new layer, move it underneath your circle layer and fill it with black.
<http://www.pixentral.com/show.php?picture=1tHMruqWn2okMPQyxftxgFTc3tgpM0>
<http://www.pixentral.com/show.php?picture=1qUbgUBPqoSQEOLcZIbU945yIXmSL>
But I'd want to do it with a vector shape, however I get there, and use the shape as a mask.
After that, then, and by copying and rotating I'd create a separate layer + mask for each image. That way each image could be moved and resized independently from the others, retaining the pie-shaped outline.
Similar to this mask-layer example I showed somebody last year, only with vector masks instead of letters.
(Click image for full size version)
<http://www.pixentral.com/show.php?picture=1L0he0TcEqWbfITy3Q2HQI18YgdGZ>
To divide a circle into 8 equal pie sections, select the circle and put it on a separate layer via CTRL-J. With that layer highlighted invoke FF and enter the following in the R G and B boxes:
R: d%128==0?0:255
G: 255
B: 255
FF splits the circle into 1024 angles. d is the angle
The R code says "If the angle divided by 128 leaves zero remainder, set R =0, else set R=255.
Thus, where d is a multiple of 128, the color is cyan (blue and green). Elsewhere it is white.
What could be cleaner?
But that's just me, and my famous pickiness.
:)
Correct me if I'm not following you. I think you are referring to what you do with the eight pie sections after they have been formed, and using them as masks is one way to go. How you delineate them is one thing, and how you make masks of them and deploy the masks is another. Using FF to form the pie sections to begin with and then selecting each piece of pie, in turn, as a mask seems as good a way to go as any.
George
I've learned the hard way too many times about trying to tweak pixel-based objects and not having them meet the quality standards I enjoy, so I do what I can to use vectors wherever possible.
(For me, anyway)
Best regards --Gernot Hoffmann
%!PS-Adobe-3.0 EPSF-3.0
%%BoundingBox: 0 0 800 800
%%Creator: Gernot Hoffmann
%%Title: N Circle Segment
%%CreationDate: November 07/2008
/mm {2.834646 mul} def
/x 150 mm def
/s 100 mm def
x x translate
s s scale % 1 --> 100mm
0.25 mm s div setlinewidth
0 0 1 setrgbcolor
0 0 1 0 360 arc
stroke
1 0 0 setrgbcolor
/N 7 def % N segments
/da 360 N div def
/an 0 def
1 1 N
{pop
0 0 moveto
an cos an sin lineto
/an an da add def
} for
stroke
showpage
When editing that code, do we replace ALL instances of 'N' with the integer of our choosing?
As long as it is not re-defined, N is 8.
I can imagine many applications, where such
a simple EPS is useful, for instance for grids.
Best regards --Gernot Hoffmann
My brain can learn and understand code, but my guts don't like it much!
: )
put it on the wishlist. AI doesn't know something simple
like a circle segment, though it's a basic PostScript
geometry element.
<http://www.adobeforums.com/webx/.59b6577c/29>
Sometimes I'm thinking that plain PostScript is more
powerful than the AI GUI (in limits, of course),
and not to talk about the idiotic blowing up of
simple PostScript graphics by AI (5kB PS --> 100kB AI).
Best regards --Gernot Hoffmann
"...AI doesn't know something simple like a circle segment..."
That's one of the reasons I really enjoy HotDoor's CadTool plugin. I don't use it all the time, but when I need it, I need it. I really ought to dig in and learn Canvas better, but the different usage conventions between Canvas and AI always bog me down.
"Sometimes I'm thinking that plain PostScript is more powerful than the
AI GUI..."
Again, I'm sure you're right, but wrapping the code in a click-friendly GUI is FAR more attractive to 99.8% of AI users...even the hard-core, long-time users. And it's quicker to use, as well.
"...and not to talk about the idiotic blowing up of simple PostScript
graphics by AI (5kB PS --> 100kB AI)."
And so we siiiiiiggggghhhh, and live with it. Short of finding another mythical vector application that has all the capability and none of the bloat, what else can we do?
My word processor doesn't have EPS as a save option. Photoshop DOES. So I enter your code as text in a PS file and save it using EPS format. When I open the saved EPS file I get an image of the rasterized text. Great! I was expecting to get an eight-segmented circle. Clearly I need more explicit help.
George
I was asked to make a choices for rasterizing; I accepted the default values— 11.111 inches for both width & height, at 72 ppi. RGB, antialiased, constrained proportions.
I added a Layer beneath the generated image and filled it with white.
Voilà!
<http://www.pixentral.com/show.php?picture=1I1EDYuszkaKZdeTV0xbYzu5Ilxos>
So then, back to you Phos---just like a TV sportscaster---Phos is going to tell us how he manages to create the vector piece of pie in PS, however arduous.
BTW, working with a rasterized shape created by Gernot's code isn't all that bad because you just change parameters for the size you want rather than suffering degradation from trying to go one-size-fits-all.
<http://home.comcast.net/~phosphor-digital/bbs/8vectorwedgemasks.psd.zip>
Siiiggghhh...explaining how I did this is a daunting task, because I've just done it twice, with some different steps each time. It's really not difficult, or super time consuming. It's just fiddly.
Both times, no matter how careful I was with anchor point placement, it seems I always end up with wedges that are just slightly under 45°. This leaves a bit of white space between adjacent wedges. I suppose that could be adjusted for, but it would take some experimentation.
That's why I like the precision of Illustrator for this type of work.
I'm thinking the third way will be extra pretty.
I get white space too, anti-aliasing in the vector render? The horizontal paths are dead on, but still make white space.
<http://home.comcast.net/~phosphor-digital/bbs/8AIvectorwedges.psd.zip>
Wow! Although you haven't revealed HOW, you have sure enough cooked up a most interesting pie. I have not worked with shape layers or vector masks, so this is adventureland for me. Your product should serve well in exploring them. There's much more there than meets the eye. Thanks for the road map.
George
There's a little thing I did with the Command + Option + Shift + T (CTRL + ALT + Shft + T) "Copy and repeat transform" function that was putting each new rotated wedge on a new layer...I did it once, but I can't figure out how I did it again. I can still do it, but for some reason there's another cumbersome little step that I don't think I had to do before.
Let me play for a little while. I'll re-float the thread again when I have something new to add.
"Spent several hours yesterday and last night into the wee hours setting up a Photoshop tutorial in prep for screen motion cap recording..." <http://www.adobeforums.com/webx?14@@.1de5f905.3c063549/945>