Today’s experiment is using the oil paint filter in Photoshop CS6:
This filter is very easy to use, with 4 sliders for brush controls and 2 for lighting controls – giving more or less detail to the strokes. It works well on this photo with simple shapes. I first tried it on a more complex scene with fairly muddy results, although that one probably would have worked out well if I had applied different settings for the oil paint filter to different parts of the image (the oil paint filter does not directly allow this, so layers and masking would have been needed).
Today’s experiment is creating a microplanet. Here’s the Statue of Liberty on top of the world (glad that Photoshop knows more about polar coordinates than I remember from school)
The technique is very easy (using Photoshop’s polar distort filter) but getting something that looks decent is tough. The big issue is finding a type of photo that works for this technique. I don’t take too many broad landscape-type photos, so it took me a while to find one of mine that could work.
The starting photo needs to be much much wider than tall – either a panorama or a photo cropped to these dimensions (which is what I used). It is best to have something where the top is plain sky, and the bottom can read correctly after it is distorted – since it ends up as the center, it gets the most distortion. And I had to experiment several times to get the proportions of the top vs. bottom right – after polarization, the Statue of Liberty and buildings came out in different height/width dimensions depending on the ratio of top (sky) to bottom (water) that the starting photo had. The other issue is the seam – the image needs to have the horizon be exactly vertical, and more or less matching areas at either end so when it wraps the seam will not be too visible. I ended up being able to get it all pretty much okay except for a visible difference in the color of the sky at the seam, which I had to smudge/blur/clone until it was less obvious.
Today I tried to photograph smoke. New to me: buying incense (I’d heard that it smoked the best), then not being able to find matches so using a lighter for the first time (how did I make it this far in life with never figuring out how to use a lighter before?)
Trying to get a great splash shot like ones I’ve seen. Well, could have done better if I had a lot more patience, and probably had I used something thicker like cream rather than mixing water, milk, red food dye, orange juice, and V-8 juice, i.e., whatever was in the refrigerator.
Also, it was one of my first times using a flash, so I was guessing on the angle and exposure. I was also guessing a bit on the focus. And I needed one more set of hands. And a better way to drop the liquid – I tried a feeble turkey baster without much success, and then moved to pouring slowly from a glass measuring cup – it would have been better if I had been able to drop the liquid from on high, and if I had waited between attempts for the frothy bubbles to disappear.
Restoring the Wye Oak
The Wye Oak, Maryland’s official tree, was about 460 years old when it died in 2002. This image is from a slide I took a number of years ago, with lots of damage – streaks, stratches, dust, mildew, etc. (hey, don’t store slides in your basement), which I now restored to useability with Photoshop’s healing tools.
I used content-aware fill, cloning, dodging, and burning. Obviously could have been better if I had spent lots more time on it, but it now at least looks like the magnificant Wye Oak of the past.
Where are the tourists?
At this time of year, Lafayette Square is loaded with tourists, but none are in this image:
The solution is to take multiple photos, each with the tourists in slightly different places, then combine them in Photoshop. Photoshop offers an automated way to do this – using “image stack”, designed for scientific purposes but working well to combine photos based on several different algorithms.
I got pretty close to the no-tourist result this way, with a few ghost areas left behind. So I refined the results manually – stacking the photos and masking out certain areas from each layer.
What would have been better for next time:
1. take more photos – I only took 6 images, and while there weren’t people standing in the same spot in each image, there were a couple of spots where one person would be there in one image and another person in the other images.
2. use a tripod. I took these shots by sitting my camera on a bench and using a cable release, but even so, when I used Photoshop to auto-align the images, it was clear that there had been camera movement between the images.
Today’s experiment is a panorama. I took 10 shots of the World War II Memorial, overlapping each one. As to how much overlap, it varied since the shots were handheld and composed by trying to stay still except to move my direction by moving my feet. I took them in vertical format so that there would be a lot of the scene left even if the merging had ragged borders top and bottom.
The process of creating the panorama in Photoshop CS6 was really easy (an automated photomerge command) but it took my computer quite a while to churn through it, even with 8GB RAM.
Today’s photo (well, actually taken last night) is attempting a silhouette – where due to the lighting and exposure, the people are almost black. Works well with simple, identifiable shapes against plain backgrounds, like the sky. This shot is from the county fair, with people on bleachers watching the tractor pull.
Today’s experiment is trying a super-high-contrast look.
The main processing was done with Topaz Adjust, with some other small changes afterwards.
Today’s experiment is trying a black and white HDR – I’ve tried a number of HDR’s in color, but this is the first time I have tried black and white.
I used HDR Efex Pro. I didn’t like any of the monochrome presets. At first, I had trouble figuring out how to make the adjustments that resulted in pleasing effects. But then when I concentrated on getting a good distribution of tones by looking at the histogram, it was easy to adjust the remaining factors.