Before I get started, there’s a different between Photoshop, with a capital “P”, which is an expensive image manipulation product sold by Adobe, versus photoshop, with a lower case ‘p’ which is a verb meaning ‘leveraging image manipulation software to alter a digital image from it’s original form’.
Here’s an example of a guy’s face with pretty extreme ‘photoshopping’ done to it to remove his acne:
And here’s a pretty extreme image of Britney Spears appearing on the cover of Marie Claire:
Makes you wonder, how old is that girl anyways, doesn’t it?
Now for this lesson, I am going to demonstrate how to take two images and merge them together to make one seamless scene.
Here’s an example of an image I created a while ago for a short story I wrote named “My Sandbox” which was about a time traveling baby who created helped the ancient Egyptian create the Pyramids:
What I am going to do today is take this image of Neil Armstrong on the moon:
And place this image of me in it:
Now being honest, this is a horrible image to do this with because the color, lighting and background colors are all off, which will wind up making this more difficult than it should be to crop and light by hand. And while this is all the homeless man who has no camera can afford, it also provides a pretty good lesson.
For photo editing, I leverage something called Paintshop Pro 3.0. It’s a tad old school – but has all the things we need to do the basics.
So first, I load up the Space Image. This will be the primary image I use.
Paintshop Pro allows me to open up multiple images at the same time so next, I open up the image of me I am going to crop, and zoom in on myself.
The first thing I need to do is get a clean copy of JUST me without the background and surrounding area.
So I FIRST select the entire photo, then I deselect a rectangle around me. Here’s what that looks like:
The region that’s selected will be cropped from this photo, so I will be left with only myself and no background.
Next, I change the selection tool to ‘Magic Wand’.
The Magic Wand Settings come up on the menu, which have these settings:
What this allows me to do is to select entire regions that matching the RGB value of a color I click on out of the photo. For instance, if I leverage these settings and then hold down the SHIFT key on the beige tile to add to the selected region, I get this:
The selection now contours my pants nicely on both sides.
The goal is to make sure no part of me is selected and to only select ‘background’ I want omitted from the scene.
Repeating this a few times, adjusting the tolerance level a bit here and there to not select as much region, and leveraging the undo function A LOT on the edit menu by pressing Control+Z if I mess up and select more into the region than I wanted, I then get an image that looks like this:
As you can see, it’s not a perfect selection, there’s a few places where my hair’s selected and shouldn’t be, and a big part of the complicated scenery on the left of my shoulder is not selected as well as my hand which should be.
To finish the selection. I go to the selection tool, this time selecting ‘Freehand Selection’.
With this, I use the Control Key AND the mousewheel to ZOOM IN and OUT of the areas I am working on, for fine control of the selection. I want to be sure to SELECT ONLY the BACKGROUND that I will crop out of the photo, leaving only me.
Pressing the SHIFT key while I start the selection on the large region on my upper left shoulder, I trace the entire region…
Which leaves me with something that looks like this :
And then, zooming in further on my left shoulder, repeating this…
This is absolutely most tedious part of doing photoshop work, is setting up the crop selection.
I have a habit of screwing up my selection or scrolling to the left and right involuntarily, so make the CTRL+Z key your best friend.
It will help your selection process.
So once I have washed, rinsed, and repeated for the entire image, I wind up with this selection:
Now mind you, this selection is inverted. Meaning I have selected the background and not myself.
So what i do is ‘invert the selection’ by selecting the menu item ‘Selections’ and the “Invert” option.
AND THEN I copy that selection to the clipboard using CONTROL+C
And then I switch to the space image using Control + TAB
And from the menu I select “EDIT” and “Paste as New Layer”
Which then insert me into the space image like this:
OOPS! Too big, right?
That’s easy to fix. I now select the ‘arrow’ icon on the left side, and make sure the top drop down box says ‘Scaling’, and then I simply select the upper left hand corner of the image to change the images size, and if I select the image or use the arrow keys on my keyboard, i can then move that image around.
After I have scaled the pasted image’s size and move it’s position, I am now left with this image:
But there’s a couple problems that make this photo not believable, right? I mean, other than the obvious I’m on the moon without a suit.
The first problem is the light source. If you look at the image of me compared to that of the astronaut, you should clearly see that the light source is to my and the astronaut’s left. However, I am uniformly lit. So what I need to do is create a light source to create some artificial imbalance in my image.
On the ‘Effects’ menu is “Illumination Effects” – and wouldn’t ya know it, there’s a “Lights….” selection there that I make.
From there, It really is a matter of tweaking the image for the best visual appeal.
When the light effects dialog pops up, I move it off to the side and select ‘preview on image’ to allow me to see what I am adjusting. Looking at the astronaut, the light source is on the left, so I place the light source off to the left, I make it a small cone, and bump the intensity up a bit on it. Basically I ‘play’ with the settings until i get a similar look to the astronaut’s.
Here’s where I landed on for my settings and how it looks :
Now that I have good lighting, there’s a problem with the vividness of me compared to the rest of the image.
This is pretty easy to fix through the sharpen function of my Photo editing software, I do this in Paintshop pro by selecting ‘Adjust’ and ‘Sharpen’, and repeat this a couple times. This leaves me looking a lot ‘sharper’
Another coloring problem is the brilliance of the strong colors on the image and the rather drab colors on me. This is indicative of the quality of the camera taking the imagery. Since it’s only in need of a touch up, all I need to adjust is the brightness and contrast of my image.
Specifically, I want to bump up the contrast. What this means to me is – I want to make the colors ‘stand out’ a bit more. Now as I adjust the contrast this typically erodes the brightness, which is fine, I counterbalance – just a bit, by decreasing the brightness. As always, I select ‘Preview on Image’ to make it so I can ‘guess less’ on the results.
One last effect I want to do is add a subtle blur effect to match the blur on the slight blur of the flag.
There are several blur options in Paintshop Pro, but they blur FAR too much, so I land on a gaussian blur, which isn’t exactly what I want, but it will work fine enough with how little I am applying it.
Now the final problem – which gets emphasized by the color changes – especially the sharpening – is the blending of the digital edges of my image and the rest of the photo. It just flat out don’t look natural. My image doesn’t quite fit in with the surrounding area because of the digital artifacts – aka jaggies.
Here’s a closeup demonstrating the problem:
To the left you see me and to the right you see the flag. Do you see the harsh and choppy edges around me and the lack of harsh edges around the flag?
Now I can ONLY start the following process of blending the imagery AFTER I have done all the prior effects, so I do a quick inventory of the image and make sure I’m happy with it. It’s decent. Not totally professional. but decent.
To fix the jaggies, I have to use something called a smudge brush.
But before I do this. I now merge the two layers.
And on the right hand side, I now see only one layer that I have to work with:
Now that I have a single image/layer to work with, I then want to LIGHTLY smudge the colors from my to the image behind me and vice versa.
So I select the smudge brush…
I then make the brush REAL small (4 pixels) – and methodically, slowly and LIGHTLY smudge the entire outline of where I had previously cropped. If you use too high density a smudge, it looks obvious. If you use too large a smudge brush, you can’t get the detail of the background lines. And a ROUND BRUSH ALWAYS looks more natural for smudging than a rectangular one especially if you’re trying to prevent jaggues from forming.
Here’s my settings:
And then I methodically trace the smudge brush along the contours of the image
Now in looking at what I thought was a finished image, I forgot something major.. MY SHADOW!
Kids, remember, when you’re doing light sources, you ALWAYS have a shadow.
They are usually easier to do at the same time!
So I copy a portion of the shadow of the lander behind us and paste that shadow under my feet, and then scale it according to the distance I am from the camera like I did before. I also copy paste the flag pole to be IN FRONT of the shadow to add to the realism, and then I use the smudge brush, again, at my left foot to add a little shadow to it, and then around the flag pole to make it look absolutely real.
I dont need to demonstrate all that again, do I?
And…. THAT’s IT (finally). Here’s the final image of yours truly walking on the moon, shadow and all:
There ya go. Not perfect, by any measure, but it’s still fun, right?