GiMP University

X-Hatch

noreply@blogger.com (FX Turk) — Sat, 01 Mar 2014 16:11:00 +0000

Hi everyone! Sorry the blog has been dormant for so long. I have had other things to do, I guess.

Back in the 90's, when Photoshop 6 cost a lot of money and there was no GiMP, there was a plug-in for Photoshop called X-Hatch by a company called Inklination. It was a ridiculously-powerful little plug in that allowed the user to do all manner of really interesting renderings which are all types of cross-hatch effects to a photo.

Unfortunately, that plug-in has simply vanished and there hasn't really been a replacement for it as far as I can tell. I have a clever little app on my Android pad which will do something like X-Hatch to photos, but it's just not the same.

This tutorial is also not exactly the same, but it will be able to move you from something like this:

Royalty-Free Image provided by http://www.sxc.hu

And with a few simple adjustments and a clever use of layers, come away with something that looks like this:

CLICK TO ENLARGE

Click that second image to get the full effect, and then download this file, which is my x-hatch base file. You will need it to create the pen strokes once we have done some basic manipulation of the image. That file is pretty big (about 61 MB) but it's worth it for this effect.

Let's think first about how an x-hatch image needs to be created. Take a look at this version of the original image:

This image is effectively a 4-color grayscale image of the original (white plus 3 gray levels), but it is made by taking the image and broadly adjusting the threshold 3 times: once for low-threshold contrast, once for middle-threshold contrast, and one for high-threshold contrast. The three layers are then flattened and we get something that effectively shows us all the brightness levels of the original, compressed into 4 categories. What we need to do is to turn each of those threshold layers into pen-stroke layers, and then flatten them to get a cross-hatch effect which makes sense when you look at it.

But wait: some of you are now asking you browser window: "What the heck is a 'Threshold'?"

That is a great question.

A decent brief explanation of what "threshold" is can be found at this link, but for the lazy or impatient, GiMP allows you the user to transform an image into a high-contrast black-and-white image by selecting the cut-off brightness value for converting the color pixels in your image to either black or white. That may sound like mumbo-jumbo to you, but think of it this way: the colors in every picture can be represented by a chart (or "histogram") ranging from dark to light like this:

On this histogram, the left side indicates the darkest shades, the right indicates the lightest. To see this histogram of an image you're using, go here: Colors > Threshold. You'll get this dialog window:

The sliders at the bottom of the histogram tell the tool how you want to convert your image. The black slider causes the threshold to set the cut-off point (the threshold) to covert the pixels to either black or white starting with the darkest color, and moving right includes more and more lighter colors. The white slider starts with the lightest color, and moving from right to left causes the tool to covert darker and darker colors to black. For this tutorial, are are only going to monkey with the black slider, but you can quickly figure out how both sliders can help you create some interesting effects.

... for example ...

For this first part of this tutorial, Let's simply set up the image into threshold layers using 3 different threshold settings. Open the sample image in GiMP, and then look at the LAYERS dialog window. You should see one layer, which is the base layer of this image.

CLICK TO ENLARGE

Under the list of layers are 6 layer control buttons. In the example above, I have highlighted in greed the "duplicate layer" button. Click it 3 times, and you'll get 3 more layers identical to the first one, and your layers dialog window will look like this:

When you have 4 layers showing, it's always useful to rename them to remember what you;re going to use them for. To do that, double-click one of the layers, and its name will appear highlighted. Type in the new layer name as you see fit.

The thing about layers is that they are exactly like sheets of paper, and the one on the top is the one you can see -- unless you do something to is which either makes it invisible, translucent, transparent, or function as a mask or filter. Since most of those are out of the scope of the tutorial, let's just show you that you can make these layers completely transparent so you can see the layers underneath them.

Simply move to the dialog box, click on the "eye" icon in the "LOCK" column each of the 4 layers, and watch what happens. Your Layers dialog will look like this:

but your image window will now look like this:

That's because you have made all the pixels in all your layers invisible down to the transparent file under the image itself. This will come in handy as we start to manipulate this image with the Threshold tool.

To finish up this first step in this effect, let's make one layer the high-threshold contrast, one a mid-threshold contrast, and the last a low-threshold contrast. You'll want to leave the base layer as-is for the last step of this effect.

So make the top layer of this image visible, and select the layer. Then go to the "Threshold"tool (Color > Threshold) and pull the black slider to the left until it is pointing at the first hash mark (which is the 20% threshold). You should get something like this:

Click "OK" to convert the layer, and then make it invisible. Make the second layer visible, and select it. Go to the threshold menu, and pull the slide to the second hash mark (which is 40% threshold). You'll get something like this:

Click "OK" to convert the layer, and then make it invisible. Make the third layer visible, and select it. Go to the threshold menu, and pull the slide to the third hash mark (which is 60% threshold). You'll get something like this:

So now you have three layers that should look like this in your layers dialog:

What that leaves us with is the three layers of contrast we need to make the hatch effect work. If you're ready to continue, open the file "4-way-hatch.xcf" which I linked you to up above. When you do, you'll find something like this in the Layers dialog when it is the active document:

It's 4 layers with 4 different hatching patterns -- diagonal slanted right, diagonal slanted left, horizontal, and vortex. We'll be using the top 3 for this tutorial, but I'll show you two alternative renderings using the vortex pattern so you can brainstorm your own uses for this file.

Before we use those hatch layers, go back to the image you want to convert to hatches. Select the top layer and make it visible. Now click the first button on the left of the layers dialog, which is the "new layer" button. If you simply press return when the "new layer" dialog pops up, you'll get a new, blank layer on the top of all your other layers. Select that layer and click the "Duplicate Layer" button twice. Now you should have something like this:

The great thing about the layers dialog is that you can move the layers around in the stack just by clicking and dragging them around. What you want to do is sort the layers so that the top layer is blank, then the next one is one of your threshold layers, then a blank, then a threshold. Make it look like this:

Now you're ready for some hatching. This next part takes some patience, but it's actually really easy.

1. Make your highest THRESHOLD layer visible, and then click and select EMPTY layer ABOVE it.
2. Select the "4-way-hatch" image, and select the TOP layer of hatch lines in the layers dialog. Now click "CTRL-A" to select the whole layer, and then "CTRL-C" to copy it.

3. Select the IMAGE window, and then PASTE the copied hatch marks into the top empty layer. After you paste, click the Anchor icon in your Layers dialog to anchor the hatch lines into the layer

At this point, you will have something that looks like this in your LAYERS dialog:

The EMPTY layer above your highest threshold level now is filled with repeating hatch lines. It's important that the hatch lines layer is ABOVE the image you're going to hatch our for the next two steps.

4. Make sure the HATCH layer is selected, and then go to the drop down "MODE" menu at the top of the LAYERS dialog. Drop down the menu so you can see this:

The MODE menu allows you to apply the layer you have selected to the layer(s) beneath it using different algorithms to create not color or lightness effects. At the end of all that geek speak, the layer's "mode" causes it to combine with the layers under it using a special effect. "Normal" mode is simply to cover the layers under the layer selected using the opacity you choose. For this tutorial, we are going to use the "Screen" mode which takes the layer you have selected and applies it to the visible layers underneath as a sort of fill. When you select "Screen" from the mode menu, you should get something like this:

All the black in the high threshold layer is now filled with the hatch marks from the top layer -- but the layers are still distinct. We want to combine them permanently into a layer for future use.

5. Right-click (Control-click on a Mac) the top layer (the one with the hatch marks) and select "merge down" from the context menu.

What this does is permanently apply the hatch layer to the image layer using the "Screen" mode, making the layer a hatched image layer.

6. Repeat steps 1-5 for the other two image layers, but as you do it, use a different hatch layer for each threshold layer. The top layer we just did has hatch marks running diagonal from bottom-left to top-right. When you are done, you'll have 3 layers plus the base layer in your Layers dialog, like this:

From this point, the rest is easy. Select the TOP layer, go to the MODE menu, and select the "Darken Only" mode. Select the SECOND layer, go to the MODE menu, and again select the "Darken Only" mode. When you're done, you should have something that looks like this in the main image window:

All that's left is to use your context menu to merge the layers into one and this is essentially finished.

The one alternative I wanted to show you uses the use of the vortex hatch layer. There's a certain flatness to the original method -- the curves of the subject's face don;t show through. The totally-straight lines look inorganic.

However, if you use the Vortex hatch for the for the darkest threshold layer, you can get something like this:

The lines running in a curve add a more natural feel to the hatch effect.

Gimp 2.8 WinXP release

noreply@blogger.com (FX Turk) — Fri, 18 May 2012 13:51:00 +0000

Well -- hi everyone. long time no see!

As I type this, I'm downloading and installing GiMP 2.8, and looking forward to testing it and reviewing it.

A quick work to readers and users of this blog: apparently the content of this blog is pretty relevant since I have posted here in years and it still gets a steady stream of traffic. Thanks to all of you who keep this blog alive by coming and learning something here.

Long Time No See

noreply@blogger.com (FX Turk) — Sat, 02 Jul 2011 18:04:00 +0000

Greetings all. Just a quick update here in the virtual academy for free graphics competence.

There's a new book out from PACKT Publishing by Juan Manuel Ferreyra -- GiMP 2.6 Cookbook.

On the "con" side:

It's not in color, which is a major setback on many of the tutorials. To see the tools actually do their stuff, you have to see the result in color.
It's $44.99 as a softbound book, which is pretty pricey. Comparable books will top out at $29.99.
The approach to grappling with GiMP in this book is not systematic, so the n00b might learn something, or might not.

But on the pro side:

While the approach to the book seems to be scatter-shot, Ferreyra really does cover a lot of ground in this book -- lots of topics I wish I had to time to cover here. There are sound foundational learnings to be gleaned from this book.
It does come in a PDF version for a much more reasonable price, and that does present the tutorials in color, so you will find it far more satisfying and helpful.
It gives you the right impression of GiMP, which is that it's a very decent intermediate graphics manipulation tool, and sort of teases you to want to play with it some more to see what you can do with it.

It's another way to become conversant with GiMP, and will be useful for anyone who wants to get started with the tool.

GiMP Labels

noreply@blogger.com (FX Turk) — Thu, 22 Oct 2009 21:32:00 +0000

A friend e-mailed me a link to this helpful site which is promoting the open source package "GiMPLabels.

Check it out.

Text Outlines

noreply@blogger.com (FX Turk) — Fri, 21 Aug 2009 14:04:00 +0000

Yes, I know: I haven't posted ONE SINGLE POST from the previous request, but I got another question/comment which is easily disposed of that we are going to deal with today: Texts as paths.

For many of you, this is already way past your tolerance for geek-speak, but it's an important issue for people who, for example, build logos or are really laying out text on graphics. The hard-core users will use Illustrator for this anyway which, with a key stroke, will convert any text to the vector outlines of the text for use in the file, but we're all here trying to save a little money and, honestly, try to keep our hobby inside a budget-friendly box.

What is a Text Path?

That's a great question. When you use any application on your computer, regardless of platform, your computer has a system of files which tell it how to use the ASCII table of characters. (yes, I know: just bear with me) those files in the "olden days" could be either a pixel map which actually renders the font letter by letter as the white and black dots on your monitor, or it could be a pixel map linked to a file full of math formulas which, when paired with TrueType technology, made fonts you could scale easily and get great results on either your monitor or on a like-minded printer.

Today, almost all fonts are, in one way or another, descendents of TrueType fonts in that they are essentially files which map the outline of the characters in a font so that you can scale the characters to any size, and you can also print them to almost any print at the scaled size without the fear of jagged edges.

In day-to-day use (for example, when you're blogging and just dumping text into the bandwidth), you really don't give a hoot about this. If you type "Sarah palin is my Homegirl" or "Sarah Palin needs to go home, girl", you really don't care if there's a bitmap font or an outline font rendering the final result on your reader's computer. But if you're doing something more robust -- for example, something like this:

The question of the outline of the font becomes a serious matter. Let me point out why. Look at this closeup of part of that image:

And because I'm a little pressed for time, in your mind switch where it says "text fill" and "inner border" -- I goofed and I don;t have time to fix it today. There you can see the multi-layered effects which, frankly, are hard to re-create in GiMP because it is not a vector tool but a raster tool: it is made to manipulate pixels and not paths.

So What?

The big "so what" is that some of you have been clever enough to find the "Path from Text" button in the Text tool dialog dock. For those of you not so clever, or really not even informed enough to try to be so clever, look here:

That button down there can give you the path for the text you have just typed -- but here's the thing: it's not a path the way you get a text path in Illustrator. Not really. You are far and away better off to use your selection tools to create a text path for all GiMP 101 and GiMP 201 application of this tool than you are to try to understand and master the output of the "Path from text" button.

So if you want to make a knock-out with your text, or do some fancy-schmancy logo work with GiMP, my suggestion is to keep your text in its own layer to work with it, use your selection tools to create your outlines, and work smarter, not harder. GiMP is not a vector design tool, and trying to use it that way will, in the end, make you less happy than you want to be over hobby work.

Which, ironically, is a great lead in to the free-hand drawing tutorials. I promise to get back to those soon.

Freehand Drawing (intro)

noreply@blogger.com (FX Turk) — Sun, 02 Aug 2009 23:04:00 +0000

Welcome Back -- it's been a long time since GiMP University had a decent update, but truth be told we keep getting visitors, so I feel compelled to come back and add stuff for you in order to keep you interested.

My friend Aaron asked me about freehand drawing in GiMP today, and it's a good question. For example, if you want to make something for a project where you can't find sufficient clip art, whaddaya do?

Well, my first reaction (and it's blasphemous, I know, so try to stick with me) is that if you can't find the component image you're looking for on the web, you're probably not looking very hard. Yes, yes: I know the purists reading this are about to launch into an jeremiad on copyright and fair use, and artists gotta eat too, blah blah blah. In theory, I agree with you: it's completely immoral to steal someone else's artwork and use it without permission to make t-shirts, or mugs, or someone's album cover, or whatever. No question.

But most of you who are GiMP n00bs (that is, the actual readers of this blog) are not making art to sell, or art to sell someone else's stuff. You have a blog, is has a handful of readers, and you like to spice it up a little, and you sell no products. You're not putting anyone out in the street by your use of their image which can be found via a Google search, and you're not putting that image someplace it wasn't before you laid your grubby GiMP on it.

But that said -- and with due respect to those who toil away at this stuff for a living -- sometimes what you can find with a Google search is simply not what you need. How do you make effective free hand drawings with GiMP? Can you?

Well, yeah. Over the course of the couple week, I'm going to show you several ways to use GiMP to make free hand drawings -- from the simple to the fairly-involved.

In the meantime, think about this: if I had to draw the object I need for my project with traditional art tools, which tools would I use? The answer to that question helps us understand what GiMP has to offer to user for freehand illustration. You look at and look for the tools in the GiMP tool box which look like those traditional tools, and I'll be back later to walk you through some basic projects to give you a feel for this sort of thing.

More on transparency and indexed color

noreply@blogger.com (FX Turk) — Tue, 26 May 2009 01:56:00 +0000

Welcome back. I had a follow-up question on transparency which asked a question about this:

Particularly, the question came up that the reader couldn't get the check box to remove unused colors to be active so he could check it and save his image as transparent. In case you don't understand, here's what I mean:

That part that's lit up in that picture. The truth is that you don't need it checked -- if you put your image on a transparent work space, change the mode to "indexed", when you save it as a ".gif" file it should leave the transparent parts transparent whether you remove the unused colors or not.

But the truth is this: if you remove unused colors, the file will save to a smaller size, and while in this world of DSL, 3G and high-speed cable you prolly don't care about file size, you can be a good neighbor and reduce your file sizes for the sake of merely improving the net load time of your images.

To get that box checked, do this:

You can check any of the color-indexing options, and the check box will magically become an active check box. And you can leave it checked in case you think you're missing something when you tell GiMP to reduce the colors for you.

One last note: without getting mired into Geek Speak, when you reduce colors you will have the best results when you reduce to these numbers of colors:

16
32
64
128

These are the natural breaks in color resolution which your computer still understands even though for the most part you all have something like 65,000+ color HD-resolution monitors. This is what will get you the best file size reductions because of the way .gif format helps your computer and your software think about color.

Transparent Background Revisited

noreply@blogger.com (FX Turk) — Mon, 13 Apr 2009 02:29:00 +0000

Yes: WOW. This is such a simple tutorial, but it's probably one which many of you have considered and (maybe) not quite gotten away with. Avid reader "D" has a Visio document which he saved in .bmp format, and now he wants to open it in GiMP and take the diagram out of the source file and create a new file in which the background is transparent.

Excellent.

I've created a source file for us to use, thus:

This is a fairly-rudimentary example of what "D" is thinking about, but it's a size we can work with and see what's about to happen to our image.

Now, this image is 488x488, so our first step is to create a target image of the same dimensions with a transparent background, thus:

Now go back to our source image and select the white background using the "magic wand" tool. After the white is selected, go under the "Select" menu and choose "Invert" and then "Grow...", telling GiMP to grow your selection by 1 pixel.

What you have now is a selection of your image which is mostly protected from the jaggies -- the rough pixels which will make your final use of the image mostly perfect. I say "mostly" perfect for a reason we will discuss in a minute.

Copy that selection and go back to your transparent target image.

Now, when you "paste" what you have copied from your source image into the target, your "Layers" dialog is going to look something like this:

Right? The pasted stuff is in a "Floating Selection (pasted layer)", which is deceiving -- because it needs to be anchored to the transparent layer so we can go on.

Just click the little anchor button in the "Layers" dialog, and let's move on.

Now, the last step here is critical -- because the n00b will think, "well, just save as a .gif now, and I'm golden.

NO. You are NOT golden. See: a .gif file requires the image to be in INDEXED color mode to save exactly as you see it, so first you have to go into the "Image" menu, and under "Mode" select "Indexed".

This will produce this dialog, for which you should make the options look exactly like this:

And this really will have two effects on this image. The first is one you may or may not care about, but by reducing the color pallet to 16 colors, you will make the file significantly smaller. But the second is the real money-maker: this allows the GiMP to save the image as a transparency when you save.

Now, before we save, let's look at this image one last time. I know all of you by now have noticed the white border around the image. This isn't really all white: it is the field in this image where all jaggies get buffered to some intermediate color so when you lay it over some other image, the original image on transparency doesn;t have all manner of jagged edges. It simply has a one-pixel wide white boundary which, in most circumstances, is totally agreeable to the eyes. And this is especially useful when your source image isn't as high-res as this one is: it protects the original image from completely disintegrating when you paste it over something else.

So now we're going to save the target image thus, using "Save as ...":

And when the "GIF" dialog comes up, just click "Save". And what you should get in the end is this:

And the chopped-off parts here are Blogger's fault. But notice that my image shows the gray of the background through without making you crazy from looking at jaggedy edges.

Makes sense? I knew it would.

GiMP vs Photoshop, revisited

noreply@blogger.com (FX Turk) — Fri, 03 Apr 2009 11:16:00 +0000

via Facebook, I found this link which talks about how to make GiMP more Photoshop-user friendly.

I have read it, but not tried all the suggestions. However, I figured you readers might find the info useful.

HT: Jon Swerens.

Welcome to Gimp 2.6

noreply@blogger.com (FX Turk) — Mon, 30 Mar 2009 02:03:00 +0000

Well! Back when this blog started, I made a promise not to upgrade until it was totally safe, and now I think it is. 2.6 was not actually dangerous at any time, but I have gotten good results from testing it in the current release on my Mac and my PC in version 2.6.6, so my suggestion to you is to go ahead and grab it as soon as possible. Still Free, so that's brilliant.

Some of the tools have been upgraded. I like what they've done with the lasso tool in particular.

All the old posts on this blog now bear the tag "GiMP 2.4/2.5", and all future posts will bear the tag "GiMP 2.6" until such a time that the upgraded version moves on sufficiently to note it.

And since this isn't much of a post, I'll open the comments for any suggestions or questions. I know one outstanding tutorial suggestion is how to make your photo into a pseudo-cartoon like my avatar is. If you have other suggestions, here's a good time to post them for my review.

Off with his Head (3b & 4)

noreply@blogger.com (FX Turk) — Sun, 22 Feb 2009 02:50:00 +0000

Yes: welcome back. Sorry I haven't been very motivated with finishing this series or posting here. In spite of the outcry to finish this series, the truth is that if you have followed this blog at all you have all the tools you need to finish the job.

What many of you have right now is this:

Phil's head has been color-adjusted so that it looks like it almost has the same skin tone as Batista does, and since it is safe in its own layer we can do almost anything we want with it.

What we want now is to orient and size Phil's head to match Batista's head, and then make the overlap seemless.

Now, no foolin': you can absolutely do this without any instruction.

Your first step is to tilt Phil's head at the same angle as Batista's head. You can do this in any number of ways, but the best reference guide to making this happen is to match the centerline of Phil's face which runs down the bridge of his nose to the same angle of the same line on Batista's nose. The more clever among you might try to match the line which runs eye-to-eye on Phil's head to the same line in Batista's head.

Either way, tilt Phil's head by selecting his head in the layer, then rotating it using the rotate tool.

At that point, you'll notice Phil's head is bigger than Batista's. Scale Phil's head so it is the same size. Again: this is barely 101 stuff.

The two things you might have a problem with is that Phil's head will not cover Batista's head adequately for a clean paste, so you're going to have to trim Batista's head off at the collar line to give Phil's head a clean space to paste into.

But the problem is that there's a gradient in the background behind Batista -- not a solid color, but a change from light to dark, right to left. So what you have to do is select his head, pick colors that will give you a decent gradient, then pull the gradient into the selected area.

Like this:

I have used the lasso tool to select Batista's head, and what you can't see is that I have feathered the selection by 2 pixels to give whatever it is we do here a softer edge.

In my sample image here I have dropped in labels "A" and "B". These represent the places I picked to use the eye-dropper to select a BG color and a FG color in order to pull out a gradient across Batista's head. When I go to the gradient tool, then, all I have to do is pull a diagonal line from "A" to "B" and I'll get something which blots out Batista's head and gives a presentable blank slate for Phil's head to fill in.

Like so:

The most very wily of you will see that, before deselecting the image, I also ran the "Gaussian Blur" filter to further mute the edges so that there are no places which are obviously not right. And since all of you are working at 2x resolution to shrink all little stuff into complete irrelevance in the final product, you will appreciate this step even more.

Anyway, that's it. That's all it takes to finish this project. Now you make Phil's head's layer visible again, drag it into position, and you get something like this:

If you are especially anal, you will use the smudge tool to make the place where Phil's neck touches Batista's neck without any cut line.

You can axe any questions you have in any of the Q&A posts.

Off with his Head (thing)

noreply@blogger.com (FX Turk) — Sat, 17 Jan 2009 16:09:00 +0000

OK -- I have the alleged "part 4" of "Off with his head" to post yet, but the rest of this tutorial seems rather obvious to me. Does anyone really want the part 4 or need it to finish what we started?

Off with his Head (3a)

noreply@blogger.com (FX Turk) — Mon, 29 Dec 2008 03:33:00 +0000

Welcome back. So far you have a basic target image for our exercise which has a layer created in it for the sake of laying on a victim's head, you have blown that image up to double resolution for the sake of our exercise, and you have selected a new head out of an old image using the scissor select tool.

Before we get to the nuts and bolts today -- and this is a long lesson, so bring a lunch -- let's look at some of the challenge we face in the project.

Here's our target image, the pro wrestler "Batista":

And here's my friend Phil:

The first major challenge to get what we want here is the way Batista has his head tilted. It's very roguish, I am sure, but it leaves his neck at an awkward angle. While we are working here, we're going to have to make sure we tilt Phil's head in a like manner. Also of note is the fact that Batista's profile is a sort of 3/4th view of his head. We're lucky (not really lucky; I picked this picture of Phil’s head for this purpose) that Phil's head is also a sort of 3/4ths view, but notice that their heads are not exactly the same size and shape. We're going to have to remove some of Batista's head before we lay in Phil's head to get a more natural look.

The last major challenge -- which is probably the key to making this project really work -- is that Batista obviously spends a lot of time in some way sunning himself. He has a marvelous bronze complexion -- and my friend Phil not so much. Phil has a normal, human skin tone. If we paste Phil's head on Batista's neck without some kind of color adjustment, we're going to have an extremely-unnatural edit, and our project is going to be less than successful.

So: last time we learned how to select Phil's head out of the "victim_head" image. Since it's relatively easy, deselect that image. We'll reselect using the scissor select tool in a few minutes, but now deselect and use your layers dialog to add a layer to the image containing Phil's head.

Now go back to the image of Batista. I have blown up a section of the image here:

Look in the area inside the red circle I made -- there you can find the darkest tones on Batista's skin. What we want to do is select the darkest of these tones into our color pallet. To do that, use your eye dropper tool (also called the "color picker"), select the Batista image, and move the dropper over one of these darker areas and click, so you get something like this in your color pallet:

Right now, some of you are saying, "Frank: the eye dropper isn't selecting any color." Yes, well, look at your layers dialog. You have the "new head" layer selected instead of the "Background". If you select the background layer and try again, you'll have better success.

So what we have is a sample of color from Batista's skin, and we want to make Phil's skin look that sunny and brown. This is a somewhat-advanced technique, so I'm going to go slowly.

What we DO NOT want to do is simply use the bucket tool and pour the color onto Phil's face. That will simply leave giant blobs of color on Phil's face and give us a completely-useless result. What we are going to do instead is create a layer of the dark skin tone we just pickered off of Batista, and saturate Phil's skin tone with that color.

Go back to the image of Phil's face. Make sure you have selected the new layer we made a minute ago, and use the bucket tool to fill that layer with the skin tone color. Your main image should look like this:

And your layers dialog should look like this:

To get a preview of what we're about to do, go to your layers dialog, and notice the "mode" menu near the top of the dialog.

Select the "new color" layer and drop down the "mode" menu to select the mode "color". In our preview of what we're about to do to Phil, you'll see something that looks like an aged photo -- something you may have seen called a "sepia tone" image. So in effect, you have just gotten a little side-lesson in how to make your normal photos into sepia-colored images.

This technique, btw, is the GiMP version of the Photoshop “apply image” command – we are blending two images to get a final result which capitalizes on various aspects of the originals. We just have one image with is, obvious, just a single color. What we are doing is using a “blend mode” on the single-color layer called “color” – a blend mode which applies the color of the selected layer to the layer immediately under it. And that’s all the geek-speak about this you’ll get today.

We don't want this image to be a sepia-tone image: we only want Phil's skin to get bronzed up -- but his hair and eyes need to look like they are the right color. So for now, it's fine to leave the "new color" layer in "color" mode. To disable it for a second, click the eyeball in the layers dialog to make the "new color" layer invisible, and we'll return to Phil.

We have a couple of choices to get rid of the bronze color in unwanted areas, and we're going to do the easiest one first. Go select your "Select by color" tool:

You will notice that Phil's hair and his jacket (at least insofar as this photo is concerned) are pretty close to the same color. In your "layers" dialog, select the "background" layer. In your "select by color" tool options dialog, set the threshold to "33.0", like this:

What that means is that when you click the tool on a color in this image, it will select that color and any color like it out to a threshold of "33". To tell you more than that is to bore you with geek speak, and you're already bored enough.

I got the best success with this image by clicking on a dark area in Phil's jacket, and then holding down the "shift" key so I could make a second selection and merge it with the first, selecting a part of the dark area in Phil's hair which wasn't selected in the first pass. My result looks like this:

You might get lucky at this threshold level and get a robust selection on the first click if you find a color in the right range. But once you have your selection, make the "new color" layer visible again and select it in the layers dialog. You'll notice your selection of the darker features of this image are still selected when the bronzing layer is added back into the mix -- which is exactly what we want. Now all you have to do is hit the "delete" key -- and BINGO! The bronze is removed from the dark areas of the image.

The very-fussy among you will want to grab the eraser tool and clean up the flecks of bronze in Phil's hair. I leave that to your discretion.

However, the rest of us are going to take the eraser tool in-hand for another purpose: unbronzing Phil's eyes. Set the tool's brush to "Circle(11)", and make sure you're in the "new color" layer still. That circle size should be the right size to easily clear the bronze out from Phil's pupils -- and leave us with a very fine result.

Now, look: it's not hardly the best result you could have that this point. You might want to de-bronze Phil's beard a bit; you might want to touch up his hairline for bronzing issues. But because this step in the cut-and-paste is so long in and of itself, I am not going to be that kind of a purist today.

What I'm going to do at this point is flatten the layers of this image to make the bronzing stick. Go to the "Image" menu, and select "Flatten Image".

And at last I'm going to use the scissors-select tool we learned to use last time and cut Phil's head out of the victim's source image and into the target image, thus:

You'll notice in this image that Phil's head is safely in its own layer, waiting for us to use it in our evil plot. Until further notice, make the layer invisible.

Because this step is so longer, we’re going to call it quits in the middle before we rotate, scale, and apply the new head after erasing parts of the old head to make sure it doesn’t show through. The really-eager among you will, of course, work (ahem) ahead. The rest of you will have to wait for my next installment.

Off with his Head (2)

noreply@blogger.com (FX Turk) — Thu, 27 Nov 2008 13:44:00 +0000

So we have our source images, and we have started building the target image. As promised, what we have to do next is take the head off the victim in a clean way that allows us to put it in the target.

Here's the victim, my friend Phil:

Today I'm working on my Mac laptop, and the screen shots are all cap'd from there. The tool we're going to use to get Phil's head out of the original picture is called the "scissors select" tool:

Now, we could use the magic wand (which is how we did this sort of thing elsewhere), or we could use the lasso tool, or the paths tool, or the foreground select tool, but I think the scissors tool actually gets the best results here. The scissors tool is a polygonal select tool which also senses the edge of the object you are trying to cut out, so in situations where the object to be cut out has a good contrast line from its bacjground, this tool can help out loads.

Let me show you what I mean. Phil, frankly, has great hair. To test out how this tool works, we're going to experiment fist on his hair to see what the tool does. First, select the tool, and click a starting point here:

I have enhanced the dot because Mac won't capture the cursor pointer, but now that you have selected one point on Phil's head, select more points around the top of his hair in a clock-wise fashion, until you get something like this:

Notice that I selected points relatively close to the edge of Phil's hair, and that I stopped where the part in his hair begins. But I didn't really "stop" -- as in, "closing my selection". I merely paused so you could keep up. Now, to show the power of the scissors select tool, move your cursor to the beginning point of your selection, close the selection area by connecting your last point to your first point, thus:

As soon as you click the final point, you'll get something like this:

The tool took it upon itself to find the high-contrast edges of the area you selected, and made the selection boundary the place where it understands one color ended and another began. I am sure that some of you long-time readers of this blog what to know why we didn't do this with the chess piece way back when, and I'll tell you why: the contrast between the eggshell-colored chess piece and the light squares on the board is not sharp enough to get a good cut without lots of selection points. Also, it was good for you to learn the magic wand tool first.

But that said, the advantage of the scissors select tool in this case is that we can get all the organic lines of Phil's head captured without too much fuss. As we select point in reasonable intervals along the border of his head, the tool will help a brother out by actually interpolating the curve between the points.

So go ahead: unselect our test, then start the selection process over again, this time tracing out something like this:

And then hit the "enter" or "return" on your keyboard to translate the pointed outline to an actual "selected" outline, thus:

And the real novices out there are thinking we are just done already, but no: the fluent users remember that the selection border needs to be feathered before we try to pull this face out of this picture, so let's feather the outline by 3 pixels to get a nice soft edge around Phil's face so we can get something approximating natural in the final result.

Next time we're going to take this selection and do almost all of the heavy lifting of this little stunt in one post. However, this prep work is pretty indispensable.

Stay tuned.

Off with his Head (1)

noreply@blogger.com (FX Turk) — Sat, 22 Nov 2008 18:50:00 +0000

UPDATED: Hey -- before your read this tutorial, some of you will notice it has been changed from its original form where I put Phil's head on a mostly-nekked man's body to where I put Phil's head on the fully-clothed body of the wrestler Batista.

I changed this tutorial for one reason only: conscience. I thought it was somewhat funny to put Phil's head on the body-builder's body -- until someone linked to it randomly and it seemed pretty lewd to me. I apologize to Phil, who is above reproach as far as I'm concerned, his wife Darlene who has always been a friend to me and my family, and to any readers I offended.

Thanks for letting me get that off my, um, chest.

Alert Reader Yahooguntu asked the following:

Can you write a tutorial on how to take the head off someone in one picture and put it in place of someone else's head in another picture?

And of course the answer is "yes, I can". However, I will write it in blocks so that the various techniques needed to do this project can be understood discretely and people can learn at their own pace.

So what we're going to do is take this picture of the pro wrestler Batista:

And this picture of my friend Phil:

And we're going to create this irresistible hunk of man-flesh:

Now, for those who are self-starters, here are the steps we are going to follow, and if you can invent steps for yourself between posts here at GiMP University, go ahead and work at your own pace. The rest of you will have to wait for my amateur pace posts.

The steps:

[1] Create a working image for our project
[2] Remove the head from the victim
[3] Place the head in the target image -- including scaling, color correcting, and cleaning up
[4] cleaning up target image for final use

And to keep you from completely giving up on me, we'll do step 1 now as it is relatively easy. If you click on the image of the young body builder, it will open up in its own window, and from there you can download it to your desktop (same with the picture of Phil, fwiw).

Now, to keep us all on the same page, open that image in GiMP and do the following simple adjustments:

[a] duplicate the image using the "Image" menu, clicking "Image > Duplicate". Close the original in case you make a fatal error and have to start over.

[b] Double the size of your target image to 782 pixels wide, and allow GiMP to scale the height using the "Image > Scale Image ..." menu

[d] Use the Layers window and add a new layer to the target image. You can name it "new head" if you need that much help.

Just to answer your question before you ask it, we are blowing up this image to double the resolution to do our editing work so that when we have finished all the cutting and pasting, tiny flaws will get minimized when we blow the image back down to it's original size -- or smaller. That's a key tip which you should have seen elsewhere on this site: you want to do your editing work at double resolution so that you can worry about big stuff and not get consumed with pixel-level perfection.

More next time.

The request line is now open

noreply@blogger.com (FX Turk) — Fri, 31 Oct 2008 00:38:00 +0000

OK people: after an interminable hiatus, I have some time to make some new tutorials.

What would you like to see me tutorialize? What's the thing you want to do which you can't with GiMP?

If you could link to images that you want me to break, that'd be great.

Color Mode

noreply@blogger.com (FX Turk) — Fri, 01 Aug 2008 21:57:00 +0000

Astute reader "JC" was reading the last post, specifically this part of it:

And he asked, "Where is the 'color mode'? Where is the 'brightness/ contrast setting'?"

To which I say, "ah, you n00bs -- youy keep this blog alive. I thank God for you." Let's start with the image which I manipulated, which is this one:

The first thing we're going to do is open up the "Image" menu, thus:

Image -> Mode opens up an extra menu which lists the three major color modes -- and to preserve your actual interest in this tutorial, we are going to not think about "Indexed" color mode right now. But you can see in the image I dropped in here that there is a "grayscale" mode. "Grayscale" means "not in regular color but in shades of gray". We're going to comvert this image to grayscale, thus:

So this image looks like a line drawing (a little), but the lines are white and the "paper" is black. What do we do about that? Well, duh, we have to invert the colors -- so go under the "Colors" menu, go down to "Invert", and click.

Now we have gray on white, which looks better, but our muitant friend here looks like he has really bad skin. How do we fix that?

Well, we fix that by altering the contrast and brightness of the image. What we really want in the lightest parts of the image all to essentially go to white, but all the darker lines to go to something closer to black. And we do that right here:

That brings up this dialog:

"Brightness" values less than 0 make the whole image fade to gray; "Contract" values less than 0 cause the darker aspects of the image to soften. But raising the "Brightness" to, say, "25", and the "Contrast" ot, say "40" gets you something like this:

Which is, more or less, what I showed you last time without a lot of explanation.

Does that help, JC?

BTW

noreply@blogger.com (FX Turk) — Thu, 17 Jul 2008 11:22:00 +0000

If you have comments or questions about filters, you can ask them here.

And I have just looked at the screenshots of the GiMP 2.6 upgrade. It's a radical interface change, and it looks like some functions will be beefed up. For those of you tracking what this blog is doing, (or lack thereof) I'm not going to change from 2.4 until we get a tested and stable version of 2.6. And to be sure, I won't change the tutorials that I have posted here ever -- I may rework them and label them with GiMP version numbers, but until the end of 2008 at least this site will be tutorializing from GiMP 2.4.

Welcome back

noreply@blogger.com (FX Turk) — Thu, 17 Jul 2008 00:43:00 +0000

You know what? I missed blogging this blog. Most of my normal blogging is "important" and "thought-provoking", and I'm just glad for a minute that I have this blog where I can be merely useful and informative.

So welcome. I missed you, too.

Most of the entries so far have been basic tool entries – how to use the straight-up tools like "magic wand" and "color select" to get something cool to happen. Well, it turns out that the GiMP has some rather mad filters which can do stuff that you could prolly do by hand if you were insane and didn’t have to eat, work, or use the potty. These filters can just save you time and create … stuff.

The difference between a "filter" and a "tool" is that "filters" reside in the "Filters" menu, and they are generally applied to the whole image at once – or for you intermediate users, to a selected area of the image you're working on.

I have an image you can practice on all picked out as we apply 3 basic filters today:

Nice face, right? That's an image downloaded from stock.xchng from a photographer whose user ID is "Sammylee". It'll be our test case for these filters. Drop down the "Filters" menu and take a look at it.

We could walk you through boring junk like blurs and red-eye removal, but what if our crazy face here is a supposed to be a super mutant with psionic energy that is bursting out of his eyes? Is knowing the difference between "Gaussian blur" and normal "blur" going to get that done? No, it's not, so let's not, let's do something worth doing. Go down the menu to "Light and Shadow" and when the next popup menu shows itself, choose "Supernova". You get something like this:

Now, obviously, the giant radioactive zit on this guy's nose is not impressing anybody. But to get that crazy psionic energy coming out his eye instead of the pores on his nose, use your mouse to move the center of the Supernova in the preview screen to the part of his face (one of his eyes) we want the effect to actually be in. I'm choosing his right eye (which is on our left).

Once the center of the effect is where I want it, I'm going to manipulate the sliders to get the filter to look like something manageable, thus:

For those of you who aren't following what happened, I reduced the diameter of the center of the effect to 9 pixels; I changed the "spokes" to 38, and I upped the "random hue" value to 77 to make the colors in the spokes rotate randomly. The higher the value there, the more random the color generated. I'm going to repeat the process for the other eye leaving all the settings exactly the same – just moving the center to the other eye, and I get this:

Weird, right? Now, check this out: select "Edge-detect" from the "Filters" menu, and choose "Edge…". Don't ask any questions – just click "OK", and you get this image:

Now, that's crazy, right? But if you fiddle with the color mode and brightness/contrast setting, you can get this:

Which looks like a line drawing or a pencil sketch. Nice, huh?

OK – that's two filters. Here's the third and then I'm out of here. Go down the "Filters" menu to "Distorts" and select "Erase every other row". You get this dialog:

Now, before you hit "OK", make sure your FG/BG color is set so that the BG color is BLACK, and then set the "Erase/fill:" drop-down to "fill with background". Now click "OK" and you get this:

Like an old-time low-res TV image.

Now, play with the filters there in the Filter menu, and we'll come back with a project that you can try out with me to see how Filters and Tools interact with each other.

Sorry to be gone so long; thanks for your patience.

GiMP University logo (2)

noreply@blogger.com (FX Turk) — Wed, 30 Apr 2008 02:40:00 +0000

Just for you people who can't wait for the next tutorial, here's the final product:

And you can download my GiMP native file with all the layers still in tact here.

And here's how today's tutorial will work; I'll leave the comments open on this post, you downlaod the files, and axe me questions that you have about how I did or why I did whatever it is you see in the file itself.

I realize that's a lazy and lame way to run a tutorial, but I used only the techniques I have shown you so far to make this logo.

Fire it up.

Template Updates

noreply@blogger.com (FX Turk) — Sun, 27 Apr 2008 04:42:00 +0000

Some of you may have noticed that this blog uses Google Analytics. You'd have to be a real G33K to notice such a thing, but I know you're out there -- you use open source software.

Anyway, GA says that none of you are viewing this page on a screen which is smaller than 1024x768, so as I begin to update the template, I'm going to format it to give you more real estate in the content field -- the place where the posts show up. No sense squinting when there's 30% more of your screen to use.

Look for more over the next week.

Just a brief update

noreply@blogger.com (FX Turk) — Wed, 23 Apr 2008 14:58:00 +0000

Hi all. I have been on the run for two weeks straight and haven't had time to update the "how to make a logo" series. Just wanted to let you know I haven't abandoned this project, and that your e-mails of encouragement have been, well, encouraging.

Will update with a real tutorial later this week.

Reciprocal Link Love

noreply@blogger.com (FX Turk) — Tue, 08 Apr 2008 01:24:00 +0000

Just as a shaddout and reciprocal link love, GimpUsers.com has somehow featured this little hobby on its front page, and we love them for it -- they are the #1 feeder to this site right now.

Nice work. Thanks for noticin'.

BTW, while we're thinking about this, if you have a simple tutorial that you think will help other BEGINNER GiMP users put a good foot forward with our friend the GiMP, e-mail me your text or a link to your mocked-up text plus images, and I'll be glad to feature your contributions (with a link back to your home page) under the heading "Guest Tutorials".

GiMP University Logo (1)

noreply@blogger.com (FX Turk) — Sat, 05 Apr 2008 11:53:00 +0000

You saw some of my handiwork in a previous entry in which, I admit, I sorta violated the rules of tutorials by serving up examples of stuff that you can't really make with the tool you're allegedly demonstrating. Well, I'm about to upgrade this site with something a little more snappy than the generic "new blogger" template I'm forcing you to look at here in the beginnings of GiMP university, but to do that I need a "GiMP University" logo that doesn't make me cry when I look at it.

(BTW, one of the most frequently used search terms that get people here is "gimp make logo", so this will be one set of lessons which, apparently, people are actually looking for.)

Now, you would think that all we would have to do is go to the "Tools" dialog, drop down the "Xtns" menu, select "Logos", and choose a logo style to make our work simple and complete.

In fact, I recommend as independent study time you doing just that with the phrase "GiMP UNIVERSITY", and if you make something you're especially proud of, e-mail me a link and I'll link it here for the readers of GiMP-U to see and heckle or praise (as they are so lead).

I don't want to surprise you or anything, but making a logo is not a one-step process. And while I want you to really sort of contemplate that, I don't want to give you 6 transferable hours in Identity Marketing. What I want you to remember, however, is that a logo is like the face of whatever endeavor you're making it for. In that, it has to make a first impression, it has to win friends and influence enemies, and it has to be memorable and expressive, but not clownish.

And the first step in making a logo is thinking about type face. As we go there, keep in mind that your slithery little collection of system fonts that came with your computer is not even nearly enough interesting typography to really get rolling. Visit any of the font sites in the sidebar to get you some really cool fonts as soon as possible. (They don't pay me to say that, either)

Now, look at this image:

It is mostly not the font -- it's mostly a big red shield with a yellow wreath around it, but notice that the word "VERITAS" and the word "HARVARD" have at least two things in common:

[1] You can read them easily.

[2] The font is not more important than the actual institution.

Here's what I mean by [2]: When they designed this seal (this isn't the original seal, btw, even though it includes the original seal; it's very clever), they realized that what makes the logo important is the centuries-long tradition of Harvard. They didn't try to overpower more than 200 years of achievement with some 800-lb gorilla font. The seal needed to say "HARVARD". They could have done that like this:

Which would instead say, "scary gothic castle erected around HARVARD which is very important". And while in one sense they may mean that, (they don't take everyone, do they?) what they mean with the font the chose is something less ostentatious.

Now, my idea here is to build a GiMP University logo which, frankly, knocks off the HARVARD logo so that our logo implies to the observer, "we're the Ivy League tutorial for GiMP users". And to do that, I'm going to pick this font, which is called "Admisi Display":

See? You can read it right away, it keeps that nutty design element for the GiMP startup screen text where the "i" is lower case and the other letters are not (yes, you can fight with me about the "G"), and it doesn't complicate the rest of our work by employing a font around which the rest of the seal has to somehow conform itself.

This is completely entry-level stuff, I admit it. But GiMP university is not for snobs. It's for people trying to improve their graphics savvy, and sometimes you have to start with something as ridiculous as font selection to get anywhere at all.

More later.