Photo of the day: HDR edition

Image

This is a pseudo-HDR photo. I really don’t know much about HDR photography at all, but I figured I would give it a shot. I used just one image taken with my Nikon D3000 in RAW mode. I set the exposure for the lighter sky, because details can be brought out of the dark areas in RAW photos. I opened the image up with GIMP/UFRaw four separate times. Each time I changed the EV in UFRaw to be two steps apart and saved it as a new image. I had -2, 0, 2 and 4. Then I ran Qtpfsgui (worst name for a program ever) on the four images and used the Mantiuk method on them. This is the result. The landscape is not as bright as I would like, and there are some strange things going on in the clouds that I’m not really happy with, but overall I think it was an interesting exercise.

Blotch Brushes for Gimp

Link

Blotch Brushes by ~Aless1984 on deviantART.

Tilt-Shift Simulation: The Movie

Posted on September 21, 2011 by Daniel Veazey

Download Video: Ogg

Grass Brush for GIMP SET 1 by *FrostBo on deviantART

Link

Grass Brush for GIMP SET 1 by *FrostBo on deviantART.

Gimp summarizes Google Summer of Code projects

Link

Google Summer of Code 2011 is over, and all of our students have successfully finished their respective projects. All of the code is available in Git repositories, but not all of it is going to be part of the next released versions of GEGL and GIMP. Here is why.

via GIMP – The GNU Image Manipulation Program.

Cookies

Image

Everything You Need to Know About Gimp Layer Masks

Featured

Posted on January 31, 2011 by Daniel Veazey

One of the most popular posts on my blog is my Gimp Layer Mask tutorial. I feel like that post doesn’t fully explain everything you need to know about layer masks, so I’m going to try to break it down here, and hopefully by the end of this post, you will be a layer mask master.

What is a layer mask?

A layer mask is pretty much exactly what it sounds like. It puts a mask over a layer of your gimp image and you manipulate it to hide parts of the layer while showing other parts. When you apply a layer mask to an image, the mask itself deals with only black and white, and shades of gray in between. White areas of the mask make those corresponding areas of the layer opaque, and black areas of the mask make the corresponding areas of the layer transparent. Shades of gray make the layer partially transparent. The darker the shade of gray, the more transparent that part of the layer becomes, and the more of the layer below it shows through. Let’s look at an example:

In Gimp, open two photos by choosing “Open As Layers” from the File menu. As you can see on the right in the layers palette, each photo is on a separate layer.

Next, right-click on the top layer and Add Layer Mask. A window pops up that says “Initialize Layer Mask to.” Choose White (full opacity).

A white rectangle appears next to the thumbnail image in the layers palette:

It’s hard to tell, but this white rectangle also has a white border. The white border means that anything you do right now will be applied to the layer mask, and not to the layer itself. Try clicking on the thumbnail of the image on that layer and you’ll see the white border move over the thumbnail. Now any changes you make will be applied directly to that layer instead of the layer mask. Now there are lots of ways you can manipulate the layer mask. you can draw on it with the paintbrush, type over it with the text tool, make a gradient over it, etc. Any black that you add to the layer mask now will make that area of the layer transparent, and you’ll be able to see the layer below it in those areas. I guess I should make it clear that when you make changes here, you’ll be doing it in the main image area, not on the thumbnail of the layer mask. I think you already knew that, but I wanted to make sure. Let’s try drawing with a large, hard-edge brush and see what happens.

Where I drew on the image with black, it masked the area, allowing the layer below to show through. I might sound repetitive, but this is the main point of this post, and if you understand it, then you can do just about anything with a layer mask. In the layers palette, the thumbnail of the layer mask now has a black streak across it where I drew on the image. Now let’s look at how shades of gray affect the layer mask.

Here I made a gradient on the right side of the image, with white on the left end and black on the right end of the gradient. You can see this in the layers palette. In the image, the layer fades out from left to right, corresponding to the gradient from white to black. One more thing, I made a circle in the lower left part of the image with a soft-edge brush, so the mask fades at the edges of the circle.

That is really all you need to know to get started with layer masks. I hope this helps. If you have any questions, feel free to drop me a line on the contact page.

Using High Pass Filter To Sharpen Images In Gimp

Featured

Posted on January 30, 2011 by Daniel Veazey

Some people are familiar with using Unsharp Mask to sharpen images in Gimp. But there is another method, which I think does a better job. It’s the High Pass Filter. Gimp doesn’t come with a High Pass Filter by default (as of version 2.6), but there is a plugin available on Rob A’s (Im)personal Blog.

Most of the time, sharpening is the last step you want to do when working on an image. If you apply other filters to a photo after you’ve sharpened it, those filters tend to affect the sharpened areas and make it look wrong.

The plugin comes with five modes you can use:

Colour
Preserve DC
Greyscale
Grescale Apply Chroma
Redrobes

Let’s compare the different modes, and Unsharp Mask. Click on any photo to see a larger version. I have to warn you, though, they are larger-than-normal files, but I wanted you to be able to see the difference between photos.

No sharpening

The photo has slightly fuzzy edges throughout.

This photo is slightly out of focus. The trees in the background are kind of foggy. The outline of the elk is a little fuzzy. The grass tends to run together. A haze hangs over the whole thing.

Unsharp Mask

Unsharp Mask applied

With Unsharp Mask, the photo has definitely improved. The trees look more defined, as well as the outline of the elk. There is a slight improvement in the sharpness of the grass and it seems like the haze has been lifted. But there are now blown-out highlights on things now. Particularly the elk’s antlers and branches and leaves that were in areas without much contrast previously. Unsharp Mask increases the contrast to sometimes unbearable levels, leaving halos around edges. Now let’s look at some high pass filter techniques.

High Pass Filter — Colour

High Pass Colour applied

This is a much gentler method of sharpening the photo. All the blown-out highlights and halos are gone, but the sharpness of the trees has increased, the outline of the elk is much clearer and the grass has more texture to it. Here is how I used it:

Click on Filters → Generic → High Pass Filter. A dialogue box appears with some options. This is where you choose which mode you want. The default is Colour. I didn’t adjust Filter Radius or Contrast Adjust from their default settings of 10 and 0.
Click OK and the filter runs its course. It creates a new layer that appears mostly grey with outlines of the image.
In the Layers palette, change the mode from Normal to either Overlay, Hard Light or Soft Light. I chose Overlay. The sharpened image appears.

I think the Colour mode will be suitable for most situations, but if you want to experiment and have a little fun, there are four other modes you can choose. Let’s look at those now.

High Pass Filter — Preserve DC

Preserve DC mode

The Preserve DC mode functions similarly to the Colour mode, but also adds the average image color back in, so in this instance the photo becomes much greener.

High Pass Filter — Greyscale

Greyscale Mode

The Greyscale mode desaturates a copy of the image before applying the High Pass Filter. With this image, I cannot see any difference between this mode and Colour mode.

High Pass Filter — Greyscale, Apply Chroma

Greyscale, Apply Chroma mode

This mode also uses a desaturated copy of the image, but blends it with the source layer colors. The result is an image with very saturated colors. It borders on looking unrealistic.

High Pass Filter — Redrobes

Redrobes mode

This mode is similar to the Preserve DC mode. However, while flipping back and forth between the two images, I can see that Redrobes produces a slightly sharper image than Preserve DC.

So those are the different methods of High Pass Filter in Gimp. Colour is probably the mode I’ll use most when simply touching up photos, but Preserve DC, Greyscale Apply Chroma and Redrobes would be fun to play around with.

I’d like to hear from you about other methods of sharpening photos.

The best way to make a vignette in Gimp

Posted on December 18, 2010 by Daniel Veazey

Vignettes look cool. There are several ways to create the effect in Gimp. Here is an easy method to make an aesthetically pleasing vignette. You can click on the screenshots in each step to see them larger.

Step 1

Duplicate the background layer on the layers palette.

Step 2

Change the mode of the new layer to Multiply.

Step 3

Right click on the layer and add a layer mask. Choose white (full opacity) for the fill option on the pop-up dialogue.

Step 4

Select the blend tool from the tools palette.

Step 5

Change the shape of the blend to Radial and change the offset to about 30. Changing the offset makes the radial gradient weighted so the change from light to dark takes place closer to the edge of the gradient instead of evenly from center to edge.

Step 6

Click and drag from where you want the lightest part of the photo to be toward the edge of the photo.

Step 7

Admire your vignette.

Gimp Layer Mask

Posted on August 1, 2010 by Daniel Veazey

NOTICE: I have posted a new tutorial explaining the basics of layer masks. Check it out here.

Here is a way to use layer masks and the paths tool to merge parts of two photos into one composite image. We’ll start by taking two photos of a television, using a tripod to keep the position of the camera exactly the same. We won’t use a flash, so the first photo will have a slower shutter speed to get the correct exposure on everything around the TV. However, this makes the image showing on the TV way overexposed and blown out. So we’ll take a second photo with a much faster shutter speed to get the correct exposure on the screen.

Continue reading →

Photo roundup

Posted on May 26, 2010 by Daniel Veazey

I got a few good pictures today. With the airplane I tried adapting a Photoshop technique in Gimp to give it the look of a polarizing filter. I think it turned out pretty decent. I’ll try to write a tutorial for the method soon. I was impressed with the bee that had all that bright yellow pollen all over it. I might make a print of that. The monarch butterfly wouldn’t sit still for more than a few seconds. That’s the only shot I got of it. I got the picture of the rabbit while I was waiting around for airplanes to fly over. I was standing right next to it for a long time before I realized it was there. On the red flower, I just wanted to see how it would look if I cropped out part of it. The bumble bee came out pretty sharp, so I included it too. I hope you enjoy these. Please leave a comment and let me know what you think.

How to read a histogram

Featured

Posted on May 19, 2010 by Daniel Veazey

When you’re out there taking pictures, how can you tell if you have a well-exposed shot? One might look at their pictures on their digital camera’s display, but that’s not going to be very accurate because it’s not calibrated to display the picture as it truly appears. If you’re out in bright sunlight, and you have the brightness on the screen turned up, there’s no telling.

But there is a solution. The wonderful folks at the camera companies make it possible (on a lot of models, but not all) to view a histogram of your photo right after taking it. But what good is a histogram if you don’t know how to read it? Let’s have a look at one and figure it out. Here is the histogram that appears when you adjust curves in Gimp. The histogram on a digital camera should look very similar.

Correct exposure

A histogram is basically just a bar graph. The horizontal axis of a histogram represents the lightness of pixels. On the left is pure black, represented by the value 0. On the right is pure white, with a value of 255. Neutral gray is smack dab in the middle. The vertical axis of a histogram represents the number of pixels of a particular lightness in the photo. In this example, we see that on the left side of the histogram, there are no pixels in the photo that are absolutely black. But just a little bit to the right of it, you see the bar graph go up, meaning there are several pixels that have a lightness value that is very close to black. Then the graph dips down just a little bit and goes way up when you get into the darker part of the middle ranges. It dips back down in the higher part of the middle ranges and then swoops way back up again to the top as you get close to white, and then drops off completely right before you get to absolute white. The exposure on this photo is pretty close to correct. There is a wide range of lightness values, but they don’t go off the chart on either side. Also with this photo, because of the U-shaped dip in the middle, it makes for a higher contrast. Because we have many pixels in the darker range, and many pixels in the lighter range, and fewer pixels in the middle range, the photo has more contrast.

Overexposure

If the bar graph were to go all the way to the right edge of the histogram, it would mean that the photo is overexposed, and there would be areas that are blown out completely white, and there’s not much you can do to fix an overexposed digital photo. Here is what an overexposed photo’s histogram looks like:

You can see there are almost no pixels showing up on the left side of the histogram, meaning there are almost no pixels in the photo that are black or close to black. Over on the right hand side, you see that the graph starts curving up at the end and goes all the way to the right side of the histogram. That means it’s overexposed and you can see in the photo itself that the grass is largely blown out, along with the white birds.

Underexposure

If the bar graph is crowded over to the left side of the histogram, it generally means your photo is underexposed. Here is an example:

You can see there are almost no pixels showing up on the lighter end of the histogram, and everything in the picture is very dark. But there is still hope if you have an underexposed digital photo. Often it can be saved (using tools like curves in Gimp), especially if you’re shooting in RAW format, but we’ll go over that another time.

What to do

The best practice is to correctly expose your photo when you take it. After taking a photo, look at the histogram and figure out if you’re overexposed or underexposed. If you are, make necessary adjustments on your camera and try again. Try to get a photo where the histogram stretches from one side to the other, but comes down to the bottom on each end (like the first example above). Don’t worry about running out of film.

There also are other situations where getting correct exposure for the entire photograph is just about impossible, and I’ll touch on those later, too. What are your thoughts? Leave a comment.