Google Earth now has entries for GigaPan. This is an amazing way to experience some places within Google Earth, but you’ll need v4.2 or later to see it.

In short, by placing a panoramic at the correct coordinates and specifying field of view, elevation, tilt, etc. it becomes possible to fly in to the image and look around it in detail. I decided to upload my night (hdr) panorama of Zurich but it wouldn’t let me — it wasn’t big enough! The image has to be at least 50 mega-pixels to be accepted. My original met this requirement but, as you may recall, I’d had to cut the image in 1/2 both horizontally and vertically in order to load it in to the HDR processing program. I could, of course, simply scale the image up but that would be cheating. So, I went back to the originals to try some new techniques.

There are two paths to follow… (1) Merge the images into an HDR with Photoshop and then do the tone-mapping with EasyHDR. Since I’d be loading just one 32-bit TIFF image, hopefully it would stay within EasyHDRs memory limitations. (2) Do HDR processing on each stack of images first and stitch those together into a panorama. This requires that the same transformation be applied to each stack in exactly the same manner or else there will be seams in the final image.

1) Merge into HDR with Photoshop

By restoring all the saved, aligned images I had made during my previous attempt, I had a good starting point. I recreated the three panoramas using PTGui and then loaded them all into Photoshop using “File::Automate::Merge to HDR”. The tone-mapping in Photoshop CS2 is poor compared to other alternatives, so at this point I saved it as a 32-bit TIFF.

Unfortunately, in the end I was unable to load even an image with 1/2 resolution in to EasyHDR. Windows programs that don’t do some sort of tiling of data (like Photoshop does) are generally limited to 2GB of memory. On to the next method…

2) Generate Multiple HDR Images and Stitch Them

The latest version of EasyHDR has some nice new features over what I used just 8 months ago. The trick here seemed to be to avoid anything that was dependent on the local image. To this end, I turned off the local mapping “mask” (which I don’t like anyway because it produces halos) and leave the general tone-mapping parameters at 1.0. I also never adjusted the black/white clip points, leaving them at the far ends of the spectrum. This would hopefully result in identical mapping for all image stacks and by saving in 16-bit mode there would be sufficient detail for me to adjust the total range in post-processing.

Before stitching, apply any filters that apply to a given image stack — noise-reduction, for example. Also, it’s likely that a lot of third-party software will not be able to handle gigapixel size images. You’ll have to run that processing on each part before stitching.

Other Things

Along with these changes, I switched the panorama generation to be “cylindrical” instead of “rectilinear” as Earth will do all the perspective alterations necessary. If you’re not familiar with the terms, the latter is the standard image that non-fisheye lenses will give you. It’s what the eye would see looking through a frame held in front of you. A cylindrical mapping, on the other hand, is what you would see if looked through a vertical slit held at arms length and then rotated your whole body, combining the all that is seen into a single image.

Here’s the final image… Click on it to browse it at full detail!  It should appear in Google Earth sometime in the future.  Look for it at 47.37593N, 8.54651E.

For the most part, it isn’t the eye that “sees” color. It’s the brain. Perhaps that it is why fantastic color is so pleasing. And when it comes to fantastic color, there is just no competing with Mother Nature.

Of course, as soon as you put a frame around it, the image loses something. I remember this moment quite well — I was sitting on an island beach with my new wife when I triggered the shutter. It was probably the most beautiful sunset of our two weeks in Greece, including those we saw in Santorini which is known for fabulous sunsets. I believe this image to be beautiful, but I’m also somewhat biased as I see it as I did then… without borders and my new love at my side.

If color stimulates the brain directly, what about something that stimulates in indirectly? Imagination, for example. It’s this that creates vivid dreams and makes appealing things like naughty on-line chat rooms. Imagination is often better than reality, and why not? There’s no limits to our imagination. As long as one doesn’t live there, it’s a great place for frequent visits.

Look at this image closely. How do you know that the walls are white, the stains are brown, the grass is green, and the little stripe at the top of the hull is blue? Your imagination fills in all the details. Unlike the color image above where you take it in all at once, this one draws the eye all around. Every place you look, you see more that what is there. There is color, but it’s like an overlay; it’s not there and yet you see it anyway.

Go on… Click on the image to see a larger version. Take a look around.

Now I’ll admit, I cheated. The image of the boat isn’t truly colorless. It’s 90% desaturated so you can’t actually see the color, but there is enough present to hint to the mind what it should be. Perhaps a child wouldn’t need it, but I find the imaginations of us adults sometimes need just a little push to get going. Could it be that we don’t use it enough?

No, I need to retire because I have too many hobbies and not enough time for any of them. As you might guess, Photography is one of them and, as you might also guess, it has slipped a bit in recent months. What with starting a new job, moving from Canada to Switzerland, and adding a new little girl to the family, I just don’t seem to have any time for photo work, video editing, programming, writing, or Christmas shopping. Sorry to throw that last one in there so early, but when you have to ship across the Atlantic Ocean, you have to start early.

I thought I’d write this article to briefly touch on some of the other things I like to do. Perhaps some other photo buff has a similar interest

Programming: I’ve written so much code over the last 25 years that I just can’t keep track of it all. Most of it just things I do for myself, touch I have contibuted some things to the open-source projects like Debian.

Travel: I moved to Zurich, Switzerland to take this job. I could have gone to any one of a number of locations around the world, but this seemed like a great chance to do some travelling. Switzerland is located pretty much in the center of everything. So far most of the destinations have been within the country but we have made a few trips to France and Italy. One day I’ll get around to processing the photos from those trips and posting them in my gallery.

Family: I’ve got a beautiful wife, a 3-year old son, and an 8-month old daughter (as of November, 2007). This is really the “hobby” that makes it so difficult to find time for all the others. People who don’t have kids really don’t understand just how much time they take and how much freedom you give up. Your life doesn’t change. Your old life ends and a new one begins. The change is that dramatic. I don’t regret it in the slighest — my kids are the most wonderful parts of my life — but the cost is significant.

Reading: Robert Jordan, Stephen King, Orson Scott Card.

Business: I’ve started or helped start three companies. One failed before it really got started, one ran for a few years, and the last did quite well (though I left to take my current job). I write a bit about it, too.

Home Renovations: I’m an engineer and though I could never get my mechano set to to take on the plans I had in my head, I’ve found that I have a pretty good knack for making things fit together when I’m the one cutting all the pieces. So far I’ve built a new kitchen (including moving 3 walls), inserted a mud-room in our garage, and added a deck on the back of the house. Each of these projects has taken several months from start to finish and I always lose 5-15 lbs over their duration.

Video Editing: Raw video is the most boring thing to watch in the world. Every time I’m doing some photography and I see a tourist with a video camera pointed at whatever-is-there, I cringe. When I see that person hold the shot for 10 seconds and/or slowly pan over the scene, I know that it’s likely some poor souls are going to be tortured with, “Want to see some shots from my vacation?”

Every year I take the collected footage from my video camera and make a DVD to send to family as a Christmas present. I set it to music, change between scenes fairly rapidly, and condense 12 months of video down to about 15 minutes. That’s a pretty good ratio.

I also enjoy time-lapse video in conjunction with my previous hobby of home renovations. Check this out… Four months of work in four minutes.

It’s a panoramic HDR image taken at night, which doesn’t sound so difficult. I certainly didn’t think it would be very difficult while I was taking the 33 shots that went in to it. The problem is that you can’t create the HDRs individually and then stitch them together because each combined section will be anywhere from subtlely to hugely different than the others. In the end, I had to…

1. load each set of 3 images as layers in photoshop
2. align them
3. crop the image to the minimum area covered by all layers
4. save the layered file
5. save the brightest layer as “a.tif”
6. repeat steps 1-5 for all image groups, output to b.tif, c.tif, … k.tif
7. load images a.tif … k.tif into PTGui
8. create points, optimize, adjust, and generate a panorama: x.tif
9. save PTGui project
10. open all the PSD files from step #4, this time saving the middle layers as images a.tif … k.tif
11. load PTGui project
12. touch nothing — generate again, saving as y.tif
13. repeat steps 10 to 12, saving the darkest layers and generating the panorama z.tif
14. load x.tif, y.tif, and z.tif in to whatever HDR program you use
15. watch it crash from lack of memory
16. load x.tif, y.tif, and z.tif in to Photoshop and shink them
17. load x.tif, y.tif, and z.tif in to whatever HDR program you use and go (remember that they’re already aligned)

See… Easy! Of course, the pano was too large for EasyHDR to handle and Photoshop (CS2, anyway) does a terrible job in comparison. I had to scale x, y, and z by 50% in each direction before it could load and process it.

Of course, all this was done for the same reason you climb a mountain: Because it’s there. The image is blurry if you look close-up in the original (not posted here) because it was so windy the night I took the photos that I could not keep the camera perfectly still for the long exposures. C’est la vie!

Photographing architecture is difficult unless you have a shift/tilt lens because a photographer usually has to point the camera upward at an angle to capture the entire building. Unfortunately, this causes the normally parallel vertical lines to tilt inward toward the center of the photo.

Photoshop offers a “distort” filter called “Lens Correction…” from which you can supposedly fix this. Unfortunately, it doesn’t work correctly and here’s why:

The two photos above are of the same coaster and CD. The one on the left is taken from directly above while the one on the right is taken at a 45° angle. It’s easy to see how the angle causes compression of the vertical and makes the lines that make up the sides of the coaster converge. After applying a correction to the latter you would expect it to look more like the original, but it does not — as the following photo illustrates.

Photoshop is doing a correct forward transform (as though tilting a piece of paper away from you) but since it’s supposed to be a corrective action, it should be doing a reverse transform (taking a tilted piece of paper and making it straight). Working around this problem is not trivial but the only practical difference is the vertical scale. In short, Photoshop reduces the vertical size by a factor x when it should increase it by that factor. Thus, if you know what x is then you can increase the vertical size by x-squared so that after the incorrect correction it will be larger by x rather than smaller by x.

To do a correct correction, first do the correction as you would normally, recording the values used because you’ll need them again later. Once the change has been applied, use the following equation to calculate y_new (the last form is easiest).

y_new is the new canvas height, y_old is the old/original canvas height, and y_corrected is the height of the image after the correction was applied (you’ll have to measure this by hand). Undo the correction, go in to Image::ImageSize, uncheck “constrain proportions”, and set the height to be y_new, scaling the image using the best algorithm (”bicubic”). Once that is done, run the lens correction again with the same parameters you recorded above and you’ll end up with a properly corrected image, as shown here:

The unfortunate part of this technique is that it does two imprecise scaling operations instead of just one, but in general this is not a problem. The better way is to not use the lens correction filter at all but rather just perform a “free” transform that gives the same result, but even that may involve multiple operations if you have to alter rotation or barrel/pincushion correction.

I’m curious to know if this same problem also exists in CS3.

There are two important elements of photography present at night that most people don’t expect: light and color. It is, after all, dark and “dark” is defined as the absence of light, right? Well unless you’re in a mine tunnel underneath the ground, there is always light. It may come from street lamps, or the moon, or even the stars, and it may not be bright enough for a human eye to see, but it is there and a camera can see it.

The same goes for color. The human eye “sensor” is made up of two types of cell: rods and cones. The sensors that are most sensitive to light, the rods, are not sensitive to color and so at night everything looks somewhat black & white. A camera sensor is equally sensitive to color in low light as it is in high light. You can get some colorful photographs at night that are especially startling because they did not appear colorful while taking the (low light) picture.

Night often has one other element important to photography: contrast. This is both good and bad. It’s good because contrast is interesting. It’s bad because it frequently exceeds the dynamic range of your camera and thus either has washed-out highlights or large areas of total blackness. Film is better than digital sensors in this regard because it over-exposes gracefully, leaving highlights that have more detail and fewer color artifacts around them.

Ah, but digital has the power of computer post-processing… if you can capture the information in the first place. To do this, you’re going to need these things:

1. DSLR: A point-and-shoot just isn’t going to cut it here. You need a camera with very low noise (i.e. the ability to go to ISO1600 or higher) and that means the physically big sensor that only DSLRs have. You also want a self-timer and “mirror lock-up” (sometimes called “exposure delay”).

2. Tripod: There is no way to hand-hold decent night exposures. You must have something that can lock all axi of movement and hold your camera completely steady for several minutes, even with you fiddling controls.

3. Adobe Photoshop: I’m sure there are other software packages that can do HDR, but I don’t know many of them off-hand. Gimp will not support this until v3 and that’s still a long time away. I’ve used EasyHDR, an inexpensive product that produces excellent results, but isn’t as flexible as Photoshop (surprise, surprise). Please leave a comment if you want to describe some other HDR-capable program.

Set up your tripod and frame the image you want. Set everything to manual, including white balance. I frequently use F8 to get a good depth-of-field. Focus and then lock the focus by switching from “auto-focus” to “manual”. Set your ISO sensitivity down as low as it can go (i.e. ISO50 or ISO100). You want to make sure there is as little noise in the images as possible and this is the best way to accomplish that. Noise is typically random and so over an extended exposure it will be reasonably uniform, like a light haze, easily reduced or removed simply by shifting the black point slightly. A high ISO setting will instead amplify individual errors and make the image look “grainy”.

Calculate the manual exposure of the darkest area you want clearly visible. You can do this by using a hand-held meter, or the spot-meter function of your camera, or you can just guess until you get it right (my preference). The shutter time will be quite long, probably on the order of 15 to 30 seconds if you’re using F8 like I do.

Program your camera with a selfdg-timer of 2 seconds or more and to have some delay between lifting the mirror and starting the exposure. The self-timer gives the camera time to stabilize on the tripod after you depress the shutter release and the exposure delay lets it stabilize again from vibrations caused by lifting the mirror. Also, if you have a “vibration reduction” or “image stabilization” lens, turn that feature off. These technologies can get confused by the steadiness of a tripod.

Take your first picture and verify the result. The dark areas you’re interested in should be well illuminated and the light areas will be flashing because they are over-exposed.

Now reduce the shutter time by a factor somewhere around 1/4 to 1/16 (3-4 stops) and take another picture. Don’t change any other setting! Repeat this until there are no over-exposed areas, usually 2-3 times. You have now captured the complete dynamic range of the scene, albeit over several exposures.

Bring these images to your computer and start Photoshop. Go to the “file” menu and select “automate”, “merge to HDR”. Then select all of the exposures for a particular image. I’ve used both RAW and JPEG sources for this and never been able to tell the difference. If you choose a raw file, the converter will use whatever settings you used last time, so open one directly, first, and configure everything as you like it.

Check the “try to automatically align images” box. It’ll take longer, but it’s worth it. There’s a good chance that miniscule movements occurred while changing the exposure settings and this is the easiest way to correct for that. Click “convert”. Wait. EasyHDR does not yet have this option, so you’ll have to open each file individually and manually align them.

When it’s done, you’ll get a full image preview and a slider you can move around to see the full dynamic range of the image. Changing this value only alters the display, not the image in any way. When you’re happy, click “OK”.

There isn’t much you can do with an HDR image until you convert it to a normal 8-bit or 16-bit image. You do this by going under the “image” menu and changing the “mode” to “16 bit”. At this point, you’ll get a dialog asking how to map the HDR image to a normal 16-bit image. The possibilities for his are more than I want to get in to in this article, but I recommend Sean McHugh’s article on tonal mapping with HDR images.

Because of the extended dynamic range in the converted image, most of these 16 bits, perhaps even all, will be data instead of noise. You can thus apply a “curves” adjustment layer to get the exact display you want without amplifying noise from the camera’s sensor.

Good luck! If you post some of your great night shots, drop me a comment with a link to them.

My very first (non-test) photo was of my son eating a snack. With no flash, the D80 went up to ISO1600 with f3.5 and the result was a surprisingly great picture. There is little noise, excellent exposure, and amazing clarity. I couldn’t have done this with my standard (for indoor) Fuji400 film. All hail digital! Right? Not so fast…

I then took the D80 up to a “lookout” point to capture a sunset. It took great pictures, but either the sky was blown-out or the landscape was too dark. I could lighten it with Gimp or Photoshop, but it was already ISO800 and the noise didn’t look so good. I’d never had that trouble with film. Why? Because film over-exposes better than a digital sensor. As it approaches maximum exposure, film begins to slow down. It effectively compresses a large exposure range in to the top end, and does it gently so that colors just desaturate (become less vivid but retain the same hue) to white. This is usually referred to as the “shoulder” of film.

Digital sensors, on the other hand, are linear. They just sit there counting photons (”light bullets”) up to their maximum value at which point they stop. That’s pure white or worse. I say “worse” because red, green, and blue are “counted” separately; a colored source will saturate one of the color chanels first while the others continue to increase. That creates wierd color bands around the overexposed object. Take a look at the following example. It’s a gradient of 2R+1G+0B counting from 0 to 100 and limited to a value of 100 per channel (just like a digital sensor). Note how it’s a nice smooth increase of the same hue until the mid-point, at which time red clips to the maximum value. For the right half, red is now fixed at 100 but green continues to increase, changing the color to magenta (or 1R+1G+0B equivalent).

Film also captures all three colors at every point. Digital sensors (all but Fovian, anyway) capture only one color at each point and then interpolate between them to create a proper RGB image. Film will therefore be sharper than digital if converted at the same number of megapixels. Exactly how many megapixels you can get from a film scan depends on the film being used, but 4000dpi (24 megapixel for 35mm film) is about the theoretical limit. Despite this huge descrepency between film (24MPx3) and digital (10MPx1), for practical purposes the 10MP digital sensors of today are about comparable to common, non-professional film for resolution in the final image. Many lenses cannot resolve finer than this.

Another advantage of film is that it only costs 10x as much to create a negative 10x the size. Increasing today’s digital sensor to 10x the physical size might cost 1000x as much to create, if you could do it at all. That’s why you can find 4×5 inch film cameras but no corresponding digital ones. It’s also why most DSLRs use smaller APS-C size sensors. If you want extreme resolution for billboards or super-sharp, poster-size prints from a single frame, you need big film.

What does all this mean? Over the past months, my basic discovery of digital vs film is this:

It’s easier to capture a good image with digital, but you can get a better capture with film.

Which is more important to you? Me, I’ll stick with my D80 for now and hopefully one day get around to playing with a 4×5 view camera.

You have to love your subject matter, or you’ll never capture it well enough to do it justice.

Now tell me… How many of you actually clicked on the link to the semi-famous landscape photgrapher?

What you need when bright sunlight abounds are huge, overhead, translucent diffusers. Meteorologists call these “clouds”. Yes, clouds are a photographer’s friend. They scatter the light so that there are no shadows, or at least no deep shadows. A camera can then record the entire dynamic range and produce images that look right when printed.

So don’t be afraid to go outside when it’s cloudy or threatening rain. That’s often when you’ll get your best pictures. The sky, too, can be more interesting this way than the nice, deep blue of clear day.