Disclosure: I get a message asking about editing this or that effect almost daily now so I have decided to create an in-depth tutorial for the techniques myself and many others use in their pictures. I am aware that Vman has already written an editing tutorial but, and I’m sure he will agree with me, times have changed since then for both of us and we have both developed newer and better ways of editing. Also, this tutorial will hopefully put right some common mistakes that I see made in a lot of pictures now-a-days. Finally, please don’t perceive this thread as anything less than an attempt at helping the community, I am aware that I am not the greatest editor on the site by any stretch of the imagination, but hopefully this tutorial will help beginners.
The shortcuts and parameters in this tutorial are for GIMP 2.0. If you know your way around Photoshop or other editing programs already, the methods outlined in this tutorial will be easily transferable to your program of choice.
For this tutorial I will be using this generic picture, made to incorporate lots of different editing effects:
It might be a good idea for you to use this picture yourself as it might make the instructions easier to follow.
Before you should even think about editing effects and more importantly lighting, you have to isolate the main subjects of your pictures. Isolation is the process of selecting an object out of your picture and pasting it onto a new layer, thus “isolating” it from the rest of the picture. Isolation can be an incredibly time-consuming and boring tasks, and perhaps the worst part of editing, but get it out of the way at the start and it will seriously increase the overall quality of your work. Nine times out of ten this is going to be a character of some sort that you are isolating. Here it is my two Jinrai soldiers from Neo Tokyo.
To select the character, I use the “lasso tool”.
Some people use the pen tool and create a selection from a line but I don’t see any benefits between the two – it is a matter of preference I guess.
Depending on how far away your subject is from the camera, you may need to zoom in more. The key is zooming in just the right amount. Don’t zoom in so far you can barely make out where one thing ends and another starts, but don’t zoom in so little that you aren’t detailed enough. Start at a point you can remember (like the bottom corner, in this case a foot) and begin clicking carefully around your character, drawing an outline like so:
Continue to outline your entire character. Zoom in a little to check that your outline is okay, and zoom in to do small curves and little details. If you stay zoomed too far for too long you can make silly mistakes, like entirely miss out limbs or start isolating a random object in the background, so make sure you keep checking your work. Remember that you can grab previous points and drag them if you make an error.
Ah the hard work has paid off… sort of. Copy this selection and then, but without clicking off your selection, create a new layer [GIMP: “layer/new layer” or “Ctrl + shift + N”). Paste the selection onto this layer. If you didn’t click off the selection, it should paste right into the correct place. Hide your background layer (click the little eye icon next to the layer) and then use the lasso tool again to select any areas of your new layer where parts of the background show beneath and delete them (in this image, there is a little gap in between the legs where you can see the wall behind, most often these areas will be between the legs and arms of characters). Now you should have a lovely isolated character on his very own layer:
There is a faster way to isolate characters using “greenscreen” techniques but, in my experience with GIMP, this method of isolating often leaves errors and aliasing issues. It may be different in Photoshop, but I wouldn’t know.
Shading and Highlighting
The most effective shading and highlighting method uses only the “dodge/burn tool”.
It’s best to do the shading first. Pick a soft brush [GIMP: “circle fuzzy 19”], select “burn” from the tool’s parameters list. For the first set of shading, the main shadowing on the character, a big brush is okay [GIMP: roughly brush scale 3-4]. Take note of the direction of light-sources (in this case the muzzle of the character’s rifle and the light in the corridor) and use the lighting already on the model to guide you. Particular areas to shade heavily are areas where limbs, armour or weapons are blocking a light-source. After you have done the basic shading, begin to vary your brush size and add in more detailed shadows. Don’t be afraid to make jet-black shadows, but make them very small and only in areas that would be naturally dark (like under the ridges of the armour on this model).
Now we need to highlight areas of the character. Highlighting should be far less strong than the shading.
Select “dodge” and set tone down the opacity and exposure [GIMP: opacity 20, exposure 20]. You might need to tone it down even more for some areas of the character. Highlight in the same way that you shaded but, whereas you could create shadows in the previous step, do not try and create highlights from scratch with the dodge tool – you will only mess up the colours in the picture. For highlighting, simply accentuate the highlighting that the model and in-game lighting already has.
This is how my character looks after all of the shading and highlighting. Note the dark areas under parts of the armour (thigh armour, and shoulder pads especially) and behind the left arm and left leg.
Highlighting and shading can be done on the background layer but you’re best not to put as much detail into it, especially if the background is blurred, as it will look out-of-place and take the attention away from the main subjects of the picture. I feel shading on the background should mostly be used for creating shadows where Source does not (eg. on things leaning up against walls).
Most pictures won’t have a lot of motion-blur, depending on how daring you want to be. Motion-blur has a fairly simple technique but miss out one vital stage and it can really look like shit.
To start with, isolate the area you want to be blurred and paste it onto a new layer, as you did with your character (in this example, we shall be blurring the shells ejecting from the rifle):
Now, use the motion blur filter [GIMP: “filters/blur/motion blur”] on the layer (if you have more than one object on the layer and they are moving in different directions, as in this example with the shells, select the each object with a box before applying the filter to the layer). The key here is to think carefully about the direction(s) the object will be travelling in and apply the blur as suitable to the weight and size of the object, as well as its speed. The shell nearest the ejection port in this picture will be blurred more than the others, for example (as a rough guide, motion blurring of 5-10 should be easily enough for a small object like a shell). After applying the filter, the object will be motion-blurred but it will have a sharp outline on the leading edge.
To remove this, we need to use the “clone tool”.
The object that has been motion-blurred still has the original, un-blurred object on the background layer below. This is what is creating the sharp edge. To remove it, clone an area of the background similar to what would be behind the object [GIMP: “ctrl + left click”] and paint it over the object (to see what you are doing, you’ll probably want to hide the layer with the blurred object). After you have removed the object from the background layer, the blur should look much better:
Obviously, if you are blurring big objects such as limbs and weapons, it will be much more difficult to remove them from the background layer at all. In such instances, you might want to take a picture from the same camera angle with the object removed from the pose (so you can see what is directly behind). Then you can just cut out the area behind and paste it onto your background layer once you have blurred the object.
This technique is better for beginners. This is the technique Vman and I used to use, back in the day, and uses references from pre-made muzzleflashes (can be poorly made/over-used) and real photographs (can be poor quality pictures/hard to use). Some people will copy and paste their flash and leave it at that, but it will look terrible and unrealistic – there is more to do to get a reference muzzleflash to look decent. It’s not the best technique in terms of quality, but it’s certainly the easiest. For this tutorial, I will be using this flash:
Copy the muzzleflash onto a new layer of your picture. Remove the black background [GIMP: “colours/colour to alpha/”pick black] and then erase the any borders or watermarks that might be on the flash (our reference for this tutorial has a small white border that you will need to get rid of). Rotate and position the muzzleflash so it lines up with the barrel of the weapon properly. This is one of the points where you will probably realise the failings of pre-made muzzleflashes – they can’t cover every possible angle that you might have in a picture.
Desaturate the flash [GIMP: “colours/hue-saturation/” slide saturation all the way to the left] so that we don’t have a coloured outline. Use a fairly large, soft brush and remove most of the glow from the flash so that you have an almost solid white shape.
Duplicate the layer and then, on the bottom of the two flash layers, apply some gaussian blur [GIMP: “filters/blur/gaussian blur/”set “horizontal” and “vertical” to 50]. Colourise the layer with a tiny amount of yellow, just off-white
(sometimes it might be better to just leave it entirely white).
Duplicate the top layer again (the one that hasn’t been blurred, your original flash layer) and apply double the amount of blur than last time [GIMP: 100]. Apply some more colour to this one, and make it slightly more orange. Keep duplicating the original layer and doubling your blur each time until you have a pretty large light source (in GIMP, I normally go as far as 800 gaussian blur). The more blurred a layer is, the more red the colour should be. I finish on this colour in most of my flashes:
The layer that is blurred the least should be at the bottom of the flash layers but remember that the original flash layer, the one with no blur or colour, should be at the very top. The muzzleflash and its layers, once it is finished:
Ah, nothing like some violence to spice up a picture. But it’s not violent enough right?
Fore this section of the tutorial I will be using this picture:
Gore! A lot of people mess this up but it’s actually surprisingly easy if you just have the right resources and method. Most use blood brushes, but I find the splatter reference images from this site much high quality and yield better results:
Open up your chosen splatter picture, paste it onto a new layer of your picture and resize it and rotate it as appropriate. “Colour to alpha” the white, so we are just left with the red. Then darken it. Darken it a lot. One of the biggest mistakes made with blood editing is having it far too bright. In GIMP, I can pull the brightness bar in the “brightness and contrast” window down to the bottom twice before the blood is anywhere near as dark as it should be. Depending on the camera angle, you will probably need to erase some of the splatter so it looks like it is coming out of the point of impact.
For a more 3D effect, zoom in quite far and, with a small soft brush, use dodge and burn to show light hitting the blood (generally, just burn the underside of droplets slightly, and dodge the top side).
All that is left to do now is apply some motion-blur so it doesn’t look like it his hanging in mid-air. Another mistake that is commonly made at this point is motion-blurring all of the blood in the same direction. This looks lazy and unrealistic – the blood isn’t coming from a single direction, it is coming from a single point. Using the “lasso” tool, isolate an area of the blood that would all be travelling in the same direction.
Lightly motion-blur the selected section [GIMP: 2-6 amount is normally enough]. Continue this technique with the rest of the blood, also taking gravity into account. Use more than one reference picture though – for an entry wound, I like to use a wide splatter with lots of droplets as well as a thin splatter to show where the bullet entered the character. For more control, it is best to put each different blood reference on a new layer. Make the different layers slightly brighter and darker than each other, to give the impression of depth.
And there you have it! Later we will go into more detail regarding wounds (with a technique that will work for entry and exit wounds as well as scratches and slashes).
For this part of the tutorial we will be using this image:
Poor Max. He’s looking a bit worse for wear isn’t he? Texturing is a technique that is a God-send for anyone working with ageing, low-resolution models but it can also be applied to decent models too. The method basically uses real-life images of materials to enhance the textures of characters and/or objects in your pictures. We will use this image to texture Max’s coat:
Copy the leather image onto a new layer of the picture. Resize it so that it isn’t too big and rotate it so that the direction of certain patterns in the image fits the object you are texturing. It can make things easier if you erase away the hard edges of the image with a soft brush.
Before pasting this, copy it so you can keep replicating it to cover the character (saves you having to resize it and erase and rotate every time). Carry on pasting the selection until all of the area you want textured is covered. Remember to keep rotating selections so that the texture follows the shape of the object.
So that looks pretty dumb right? Now we need to remove all of the texture that is overlapping different materials and the background. To do this, hide the layer with the texture on it and use the lasso tool to select areas you don’t want to be textured.
Delete the texture from this selected area. Now this must be repeated all over the picture so that only the right area is being texture (in this case, the coat). Like isolating a character for shading and highlighting, this is the dullest part of texturing. To speed things up though, assuming you have isolated your character first for shading and highlighting, you can use the magic wand to quickly select everything outside of your character (by clicking off the character on the character’s layer) and then by switching to the texture layer and deleting everything in the selected area. For this example though, we will have to manually remove all of the texture from outside the character because we have not isolated him. When you cleaned up the texture, it your picture should look something like this:
Bear in mind that to cover the whole area, you might need to use multiple layers to stop things getting too confusing. When you’ve got the texture on the whole area, it will look pretty stupid:
Merge all of the texture layers down. The texture will have become slightly blurry from all of the resizing and rotating, so sharpen it [GIMP: 20-40 is enough]. Desaturate the entire layer and set it to “overlay” mode. Play with the brightness and contrast to make the texture stand out less or more as needed and attempt to get it to roughly the same brightness as before you applied the texture. And voila; all of your hard work has paid off:
Much better, huh? If you work hard at texturing, posing with low-resolution textures should worry you no longer. You will even be able to make the crappiest models look half decent with practice, for example:
… compared with:
Still to come!
Muzzleflashes and explosions (two different techniques for varying abilities).
Rim-lighting (when, where and how to use it realistically).
Blood and gore (splatter, wounds etc).
Dust and smoke (weapon smoke, missile trails, dust and fog).
Lighting (overlays, gradients, colour-mod, layers, saturation and contrast).
Rain, snow and splashes.
Thanks for reading!