Tips I Index
|The following are posts written to the Usenet Photoshop newsgroups.
Triangles Want precise angles? (This will make an isosceles triangle which will be equilateral if you use 60 degree angles.) 1) Open a new file in white.
2) Draw a straight horizontal line with the line tool holding down the shift key.
3) Create a new layer.
4) Draw another straight line.
5) Copy this layer.
6) In the layers palette activate the layer you just made and go to Edit -> Transform -> Numeric and type in the angle you want, say 60 degrees.
7) Go back to the first line you made, duplicate the layer, and again on this
new layer, do Edit -> Transform -> Numeric but this time put a Minus (-) sign in front of the
number to get a negative angle.
8) Position the lines how you want them and let them overlap.
9) Flatten the Image.
10) With the magic wand tool click inside the triangle and fill with a different color, Then Copy and File New< OK and paste. If you want a border around the triangle in a different color select the triangle with the magic wand tool and go to Select< Modify< Contract and type in the pixels, then fill with an other color.
|Rectangle with Rounded
Here is one way to do it:
1) Make a rectangle. Ctrl-d to deselect
2) Filter -> blur -> Gaussian Blur and make it so that the part that you want to be black is still
black, the rest is blurred.
3) Image -> adjust -> levels and then bring the sliders to the center.
|Rectangle with Rounded
Your best way to do this, for the best most flexible result, is to do it by brute force. First of all, think of what a rounded rectangle really IS. Think of this from the "building" perspective, not from the "taking-away" perspective. If, for example, you wanted a rounded-rectangle cake where the edges didnt have to be exactly precise, you would just make a rectangle cake and then cut the corners round with a knife. That is the "take-away" method, and that is how some people will tell you to do it. (Making a rectangle, blurring it, then upping the contrast till you get something that looks about right. It is not really precise, your result is not always really predictable, and you will end up with jaggy edges.
On the other hand, if you use a "building" method, you think of what photoshop can BUILD precisely and then you go from there. Photoshop is great with rectangles and with circles. Can you make a rounded corner rectangle from smaller rectangles and circles? Sure. Here's what you do:
1. Open your file.
2. Drag guidelines out to where you want the straight edges of your rounded rectangle to be.
||3. Drag more guidelines to outline where you want the curve to start. If you want an oval that is not a circle, make 2 guidelines to show where the curve is going to start on each side. Where they intersect will be the center of the circle or oval that represents the curve.
4. With your elliptical marquee tool, beginning at the inside-the-rectangle intersection of guidelines (read that again) hold the alt key and drag your ellipse up till it becomes your curve. (Holding alt makes your ellipse start build out from the center. If you want a perfect circle, hold your shift key as you drag.
5. With your foreground color selected, alt-backspace to fill the selection.
6. Drag your selection to the next corner and do the same.. Alt-backspace to fill. Continue with the other 2 corners.
7. You need to do two rectangles to finish this. Look at your picture and try to figure out where they would be. (Okay, if you need to know, the vertices of the two rectangles are the points of tangency of the ovals with the outer perimeter of the rectangle. Which, come to think of it, could be just as easily, if not more so, selected with the polygonal lasso tool. hmm.. )
Screen Capture When you select file -> new, the default new image size, will be the size of the image in memory (the screen shot). It's very convenient... --Jack Garvey
You can even use Ctrl+Alt+N (Alt + File » New...) to override the default size.
1. press the Print Screen key (to capture the entire screen, or
Screen to capture the active window)
>2. in PS, choose File » New... [Ctrl+N]
3. choose Edit » Paste [Ctrl+V]
|Screen capture on a Mac
Here are some keyboard commands for capturing screens.
1. Apple + Shift + 4 Position crosshairs, click and drag as you would
the marquee tool to select and capture the area you want to save. Release
2. Apple + Shift + 3 takes a picture of the whole screen.
3. Apple + Shift + Caps Lock + 4 takes a picture of the top window.
Position round icon and click.
||You will hear a camera shutter sound letting you know the capture has been
made and the picture should be found on your startup drive under the names
Picture 1, Picture 2, etc. You can open these from PS.
If you wish to cancel variations 1 or 3 before capture is made, press Command + period .
If the contrast is
significant between the object and its background, the magnetic lasso
tool is probably your best bet. Just click it near/on the border of the
image and drag it around the image kind of close to the image.
You can also use a "snap to" lasso if the contrast is great. Just use the lasso to draw a loose line around the object and then choose the magic wand and hit alt-click
This will bring the ants right up to the border of the object. If the object is
not in sharp enough color contrast with the background, then you will need to
use the polygonal lasso tool, probably. Magic wand can isolate colors for you.
(Shift-wand adds to your selection and alt- wand subtracts.)
Q: If I am using the magnetic lasso tool and need
to scroll to get to a part of the image currently not available, how do
I do that without messing up my selection process?
If you hold down the Option (PC: Alt) key, you can
release the mouse button without the selection snapping
shut. If you're using the MAGNETIC lasso, then you would
simply access the hand tool with the space bar and move the image. If,
on the other hand, you are using the STANDARD lasso tool, then you would
need to first hold the Opt/Alt key (to access the polygonal lasso) then
press the space bar to move the image. --TJ
Here is what i'd do for the save-the-layer-mask problem. As with anything in PS, there are lots of ways to save layer masks for later use on the same file or for use on another file (of the same size). Here is what you do.
1. Make your layer mask.
2. Ctrl-click in the mask in the layers palette to select the masked area. (The marching ants may look weird if you did it with a gradient or an airbrush.. Don't worry about that.)
3. Click the tab for your channels palette and then click the "Make Selection into Mask" button (next to the new channel icon.. little circle inside a square). This makes your selection into a channel of its own.
||To use the mask/alpha channel for a different image here's what you do:
1. Open both of your files, the one with your alpha channel you want to use (source), and the one where you want to apply it (target).
2. Ctrl-click the source alpha channel.
3. Click the Channels tab on your target file and then click the new channel icon next to the trash can there. Ctrl-v. This copies the selection to your alpha channel in the new file!
4. Click the layers tab and click the layer where you want to apply the mask. Click the "Add layer mask" icon at the bottom of the layers palette. Et voila`! What fun, eh?
Let me know if you run into any problems with this. I wrote this out pretty fast. Good luck!
Always me, Janee
Fading two images together Q: I've forgotten how to do the old fade off an image / two images together trick. A: Place one image on one layer and another image on another. Create a layer mask on the top image and paint black to white color in it. You can use the gradient tool in the layer mask with a black to white fade set. The layer mask will reveal or hide portions of the underlying layer depending on the white/black values in the layer mask. And the changes are not permanent to either one of your images.
|Or you can simply paint away parts of your top
image using the eraser tool with different brush settings. Or you can work on a
single layer with one image on the background. Create an alpha
channel and paint your black to white fade in it. Create a selection from that
alpha channel and hit apply image and choose the second image as your target for
the apply image effect.
|Fading two images together
(a short tutorial)
For your fadeaway, here's what you do: 1. Doubleclick the building layer in your layers palette to make it a regular layer. 2. Click the new layer icon next to the trashcan. This will be your new background, so drag it beneath the building in the layers palette. Fill it with whatever you want. 3. Select the building layer in your layers palette again and then look to the bottom of that palette.. the little icons... Click the one with the dotted circle. This is your add layer mask icon. 4. Notice in your layers palette that your building layer thumbnail picture has next to it a plain blank square. Ok.. this is going to be the "window" through which you will make your building peek. It is called a layer mask. Click it.
|5. Ok.. now this is not really part of the procedure, but i want you to do it anyway because i want you to see what you are doing.
a. Fill the layer mask with black. Your building appears to have disappeared, but if you look in your layers palette, you still see it in your thumbnail.
b. Now with the layer mask still clicked in your layers palette, choose a sizeable paintbrush and draw in white over the canvas.. Just scribble. You are drawing on your layer mask. See what it does?
c. Ok.. so drawing in white on the black layer mask lets part of your building peek through. Think about how to make the top peek through but the bottom be hidden by the background.
6. Ok .. back to the directions, in case you need them at this point. With black and white as foreground and background colors, choose the linear gradient tool.
7. Make a gradient from top to bottom. If it turns out that the wrong side of the building is peeking out, then just do the gradient the opposite direction.
MAKING ONE PICTURE "FADE INTO" ANOTHER USING LAYER MASKS
Q: I am having trouble merging a couple of jpegs together and need a little help. I have two pics identical in size etc., and I am trying to lay them side by side - horizontal - and want the right edge of the left pic to fade onto the left edge of the right pic. The left pic is a sunset view and a little darker and the right pic is a blue sky type view and much lighter, so I need a gradual fade to avoid a straight line joint between them. I have tried all sorts of stuff but cannot get it right. Any ideas?
||A: Put your pics on separate layers.
They should overlap a bit.
Add a layer mask to the topmost pic
With the rectangular marquis make a selection of the overlapping horizontal gradient (foreground to background - white to black) on the mask
--Greetings from Belgium, Ronald
Creating realistic blood
 Take the texture you want to apply blood to
 Create a new layer and change its mode to 'Colour Burn'
 Take a nice red colour and airbrush away =)
 To finish off, take a darker red and airbrush darker stains in the
center of the main blood stains.
I used two layers of yellow text, but instead of an inner bevel, I used a centered inner glow of black on the top layer.
the bottom layer, I used a large soft spray tool set to dissolve. Then I
adjusted the opacity of the top layer to let the "twinkles" show
Create a new layer
Set the colors to default black & white.
Select the Foreground To Background Gradient Tool.
Drag the Gradient from the left edge to right edge.
Filter > Render > Difference Clouds.
Image > Adjust > Invert.
Image > Adjust > Levels.
Slide the middle triangle under the histogram to the far right
until the lightning effect appears.
Change the Layer Mode to Hard Light.
If you want you can do the following to add some color to the
Image > Adjust > Hue/Saturation
--Bruce William Johnson
|Explosion No.1: The
1. Make a new layer with a transparent background. Choose the airbrush in 'dissolve' mode. Choose a nice fiery orange for the foreground colour and a suitable yellow for the background colour. This makes it easy to alternate between the two as necessary by pressing 'X'.
2. Make a nice orange ball with the airbrush (not too dense, better to 'squirt' a few pixels at a time by tapping the mouse button rather than hold the button in) then swap colours and add a few yellow highlights, then apply a light Gaussian blur. Repeat until you are satisfied with the fireball. The 'secret' is to build it up gradually. Billowing clouds or smoke can be created in a similer way if you choose suitable shades of colours to work with.
|Explosion No. 2: The Starburst
1. Make a new layer with a transparent background. Choose fiery orange as above.
2. Use a large soft edged brush to make an orange blob, then use a smaller (about half or two thirds the size) soft edged brush to put a yellow blob in the middle of the orange one.
3. Then just use the smudge tool to drag out your spikes of flame from the center.
4. Copy onto a new layer and rotate the copy to double the number of spikes. --Ron
1. Make a new canvas
work space with a transparent background that is one pixel wide and two pixels
2. Enlarge it as far as it will go 1600%.
3. Using the pencil add one pixel of color to the top leaving the bottom pixel empty.
4. Select all.
5. Click on edit, click on define pattern.
|6. Make a new file the size
you want it to be with a transparent background.
7. While this new file is selected go back to the edit screen and click on fill.
8. Choose Contents, use: pattern. Your second canvas is filled with lines and spaces. Add a background layer in another color. Or make your pattern two or more pixels high and use two or three different colors. It is always good to keep the background separate so that you can easily adjust the colors of the lines. --Dorothy
| CIRCULAR TEXT
1. Create a new doc in Photoshop and make it square... say 6X6 inches.
2. Drag guides horizontally and vertically to create a crosshair in the
center of the square.
3. Select the circle/oval marquee tool and position it in the center of the
crosshairs and hold down shift/alt (windows) while dragging the circle
open. This should create the circle from the center point.
4. Create a new layer while the circle is selected and stroke . A pixel
width of 1 should be fine. You are just creating a guide.
5. Now select the text tool and create one letter of the text you wish to
When sized to you satisfaction, position it at the outside, center top of
the circle guide you created.
||6. Go to edit>transform>rotate.
7. A box will now surround your text and in this box's center, is a pivot
point. Drag the pivot point to the center of the crosshair guides you
8. Now pull your letter around the outside of the circle to where you think
it should be positioned.
9. Do the same for each subsequent letter.
I don't know if you can afford Adobe Illustrator (we ALL know how
expensive Adobe products are) but I do all text bending in Illustrator
I then save as a JPG or TIFF, and layer it on my Photoshop image.
--Gary L. Brock
Text in a circle or oval.
This question comes up frequently in this newsgroup.
My solution in Photoshop is simple, even in version 3.04.
1. Type text
2. Scale-up the selected text vertically, about two times, with Free-Scale.
3. Rotate text 180 degrees.
4. Fit text symmetrically in a SQUARE selection.
5. Use Filter-Distort- Polarcoord. and choose Rectangular to Polar.
6. Rotate 180 degrees.
7. Adjust the dimension of the circle with Free-Scale.
a. Text rotates Clockwise. Skip 180 degrees rotation to obtain
Anti-Clockwise text rotation.
b. A RECTANGULAR selection gives text in an oval or ellipse.
c. A larger SQUARE selection "opens" the text circle.
d. A wider RECTANGULAR selection "opens" the text oval.
on Circular Text
I have found that this is best done with the polar coordinates filter, not
spherize filter. To experiment and learn how and where to place the text to
the result, type a page full of lines, then choose which one you like; this
will tell you where to place a single line of text to get the right
Vertical and horizontal placement determine the wrapping effect. You may have
to turn your text upside down to get the right effect - I have had to do this
before. One thing that helps is to make a grid page of lines and margins;
can later be used once you determine where you want to place the text.
More on Circular Text
If you are using a PC type computer and have Microsoft Word installed,
you can use the Wordart object. Go into Word and select insert
object. From the object list select Microsoft Wordart. This will
allow you to do text around a circle or arc.
For Janee's detailed tutorial on one way to make this effect see THIS LINK.
Step one: Open background image. Here we used 400 X 250 pixels at 72dpi. Step two: Create a new layer. Press D to set default foreground and background colors. Filter > Render > Clouds. Step three: Filter > Stylize > Find edges. Step four: Now adjust the levels for more definition in the image: Image > Adjust > Levels (Cntl L) Step five: To create the drops: Filter > Sketch > Plaster. You will have to experiment to find the best settings for your image size and intensity of the effect, here we used: Balance 40, Smooth 14, Light - Top Left. Step six: The result should resemble this. Step seven: To define the edges a bit more: Filter > Sharpen > Unsharp Mask: Amount 500, Radius 1.0, Threshold 0. Step eight: Select the magic-wand tool, Check the contiguous box on the status bar if using Photoshop 6.0, otherwise it is in the options palette. Set the appropriate tolerance setting, we used 10. Now select the background around the drops.
Hit the del key to erase the unwanted background. Invert the selection to the
drops. Select > Inverse or Cntrl + Shift + I. You should now see the original
image and the drops. If you wanted the image to appear as if mercury had been
spilled on it you can stop right here. For the water - proceed...
To make the drops look more like water change the layer blending mode to soft
light. We now have an image that looks like water drops on it. But let's have
some fun and give it that lens-type distortion typical to real water drops.
With the drops still selected, click the background layer on the layers palette
activate the background layer. 12/ Make the distortion:
Filter > Distort > Spherize. 35% (experiment to find what best works for you.)
You now have an image that looks like it was left out in the rain with realistic
a CD cover (Distorting a
square to make it round)
Q: I want to take a CD front cover and turn it into a disk label so that all the original picture is in the round.
A: Here's a way to get a comical distortion of a square picture to a round one
with a hole in the center, as for a CD label:
1) Use Image Size with resample turned on to make your image square if necessary.
2) Use Filter>Distort>Polar Coordinates, with polar to rectangular selected
to distort you image temporarily to something truly obscure
Use Canvas size to add a strip of white to the *top* of your image. This
will eventually be the hole in the center of your CD label. The strip
should be about 20 percent of the original image height.
4) Use the Polar Coordinates filter to convert from rectangular to polar,
and voila!, you're picture is recognizable, but with the outer edges drawn
into a sphere, and the center distorted to make room for the hole
5) Use ctrl-A (command-A on Mac) to select all, then ctrl-T to transform,
and reshape your image back into a square shape.
6) Fine tune the final image using Filter>Distort>Spherize - you can adjust
the size of the hole this way too.
the Displacement Filter
used the Displacement filter thousands of times with the few maps that come with PS, but I've never really understood how it works. I've
heard that you can create custom maps that allow you to 'bend' an image to
fit a shape, useful in creating good reflections or following contours, but I
don't understand how it works. Does anyone have a good
explanation of this filter?
A: It's fairly simple to understand, but more difficult to visualize what it's going to do. Displace copies neighboring pixel values based on grayscale values of corresponding pixels in a displacement map. Okay, not -that- easy to understand. :)
A 50% gray pixel in the displacement map is neutral and any corresponding pixel (in the same x,y position in the image being displaced) will not change value when the dmap is used. Darker pixels in the dmap will cause the corresponding pixels in the image to take on a value from pixels above and/or left, making it appear to move (actually copy) pixels down and/or right. Black causes the most apparent movement, which is determined by the horizontal and vertical percentage values chosen in the Displace dialog.
Lighter pixels in the dmap will cause the corresponding pixels in the
image to take on values from pixels below and/or right, making it
appear to move (copy) pixels up and/or left. White causes the most
apparent movement, which is determined by the horizontal and vertical
percentage values chosen in the Displace dialog.
Basically, I think of dmap pixels as pulling in values from neighboring pixels. If you made a dmap of a single white pixel on 50% gray, and used both horizontal and vertical displacement, the same pixel in the image being displaced would acquire a value from a neighboring pixel down and to the right. How far away the neighboring pixel is, is dependent on the percentage values you use.
Does that make sense? If not, try a few experiments with simple dmaps and images with which you can easily determine what has changed, and correlate with the descriptions above. Remember that dmaps must be saved as flattened .psd images. -- Ross
Photos -- Resizing
Q: My digital camera takes pics at 72dpi (jpg). What is the correct way to
resize the pictures? I want to make them smaller and 300dpi. I'm confused on the resampling, etc. area. Thanks for any help!
A: Turn off resampling in the Image Size dialog, and change the Print
Size. You can change the resolution or dimensions. The larger the print size, the lower the resolution - the higher the resolution, the
smaller the print size. There's a limit to how large you can make the print, and still maintain a decent resolution. That depends on how
many pixels you got from your camera - the more the better. An ideal resolution for your printer is dependent on your printer.
||By turning off resampling, you keep the same number of pixels - you
only change the size they're printed. If you resample, PS will throw out old pixels, or make up new pixels using interpolation.
Interpolation softens or distorts the image, and should only be used in certain instances. Resampling down (eliminating pixels) is okay -
you can use Unsharp Mask to restore edge contrast lost. Resampling up (making more pixels) won't give you higher resolution, just larger
and/or blurrier looking pixels.
Q: What is the best way to reduce file size? record of what level of compression was used or is it meaningful to save them as tiffs to protect against further degradation for my originals? A: Make the picture physically smaller. Reduce resolution to 72ppi and resample to smaller physical dimensions. That will hold onto the detail while decreasing size and load time.
You can also crop to a small portion of the picture. If you want people to be able to see a larger image, use the smaller image as a link and warn them of the file size of the larger one. Jpegs do lose data everytime you open and make CHANGES and then RESAVE. Just opening them to look at them does not cause loss.
To avoid this, I suggest the following. I shoot professionally and this works well. 1. download/scan photos into folders in an Originals file. 2. To work on a photo, get it up from the Originals photo and then go to File > Save a Copy As and name it and save as a .psd. This is the native photoshop file and is lossless. You can save it to the same file folder or a new one. I usually keep them in the same one so I'll know what I've got a working copy of. This also saves layers. 3. To print or move it to the web, save as a .jpg, either using File>Save As or the Export function. (you can print from .psd but all my printers seem to be happier with .jpgs). --LauraK
Reducing the number of colors before saving as a JPG won't make the file smaller. Reducing the level of detail in the picture will. For instance, a person standing against a background of hedges and trees; much detail is wasted drawing the greenery. If you blur that background, or shoot the photo so there's more blue sky and less greenery, the file will be much smaller. You'll definitely want to save TIFF, BMP, or PSD masters of the files you save as JPGs or any other lossy compression system. --Art I would like to remind you that you should never download the pictures of your digital camera by dragging the camera folder to another folder you create. (that is, when the pictures are downloaded with the default names and formats saved on your camera.) To download a picture or group of pictures from digital camera without a great lossiness proceed as follows: 1- Activate Photoshop 2- From File menu choose Import > Acquire Twain The camera's file manager will be activated and you choose pictures you like. You will notice that the pictures are positioned within (Untitled-1, Untitled-2, etc. windows). Now you have the peace of mind that the picture was downloaded in it's best format. Be sure that you have proper resolution before you begin tampering with the photo. --m_dabbagh
Q: My Web site looks ok in Internet Explorer, but when I look at
it in AOL, it doesn't look right.
A: When you browse a web site using the America Online browser, by default, AOL converts the images to its proprietary .ART format. The .ART format is very heavily compressed, and this conversion *substantially* degrades the quality of the image, making it look blurry. In some cases, it degrades Web site backgrounds so much that they turn into solid color.
can disable this. To turn it off, use Preferences->My AOL. On the
Mac, click the WWW button and uncheck
"Use Compressed Images." On the PC, click WWW
and select the "Graphics" tab, and turn off "Use
Now the Web sites you visit will look like the creator intended. Pictures will be sharper and clearer, and when you save pictures from the Web browser, they will save in the original GIF or JPEG format, not ART.
| IMPROVING PHOTOS
Improving the quality of photos
How can I improve the quality of pictures with bad quality, using Photoshop?
Oh my goodness, let me count the ways! Well, to cut to the chase, there are
some basic things to do that almost ALWAYS improve the quality.
First of all (if you scan the photo), no matter what you intend to do with
photo, scan it at 300 dpi to pick up all the information there.
Then, whether the photo was scanned or was scanned or brought in to Photoshop
some other way, go to Image > Adjust > Levels. Find the three little
on the middle bar and, usually, just moving the one on the extreme right to
left some will result in immediate improvement. Experiment with moving the
middle one a little to the right, etc.
Hit OK, then to go Edit > Undo,
then hit Redo and compare.
If that's not good enough, go to Image > Adjust > Curves, and you can
what happens to the photo when you play with the 45 degree angle line.
And don't forget to sharpen. But don't use a sharpen filter; use the Unsharp
Mask under Filters > Sharpen. Set the boxes from top to bottom to about
100, 3 and 2 and see what that does. Experiment.
There's probably not a photo that can't be improved either somewhat or a
lot using these three procedures, and of course there are many, many more
things you can do.
|On colouring grayscale
First, make sure to convert your grayscale image to a color mode if needed - RGB, or CMYK.
The easiest way to color a grayscale image is to use the "Color" blending mode. You can use any of the paint tools in Color mode, and
paint right over the grayscale pic. The colors will take on the brightness of the grays underneath.
Better yet, make a new blank layer above your gray pic, put the layer in Color mode ( in the Layers palette), and then you can paint (in
Normal mode) on the new layer. You can work faster because erasing color is as easy as applying it.
Another method that gives you more control is to make a selection of something you want to make a particular color, then create an
Adjustment layer (Ctrl-click the "Create new layer" button in the Layers palette) using Hue/Saturation. Enable the "Colorize" option in
the Hue/Sat dialog, and adjust the sliders to get the color you want. You can go back and change the settings at any time so you're not
committed. You can make a separate Adjustment layer for each different area you color.
TIP: Choose the color you want before using Hue/Sat. When you enable "Colorize", that foreground color will be used.
Notice that when you make an Adjustment layer with a selection active, you make a mask for it. Look at the thumbnail in the Layers palette
for the Adjustment layer. See that white indicates where the adjustment is applied, and black where it's not. You can paint, fill,
filter, use a gradient, or anything you need to change the mask. Gray
values on the mask make a partial mask, meaning you can apply the adjustment partially wherever you want. For example, say you had a
checkered tablecloth that you'd colorized with red. You could choose one of the values from the grayscale image (say, the lighter squares
of the pattern), then fill the mask with gray to lessen the adjustment, making the lighter squares less saturated than the darker
Selecting areas before colorizing them with an Adjustment layer is more work up front, but it allows you a lot of flexibility. Making a
colorization less flat looking is easier when you can alter the mask that applies the adjustment. Making selections based on highlights or
shadows, for example, allows you to lower saturation in the highlights, or make a cooler hue in the shadows. You can add noise or
mottling (Clouds filter) to a mask so that colors are not so uniform.
You can darken/lighten areas to create more contrast in the grayscale pic. You can paint on the mask to remove color from spots (eyes,
specular highlights), soften edges, etc...
|Solution for skin blemishes
1: Convert the image to LAB color space.
2: In the Channel menu, select the "L" channel.
3: Use the clone and blur tools to remove any "dark" blemishes on the L channel.
4: (If you want to sharpen, doing it only to the L channel is the BEST way to sharpen by far.)
5: In the Channel menu, select the "A" or "B" channels
6: Use the clone and blur tools to remove any "color"patchiness.
7: Convert back to RGB or whatever.
|The "L" channel is the "Lightness", and does not contain any color
information. The "A" and "B" channels ONLY contain color information, not
lightness/darkness. So, if it's a color problem, use A/B, if it's density, use the L. This way, you're not applying corrections to places that don't
need them, and it preserves the other elements perfectly.
Try it for this type of thing and sharpening... it works wonders.
Adobe Certified Instructor For Photoshop
1. Make a copy of the layer with the picture. 2. Apply enough Gaussian blur to get rid of any imperfections. The picture won't look good; don't worry. 3. Drag your duplicate layer above this one in the layers palette. 4. With a gentle hand and a feathered brush, "erase" away the problem areas.
If you need to, you can also use the clone tool with a part of the person's face that is RIGHT next to the offending part. -- Janee
|For additional "glamour shot" effects you can also...
Apply filter>distort>diffuse glow to that second upper blurred layer, then reduce it's
opacity with the luminosity pull-down selected. Additionally you might want to apply an
extra dose of filter>sharpen>unsharp mask or standard sharpen to the first layer to
conterbalance the blurred layer.
You'll have to play around with the diffuse glow sliders for your preferred effect. The
final effect should give your subject a smooth radiant soft focus feel, depending on the
amount that you dial in. You can use it a little or a lot.
This procedure appears in several publications and I have used it with good results.