More luminosity

I'll say it again: luminosity is the key to making good pictures; for those who don't know , PHOTOgraphy means "light-print".Photo means light (photon), when your eyes see an object , it doesn't see the object but the light it reflects, absorbs, and/or emits.The intensity of light , it's tint, falloff , difuse map are what make your brain recognize an object.Therefore, everything you do when you retouch a photo is to alter the caracteristics of its light; a good , simple example is when you enhance a landscape : you'll use curves, hue, etc, that all are commands that alter the luminosity.Since 99% of you work on RGB monitors, photo retouching consists in altering the value of each red, green and blue pixel in a picture, modyfying the light it emits.

I just wrote this part because most people don't think of that and therefore dont understand the way photoshop works, and wonder why I emphasize so much on luminosity.Now I'll share some luminosity correction tips so you can practice :

1: bad exposition

You just shot a pic , and the fact is that the exposition is bad: some parts are visible, some are too dark.uneven luminosity.Here's a pic of my nephew Lucas:

O bviously the background on the left is too dark, so let's try this :

D uplicate the layer, (ctrl+j), invert the luminosity of the duplicate (ctrl+i), put it in soft light mode, and give it a 9.0 gaussian blur:hooray the dark part is now visible !

The problem is that it also altered the luminosity of the kid, therefore just duplicate the soft light layer for safety, turn it off, and erase the part of the visible one that overlaps Lucas with a large soft brush to get rid of the luminosity settings the layer contains , and so you restore his original luminosity


You can see the shape of the layer in the layer palette: that's a way to correct bad exposition , and even a pic's luminosity in a non destructive way.

2: Low contrast

If you shoot a pic and it has a low contrast, it means that his alpha channel, or luminosity values are too modrate, low, even; in other words, the black point and light point are not visible enough: I lowered the contrast on the pic of Lucas to make an example


You have two very good ways to fix it:

A: create a luminosity map:

Duplicate the layer, desaturate it (ctrl+shift+U), put it in luminosity mode, and press ctrl+L


You can fix the saturation by adding a little, and it's done!


Just compare with pic 4 to see if you like it.

B:Jerry's trick:

Jerry717 taught me and a few others that were on the discussion board this trick; it's a very good example as it is efficient, easy, and practicing it will make you understand the whole luminosity thing.

Take pic#4, create a curves adjustment layer over it; double click the eyedropper on the right


And fill in the rgb values with "250" : this prevents the picture to contain a white point that would contain extreme, non printable color values.Press enter, and , again with the same eyedropper, click on the bright part of Lucas' t-shirt.

Now double click on the left eyedropper (set black point), and enter "3" as rgb values for r, g, and b.press enter and click in the darkest zone in the left side upper part of the pic .

Now do the same with the center eyedropper (set gray point), enter "128" as rgb values: this makes your pic have the right, neutral color values.then click the eyedropper where the arrow points at in pic 8


This sets the gray point of the picture , and makes it get rid of the opposite color that it contains.Try it everywhere on the pic to understand what I mean.

Now just press enter to get out of the curves layer, duplicate the layer and do this :put the adjustment layer in luminosity mode, duplicate it, and put the copy in color mode.This lets you tweak each different settings aprt from the others.Otherwise if you like the original result, just put your adjustment layer in normal mode...the result I have is this one

The great thing about jerry's trick is that it also prepares the pic for printing, since it implies good black point and light point values, both for the eye and the printer.Thank you Jerry !!!!

Photoshop CS added a "shadow/highlight" command that basically does the same, but it doesn't explain you how it works....I prefer the old way.

You could have tried the second trick "luminosity layer" with an overlay layer instead.....just try some more, using opposite values, deasturated layers so you understand precisely how this works.

Pen Tool as Mask Selector

My Happy Toaster

One of the main things that every PhotoChopper needs to know is how to free-up, or separate, different parts of an image. With programs like Photoshop there are a number of ways to do this. But this particular tutorial will deal with one of my favorite way to get this sometimes challenging job done. And the main tool for this tutorial will be:

THE PEN TOOL

Although there are a number of sub-tool-types for this item, we will be focusing on the basic Pen Tool which is accessed under the icon shown above on your PS tool bar. Choose it by clicking on the icon, or for bonus points use the keyboard shortcut by pressing "P" to select it.

Open a simple (160 x 120) pixel NEW IMAGE. Take the pen tool and click on it in two places: You'll note that you get a straight line connecting where you click (see A). If you keep clicking different places, you'll continue to get more straight lines connecting the dots.

You've just created a PATH. So far so boring. Let's get into the FUN part of the pen tool - Bezier Curves!

But first, clear your current 'work path' by going to your PATHS palette (find under Windows: PATHS if you can't find it already open somewhere). There should be a 'layer' labeled WORK PATH. Click on it and then click the TRASH CAN at the bottom of the path palette (or drag that path to the TRASH CAN). This clears the previous path you were working on.

Now create another NEW IMAGE. This time click once and then, when you do the following clicks HOLD DOWN THE MOUSE BUTTON (do not release). WOW, things start happening!

First thing you'll notice is that the connecting line curves, and depending on how you move the mouse around BEFORE RELEASING IT it curves differently. The second thing you'll notice is that little "knobs" or "antenna" sprout out of the line. The places you've clicked will have empty circles, and are called ANCHOR POINTS. The "knobs" will show up as smaller and black. These are called CONTROL POINTS, or DIRECTION POINTS and will normally show up on both sides of the ANCHOR points.

While the PEN TOOL is active you can change how the curve behaves in real time. If you (on PCs) hold down the CNTRL key you'll get a hollow white arrow. With this tool you can move the ANCHOR points around and reposition them, and also move the DIRECTION POINTS to adjust how the line curves. If you instead hold down the ALT key you'll get a little "V" arrow - with this tool you can change the angle of the CONTROL points for more drastic (and more accurate) curve behaviors.

In this case, to free the toaster!!

You can use the picture of any toaster to practice.

By using the PEN tool and carefully 'tracing' around the toaster I have created a PATH. The picture below will show you a closer detail of the PATH tracing.

As you can see, the path is not pixilated but remains flexible. With practice you learn how to "ride the edge of the pixels" to best advantage. If you make an error you can use the PATH adjustment tools to bend it to your will until you have your object perfectly surrounded!

Once you are happy with your path, you can click on the PATH OPTIONS arrow in the PATHS palette, and then choose SAVE PATH. This PATH will be saved with your PSD and you can still edit it to your heart's content at any time in the future. Or copy it and have a few "edits" of the path saved. A PATH is much more flexible for solid edged objects than a normal Selection.

With the PEN TOOL selected RIGHT CLICK (PCs) on the path you've just created. You'll get a pop-up selection box similar to the one show at left. Choose MAKE SELECTION.

The toaster is now selected exactly following your PATH. Depending on which version of PS you have, the PATH is deselected and the Selection remains. If the path is still there, go to the PATH palette and select the gray space UNDER the PATH layer to deselect it.

Then, you can COPY the selected area into a new document, or CNTRL+J (PCs) to have PS copy the selected area into a new layer above the original. The TOASTER IS NOW AN INDEPENDENT TOASTER.

Some of the benefits of using PATHS to create selections are:

• you can save and re-edit the PATH at any time, and it is saved as part of the PSD file.
  
• it creates a very clean 'cut' when forming a selection. Great for geometric shapes or objects with defined edges.
  
• a PATH can also be TRACED with another tool, a Brush or Pencil. Great for creating a complex outline of an object.
  
• once you master the PEN tools ANCHOR POINTS and DIRECTION POINTS (those knobby things!) you can create complex yet clean selections very quickly.

Cutting Up

Walking the Dog

Our gal walking the dog has just come from the beach. Well, actually she never was in that field above, but she did have a walk on the beach.... stand bye for more!

Taking things from one image and putting them into another image is one of the basic concepts that one needs to master if one is to continue down the path to Photoshop Guruhood. I'll go step by step through this process and perhaps you'll pick up a few tips along the way.

ORIGINAL PIC #1

ORIGINAL PIC #2

Get chopping!

By using the PEN tool and carefully 'tracing' around the girl and dog I created a path, which I then converted into a selection. ash. No telling what that dog would do if we didn't keep it on that leash!

We'll need our three-legged friend under separate control, so we'll chop him out now. Put a ruff lassoo around him cutting close where the lease attaches to his body.

Then press CNTRL+SHFT+J (PC) and he will be magically transported to a new separate layer.

(he really does only have three legs, another reason to keep that leash on him! Don't want him to loose another leg or he's in REAL trouble!)

So we've got a gal and a leash (be still my happy heart!) and a three legged dog. Let's get them off the beach.

Open up the original pic #2, the tree and the field. Here's a neat trick. Have the pic with the cut out gal and dog open on your desktop. And have the pic#2 also open on your desktop. Size them up so you can see both. MAKE SURE that you have the DOG layer and teh GIRL layer LINKED (click on the GIRL layer and then click in the gray empty box to the left of the DOG layer and a little "chain link" will appear).

Select the PICK tool and make sure that is checked. Click on the girl.
While holding the right mouse down drag the girl from her window into the tree/field pic and release. If you did it right the dog went obediently along with her, while maintaining his own layer! (you can use this method to drag multiple layers from pic to pic. You can also first put the layers in a folder as well, and then drag the folder from one pic to the other.)

okay, they are off the beach and in the field. One important point here is that I chose the field picture because it had a simular perspective to the beach picture. Also I made sure that the lighting was simular. Although you can play a lot with lighting and shadows in PS, it is always easier to pick well matched bits!

But it still looks a wee bit faked.

Firstly, she looks weird up there above the doggy. This is why I wanted to control Lucky (the three legged dog) separately.

make sure the dog layer is on TOP of the girl layer. Go to the LINK button and UN-LINK the girl and the dog (unleash them, so to speak!).

move the doggy UP the leash and towards the girl, then re-link them. Move them both into a more 'artisitaclly pleasing' position in the forground.

better, but not quite right yet. Let's see what be can do about their feet. And we also need to deal with the fact that they are casting NO shadows. Unless we are working on a vampire pic, this won't do!

zoom in on her feet. Take the lasso and do a ruff outline that roughly follows the shape of the grass and flowers around her feet. Don't worry about the bottom of the selection too much, the important part is the part that overlaps her shoes.

With the FIELD/TREE layer selected press CNTRL+J (no SHFT). This will magically put a copy of the selected area in a layer above the FIELD/TREE layer. Now take the copied layer and move it ABOVE the girl layer.

you'll see that her shoes look more "IN THE GRASS" now, but still a bit fake looking. Take the ERASER tool with about 40% opacitiy and a medium feathered edge and rub out and soften the sharp edges of the patched grass and flowers. You should get an effect simular to what is at left bottom.

You can get the same idea without cutting out extra grass but using the clone tool / rubber stamp tool. But I like this way as it leaves me a layer that I can change or modify later.

Do the same to the doggies three legs!

Okay, looking at the original beach picture we can see that there is some shadow action happening there. Analizing the original shadows we see the size and angle they need to be. Create a NEW LAYER above the FIELD/TREE layer but BELOW the dog and girl layer. I used the lassoo to draw a rough selection approximating the shadow shape. I then filled it with 100% black. Since the grass and flowers is a less flat surface than the sand, the shadow need to be softer. I used GUASSIAN BLUR and then put the opacity of the shadow layer to about 40%. The girl is done, the dog is still in the unedited shadow stage.

Almost Done! As it stands, not too bad. But being a photographer I know that there is one more thing we can do to make this an interesting picture. The composition is good with the dog and the girl and the tree forming a nice triangle flow. But that tree is just a wee bit too overbearing. I want the viewer to really track the girl and Lucky, and not notice the tree as much. So I'll apply some "depth of field" to the background.

Depth of field is a great way to control a busy background, and keep the viewers eyes on the forground. With a pro camera you set the lense in a way that the focus drops off behind the subject. Since my background was originally shot to be sharp all over we'll have to let Photoshop help us out in the post-production phase.

Hide the Girl/Dog/Shadows/grass layers for now. Select the FIELD/TREE layer and then using the lassoo tool draw a rough line along the field just in front of the tree. If you look close you'll see there is a natural line of the field grass where you can divide your picture. I've show the selection at left with the QUICK MASK mode on so you can see where I've selected better. For a even better effect fade the selection from the front (advance work!). Hit ALT+CNTRL+D to pick the selection feather and put in 1 or 2 pixels.

With the FIELD/TREE layer selected go to filters and pick Guassian Blur. I used about 1 pixel, but do it to your taste. Try to not go too crazy with this or it will look really fake. It is like a seasoning!

Okay. Turn on all the layers and you should have a happy three-legged dog taking his two-legged owner for a walk in a nice meadow.

Some things we used to make this image are:
   
1) sectioning selected layers for more control of image placement

2) direct dragging of multiple layers from one image window to another.

3) CNTRL+SHFT+J to cut a selection and automatically place it in a new layer. And CNTRL+J to copy a selection into a new layer above.

4) adding forground bits to 'blend' your cut-outs into the background

5) faking "depth of field" to add depth to images, and basic composition control.

Perspective

I'd like to talk about the importance of perspective - often the most forgotten ingredient in PSC
entries (with reflection and shadow realism a close second/third).

Have you ever looked at an entry and thought "something looks wrong here, but I can't quite put
my finger on it"? Well, 9 times out of 10, it's probably a perspective problem.

Sometimes you can cheat or "force the perspective" without it being an obvious problem, just as you
can cheat the realism of shadows/reflections to some degree. A recent contest example is illustrated by
this entry of Patre's, which did quite well:

Being a secure and well-adjusted newbie, Patre welcomed my comment and asked for clarification of this issue,
so here we are. The perspective problem here is caused by a conflict between the camera point-of-view (POV) in
the original background image versus the new source (foreground). One POV is higher than the other. The tell-tale
signs are the areas I've highlighted below plus what can be seen of the horse itself:

Note how in the foreground, the side rails on either side of the track (flashing blue) are vertically close together -
indicating that they are approaching alignment with the horizon. Also note that we can see little of the top of the
horse's body (we are not looking DOWN onto the horse at all) - indicating that the horizon runs through it somewhere.
These, and similar key pointers in the background image, allow us to project an imaginary horizon-line for each POV:

The mis-match between the 2 camera POV's (and their horizon-lines) is what causes the final image to look odd.
To remedy this, we can move the original background image down so that its horizon approaches the same level as
the new foreground image:

So... when choosing a compatible source image to combine with another, ask yourself "Does the camera have a similar
enough point-of-view in each image?"... and discover the joy of good perspective.

Depth of Field

Well, i knew the day would come when my photoshop skills would advance so much that somebody
would accuse me of cheating! The picture in question is this:

just to cover certain questions about my post, i have made a step-by-step guide as to how and why I did this.
The original entry took me nearly a whole day to complete, and as far as i've noticed, i can't see anything else
i could do to improve it. This step-by-step guide will cover just the basics of what i did as I have lost much of
the original material i used (mainly by flattening files and so on), and I only spent a short time covering the basics
- so for those of you who think the new attempt looks nothing like the original, then it won't because I haven't spent
any time perfecting it.

Before we begin, you will need a copy of this:

and this too:

(Jokes apart, the book is very good!)

First I created a background (anyone would have worked fine, but the stripes give the depth of field
more emphasis) and then blurred it using Filter - Blur - Gaussian Blur

then I opened the original image:

For the effect i wanted to create, I needed to work on the original image before it was to be pasted into the
main file. I wanted the image to look printed, with bleeding lines (as if the head wasn't aligned properly) and
I wanted the colours to be slightly different, and a bit more plain.

Above, I used 'Select - Colour Range' and selected the red in the picture with a fuzziness of 157. I then
created a new layer, filling the selection with a lighter red. I did the same for the green and then changed
the opacity of both of them to blend it in.
I then flattened the image and altered the Brightness/Contrast.

I then Copied the flattened image layer (Background) and selected the mode to 'Soft Light'

I then created the 'print lines' by roatating the canvas 90'CCW then applyed 'Filter - Sketch - Halftone Pattern'.
The pattern was set to 'Line' and the size =2, Contrast=3
I then rotated the canvas back to normal (90'CW) - i had to rotate the canvas as applying this filter without doing
so would have given me lines in the other direction!

I then flattened the image and increased the canvas size to give me a white border.

I then dragged the final image onto the background i had created:

A bit on depth of field.......

When a lens focuses on a subject at a distance, all subjects at that distance are sharply focused. Subjects
that are not at the same distance are out of focus and theoretically are not sharp. However, since human eyes
cannot distinguish very small degree of unsharpness, some subjects that are in front of and behind the sharply
focused subjects can still appear sharp. The zone of acceptable sharpness is referred to as the depth of field.
Thus, increasing the depth of field increases the sharpness of an image. We can use smaller apertures for
increasing the depth of field.

the pic below is taken with a small aperture rating and has a greater depth of field:

and this one has a higher aperture rating, a smallere depth of field:

I wanted to create a picture that looked realistic, so I wanted to use a greater depth of field. When you look
at the final pic i created, the front (left) and the back (right) are out of focus. This is because the focal point is
towards the centre.

To help me with the perspective, I drew a grid on the background with lines. I then distorted the grid to follow
the line of a 'hanging' line.

the top of the grid will represent my imaginary hanging line

I then set about distorting the paper to make it look as if it was at an angle to the viewer. I used 'Edit - Transform - Distort'
and aligned it to the grid:

After having copied the paper image five times, I then went about distorting them all, then scaling them
smaller and smaller to give the 'distance' more feel.

Curling the paper.........

This one took me ages to get right, and many attempts. To give the paper it's 'curled' look, I used the 'Filter - Distort - Shear'.

I then did this to the rest:

The only other source pic i used was one of some pegs:

Here, I quickly cut them out, but in the original effort, I obviously spent more time on them.
Using 'Edit - Transform - Distort' again, I added them to the pic. Using the eraser tool, I shaped the pegs to
make them look as if the paper was through them.
I then scaled them down more progressivly each time so they would match the paper size. I then drew the
line in with the shapes tool - 'line'

A bit about shadows......

I obviously spent a lot of time on the original completing the shadows. Implementing shadows into the picture
will help enhance distance, depth of field and lighting.

Take a look at the first picture. It depicts two spheres sitting on, or floating above a green base. It's hard to tell
whether they are touching the base or if they are floating above it.
Now look at the second picture. Suddenly it has all become clear. The red sphere is the larger of the two, and is
resting at the back of the green base. The blue one is floating above the base, and is nearer. The only thing that is
different between the two is that the second picture has shadows, otherwise they are identical.
As you can see, shadows provide the viewer with a great deal of information about a scene. An object with a shadow
has more presence than one without. Shadows give depth to a scene and reinforce our perception of the location of lightsources.
Take a look at Quake. It's shadows add mood and atmosphere. Imagine how boring it would look with everything at the same brightness.

I imagined a light source from the angle below:

I then - quickly - added some shadows:

Now for the Depth of field conversion!
I flattened the image and chose the blur tool. With reference to the depth of field, I chose a focal point
(just off centre, to the left)
and then blurred the far left hand side. I then started from the centre, blurring progressively more the further
right I went, until the distance was not in focus.