Welcome Web SIG Cleveland!
On this page you will find the photos I was editing with, as well as the steps I took during the talk. I hope you find it helpful!
Disclaimer: I’m typing up this page before doing the talk, so this is what I plan on talking about. It is, of course, very possible we get side-tracked, or that I don’t have enough time to cover this all, but this should still be a handy reference! Oh, also, my talk will be a live demo, so the photos might not end up looking exactly the same there as they do on this page.
First off, this is the photo I’m editing. It’s of a very nice couple taken in downtown Akron. It’s not a bad photo straight out of camera, but let’s see what we can do to make it just a little more exciting!
Rotate and Crop
The first thing we want to fix this this photo is to rotate it, as it’s not exactly level.
The rotate tool can be found on the toolbox, and my personal preference is to Select the Corrective (Backward) direction and then uncheck the Show image preview box. What this does is, instead of rotating the photo itself, it will rotate a grid. You can then line the grid up with the part of your photo that should be level, and when you click the Rotate button, it will rotate it as such.
That makes no sense when reading it, I know. Perhaps the screenshot below will help. The grid shows up as soon as I’ve clicked on the photo and I’ve rotated it so that one of the lines goes right across the top of the door.
When I click the “Rotate” button, my image now looks like this:
Much better, though you can see that there is now some transparancy we have to deal with by cropping the photo.
While it ultimately turned out not to be useful in this photo, when cropping, I usually turn on the Rule of Thirds guide so I can see if there might be a way to crop the photo that would take advantage of this rule (where you place subjects of your photo in the intersection of these lines or along one of the lines itself – often done with the horizon line). Here’s what I had selected as I did the crop on this photo:
After the crop, my photo looked like this:
The first thing I do after a crop/rotate (or the first thing I do at all if a photo doesn’t need a crop or rotate) is to create a duplicate layer. There are multiple ways to do this. There are multiple ways to do many things in Gimp – things can be in more than one menu, for one, but more importantly, there are shortcut keys. Instead of me telling you each one (Some I use often, some I don’t, and I’m really not sure why or why not), you can see them as you look through the menu options, or, if you go to Edit –> Keyboard Shortcuts, you can see what all is defined, and edit them if you’d like.
Back to the Duplicate Layer. I’ll create it by right clicking on the layer in the Layer dialog and selecting Duplicate Layer.
You may wonder why I create the layer to begin with. I prefer to do this for two reasons. 1) I always have my original image to back to in case I screw up so much that I want to start completely over. It’s quick and easy to just remove all the other layers and have my original back again. 2) I like to compare my edits to the original, and it’s quick and easy to turn layers on and off in order to do so.
Now that I have my duplicate layer, the next thing I do to almost every photo is adjust the levels by going to Colors –> Levels. You can read more about levels here, but the key here is that I wanted to make my darks darker and my lights lighter (thus increasing the contrast in the photo), so I moved the sliders until I got about here:
That made my photo look like this:
The next thing I want to fix is to remove the pieces of trash (or leaves?) near their feet. I use the clone tool for this:
With the clone tool, you select an area of the photo that you want to use to paint over another area of the photo. In this case, I want to paint over the trash (or leaves…) on the sidewalk with another piece of the sidewalk as my brush. To select an area, hold down the ctrl key while clicking (at this point, I should point out that I use Windows. I don’t know if the commands are different on other OSes). When you move your mouse away, you’ll see that the spot you selected is still selected – that’s just to remind you that when you start painting, that’s what it will use as a reference. This is what my screen looked like after selecting a spot to paint from (Note: I zoomed in a bunch before taking this screenshot):
Now you can start painting. You’ll notice as you do this, your reference point moves as well, to show you that as you move to the left, for example, it will pick up parts of the photo to the left of where you started from. Pretty neat. After removing the trash (leaf) it looked like this:
I also removed the piece of trash (leaf) from between their feet and now the photo looks like this:
More About Layers (and Blending Modes)
Early on in this whole process, I talked about creating a Duplicate Layer, so we’d always have the original to go back to. Layers can be a bit confusing before if you’ve never used them before. Think of the photo on your screen as a piece of paper. Let’s say you wanted try painting on that photo – adding some devil horns to our friend Bryan in this photo, for example. But you didn’t want to ruin the original photo. What would you do? you’d create another copy of the photo, and perhaps lay it right on top of the original, and then draw on the devil horns.
That’s what layers are, except, through the magic of computers, there’s more you can do with them than just create additional photos to draw on. The most obvious is the layer opacity. So imagine you placed a red piece of paper on your photo. You wouldn’t be able to see the photo below it, but if you could change that red piece of paper so it was semi-transparent, you would just have a red tinted photo. You can do that in Gimp by creating a new layer, filling it with red, and then decreasing the opacity.
None of those examples are very practical with photo editing, so let’s take a look at a few things I actually do sometimes with the photos!
Right now, we have two layers – our original unedited (outside of a rotate/crop) photo and the one with the levels and cloning. Let’s create a duplicate layer of the leveled/cloned layer (note: most people would tell you to label all your layers so you know exactly what they are. I completely agree. Still, I rarely do it, so this is probably one of those “do as I say, not as a I do” moments), so now my layers dialog looks like this:
The circled part is what I’m about to change. It’s a the Blending Mode of the layer. This is where the layers of paper analogy falls apart, but, you can change the way a layer affects the layer above it. For a list of all the different blending modes and what they do, you can read this, but for now I’m just going to show you two useful ones. The first is Screen. Screen mode is a very quick and easy way to lighten up photos that are underexposed. the photo we’re looking at now is not one we’d want to use Screen mode on, but here’s what it looks like if I change the Blending Mode of our top layer to Screen:
As previously mentioned, I wouldn’t do this to this photo, because, as you can see, it lightens it a bit too much. However, you can decrease the impact by decreasing the opacity:
And that results in this photo:
In my opinion, it’s still overexposed, so I won’t keep that. Let’s switch the blending mode to Overlay. This will add some more contrast and “punch”. Here’s the photo after the blending mode switch:
(Note: I also switched the opacity to just about 40)
Another way you can use layers is to change the color or texture of your photo by adding a colored or texture as a layer.
Let’s start out with this colored layer to warm the photo up. There are, of course, other ways to warm up a photo, but I often like to use this photo…
because it adds a little bit of randomness with the pattern and the edges are darker than the middle, so it adds a bit of a “natural” vignette as well.
(in case you’re curious, the photo is of some confetti on my floor, taken out of focus on purpose)
The easiest way to add a photo as a new layer is to drag and drop it from the file system onto your photo. It now shows up as an additional layer on my photo:
As you can see I’ve changed the blending mode to Overlay and the opacity to about 70. This is what my photo looks like after that:
As you can see, it warmed up the photo, darkened the corners just slightly, and also (you can kinda see this in the door) it adds a bit of “splotchiness”. You wouldn’t notice this if I didn’t tell you (and perhaps you don’t notice it even when I do), but I like the really subtleness of it (I’m a big fan of subtle).
Ok. Now let’s add a layer that’s actually a texture. Here’s what I’m going to use:
That’s actually a photo of the old tile floor that used to be in the entryway of my house. Yup. I know this talk is about photo editing, but I can’t help but tell you to look at some everyday objects around you and see how they can be used as textures. Another one of my favorite texture layers I use is photo of fresh snowflakes on a frozen lake. Ok, sorry for that aside. Let’s drag this image onto the photo and once again set the blending mode to Overlay. This time I’m going to keep the opacity at 100. Here’s the result:
As you can see it adds just a bit of grunge to the photo. If you want to make it MORE grungy, set the blending mode to Burn and you’ll get a photo like this (I also set the opacity to about 45 here):
I prefer the overlay version. However, I don’t like seeing the texture on the couple. I’d prefer the texture to only be on the background photo. And that’s where Layer Masks come in.
(Before I get to that, here’s a bonus hint: If you want to make your texture a bit more obvious without going the Burn route, you can always adjust the levels on your texture layer, too! Remember, this will make your darks darker and your lights lighter, adding more contrast to the layer, making the texture a bit more pronounced)
To go back to my layers of paper analogy, a Layer mask is like giving you the abililty to make just parts of the top piece of paper either partially transparant or just completely cut it out.
Let’s add a layer mask to the texture layer by right clicking on it and selecting Add Layer Mask. You’ll get a pop-up asking what you want the layer mask to start off as. I’m not sure I’ve ever picked anything other than white, so I’ll pick that now, but this will give you a glimpse into how layer masks are used. Anywhere that your layer mask is white, the layer that mask is on will be completely opaque – so you don’t see any of the layer below. Anywhere that the layer mask is black, that layer is basically erased and you see the layer below. This also means you can make part of the layer mask gray and the layer will be semi-transparent in those spots.
So. I create a layer mask that is white and my photo looks exactly the same.
Now I’m going to paint some black on that layer mask where the two people are. This means that layer that has the mask (the texture layer) will be erased where I paint the black, thus removing the texture from our two friends here. So I select the brush tool and select a brush that’s a little soft (so there’s no harsh lines where it’s obvious I removed the texture), decrease the opacity, so I’m not completely “cutting out” the layer, just decreasing the opacity (once again, so it’s not so harsh and obvious that I’m removing the texture) and bump the size up a bit:
Then I simply paint over the people on the photo. As you paint, you’ll see the texture removed from the couple. One of the great things about the layer masks, instead of just actually erasing, is that if you mess up (like if you sneeze and you accidentally erase a giant line), you can always paint white back over and your texture layer is back again! Anyway. After you’re done painting, you’ll see what you painted there on your layer mask:
And here’s the photo with the texture masked out:
Once I’ve got a photo that I think is my final edit, I like to do a quick compare to the original. Since I have all these layers stacked on top of each other like this, there’s no quick way for me to all these layers invisible at once for me to quickly compare to the original photo.
A new feature in Gimp are Layer Groups. You can create one by clicking the little folder icon at the bottom of all your layers:
Now drag all the layers except the original into that layer group:
Now you can toggle the visible of all of the layers at once by clicking the little eye next to the entire layer group so you can compare the original with all your edits at once.
You can also adjust the opacity and blending mode of the entire group all at once, too!
I used to have to do all this by creating a “new layer from visible”, then I could make all the other layers invisible. This is easier (in my opinion). I like these new Layer Groups!
Is there anything like Photoshop Actions in Gimp?
Yes! Something like that, at least.
Gimp has something called Scripts. You can download them and run them and it will automatically perform a bunch of stuff and add effects to your photos like cross-processing, vintage, duotone, split-tone, lomo, and more.
What you can’t do (and you can in Photoshop) is easily create your own Actions. You can create your own scripts, but you need to know the scripting language that they are written in, so it’s not like pressing a “record” button and performing what you want in the script.
Once you have it installed, you’ll see there’s a whole selection of scripts just for editing photos (I go into detail on all of those in the link above, so I won’t go into it here), so if you’re only going to install one script, I highly recommend that one.
Converting to Black & White
You may wonder why it took me this long to talk about something so simple, after all, right there in the Color menu is “Desatuate”!
Well. That is true, but that’s my least favorite way of converting to black and white.
First off, these methods will change the current layer to black and white, so if you want to change your “final product” that you seen on the screen, and not just the current layer, I suggest you go to Layer –> New from Visible. This will create a new layer that is a flattened image of all the layers that are currently visible.
You’ve perhaps heard of using a channel mixer to convert to black & white. The idea here is that every photo is made up of red, blue, and green. and you can desaturate based on these values. For instance, to desaturate based on red alone, you’d be saying you wanted to make any put that was pure red to be completely black and any point that had no red in it to be completely white. Same idea goes for the blue and green channels. You could use a channel mixer to say you wanted to base it 75% on red and 25% on green. You can do that in Gimp by going to Colors –> Components –> Channel Mixer:
You can play around with the sliders and see a very small preview of your new image. I like that you have more control over the default “desaturate” but I find it’s hard to know what exactly I’m looking for because my brain can’t really grasp what each channel will look like individually so it’s hard to know what percentages to use, so I prefer to use a script that comes with FX-Foundry.
Under the FX-Foundry menu go to Toolbox –> Lasm’s Channel Extract. Here you can select a whole bunch of different channels – more than just red, green and blue to desaturate on. In the screenshot you can see which ones I typically use, but I recommend trying all the channels at some point to get a feel for them all (with large images, this can take a while to process which is why I tend to stick to the “tried and true” ones):
You’ll see this creates a bunch of new layers for you, that are all black & white, for the channels you selected:
Now it’s just a matter of turning the visibility on and off of each layer to decide which one you like as a good starting point.
If you happen to like both Red and Green, and want to do the 75% Red and 25% Green, all you have to do is make sure your Red layer is directly above the Green one and set the opacity of the Red layer to 75:
Which, btw, makes this image look like so:
Or, don’t forget, you can still adjust the levels. So, let’s say, for instance, you liked the L channel, which looks like this:
But you want it to have more contrast. Just adjust the levels:
And now it looks like this:
What to Try Next
Here are some suggested things to try based upon some of the stuff we’ve talked about today:
- Now that you know about blending modes, and you have a black & white layer you like, see what happens when you adjust the blending mode and opacity of the black and white layer over your color layer
- There’s another script I like using to create Vignettes (making your photo darker around the edges). You can read about that here
- Some of the stuff in these two posts were covered in today’s talk, but here are some quick Gimp tips and tricks: part one and part two.
- When editing photos of people (especially kids), you sometimes want to make their eyes “pop”. Here are a couple ways to do that in Gimp.
- If you just want more examples of how I’ve edited some photos, check out: one photo, four edits, how I did it: the hand photo, how I did it: the vines on the wall, how I did it: the boots photo.
I hope that helps! I really appreciate getting the opportunity to talk to this group, and you can contact me at jennifer (at) iffles (dot) com with any questions!