Smart objects allow you to apply non-destructive transformations, duplicate the objects multiple times and base all the duplicates on a single original. It also allows you to import an inside file, like am illustrator file, and manipulate it over and over again after you've placed it in as a smart object. Non-destructive transformations are a milestone topic in Photoshop development.
In the Layers palette I've got two layers, upper girl and lower girl, that correspond to the 2 objects in the picture. We're going to apply a standard destructive transformation to the upper girl and then convert the lower girl into a smart object and apply a non-destructive transformation to it.
So, with the upper girl layer selected, I'm going to press Ctrl+T to enter the free transform mode and scale the object down to a very tiny size like 33.3% (I'm pressing the shift button as I'm scaling it down) Now I'm going to accept that modification, and you'll know that we've just applied a destructive transformation, i.e. interpolated the object down to a very small size and reduced the number of pixels inside this girl layer. Now try and press Ctrl+T to once again enter the free transform mode, and let's try to make the objects as big or roughly as big as it was before. Shift-drag the corner handle and you'll see that it's not working out. You can see those big chunky pixels and of course when you press the Enter key, the image becomes all fuzzy. This is a destructive transformation and this is why the rule of thumb inside the Photoshop has been for years and years: you apply one transformation pass, you get all your scaling and rotating done, and that's that. Well, no more. Inside of Photoshop CS2 you no longer have to worry about that as long as you take some time to first convert the layer to a smart object. I'll show you how it works.
Let's go down to the layers palette here, click the lower girl top select the layer, then right-click to the right of the layer girl name and choose "Group into new smart object". As soon as you do that, you get the special little icon next to the thumbnail, which is telling you that this is now a separate file that is saved inside of your original file (not as a separate file on disc, but as a separate image inside the same file)
If you prefer to get to that command from the menu, go to the Layers palette>Smart objects>Group into new smart object.
I'm going to press Ctrl+T again, scale this guy down to a very a teeny-weeny size and press the Enter key to accept that transformation. It is very-very dinky, it looks like it's been interpolated down and has lost all those pixels. But try pressing Ctrl+T again and look up here inside the options bar.
Keep in mind that Photoshop automatically saves my transformation settings, because these settings are applied parametrically, that is to say non-destructively, and I could change these settings if I wanted to by turning on the little chain icon, so that I'm changing my settings proportionately. So, if you raise it to 50% or larger (up to 100%), it actually makes the image bigger and while keeping it nice and smooth looking. You probably don't want to make the object any bigger than your original one, because Smart Objects don't magically produce pixels, they just keep your original pixels intact. When you press Ctrl key, you're going to get a sharp smooth object – the result of a truly non-destructive transformation, the kind of thing you've never seen in Photoshop before.
Delete the layer with the right girl, then duplicate another one and place it to the right. Let's try making a modification to both of these girls. Let's say I want these layers to appear de-saturated. For this, we'll have to de-saturate the objects. If these were separate layers, we'd have to apply a layers style to each and every one of them. But because they are all linked to a single original, the process is simplified.
In the layers palette, double-click on a layer's thumbnail (first copy, for instance) in order to open up the original version of the image that is saved as part of this file. This allows you to make adjustments that are then represented in the original background composition. You can also make adjustments just to the smart object if you want. You can create as many layers as you can inside of a normal Photoshop file.
To apply an effect to this girl, I'm going to have to de-saturate the photo. This will give me a wonderful black & white effect here. I will click OK to accept that modification. Now the great thing about Smart Objects is that you can make the change once and it's updated inside of all the images in the background. At this point I could choose the Save command from the File menu or I could just close this item and say Yes, I want to save the changes. And now those changes will be represented inside this original composition right here applied to each one of the layers based on the original girl Smart object.