Sunday, November 28, 2010

Night Pier, with Surf

The Surfside Beach Pier (South Carolina) stands stoically as the the tide rolls in.

Looking at it in real life, you'd probably think it looked neat, but not this epic. That's because your eyes are so much better than a camera sensor. Let's walk through the differences visible here.

1) The lights looked white in real life, not green.

Our eyes color correct a lot of light sources to white - they even correct mixed light sources, such as daylight mixed with tungsten, to white. Camera sensors can't. As far the sensor is concerned, there's 1 color temperature (expressed in degrees Kelvin or Kelvins) that's neutral, and any light source that's not that color ain't white (usually bluer, oranger, yellower or greener), despite what your eyes tell you.

Here's a color temperature chart that lays it out nicely. Note that fluorescent lights generally come with a side of green tint that isn't reflected in that chart.

Here, those lights on the pier, except for that red one at the end, are pumping out light that looks green to a camera set for daylight color balance. I could have corrected this by changing my color balance to fluorescent on the camera, or on the computer (in post), but didn't. I saw what was happening as I reviewed the photos I was taking as I moved up the beach towards the pier, and liked the effect.

2) The sea had surf with distinct waves, not the smooth, misty, mystical water you see in the photo.

Pier lights seem bright to our eyes, but to the camera sensor, they are damn dim. This is a 2 minute exposure at f/11, ISO 100. That's approximately 15 stops less light than on a sunny afternoon. In other words, those pier lights are 1/32,768th as bright as the noon sun.

Two minutes is far too long a shutter speed to stop any sort of motion. Thus, the waves disappear, their white tops recorded only as white mist clinging to the timbers of the pier. Likewise, the water goes smooth and silky.

3) In person, it doesn't seem so immense - it's big, but not the way it looks here.

The last divergence from what the eye perceives is the result of the compressed perspective offered by the 100mm lens. Since the 60D has a 1.6 crop factor, I effectively shot this with a 160mm lens. That telephoto perspective allowed me to stack the support timbers visually to create an illusion of massiveness.

Obviously, I used a tripod to get this shot. I also used a remote shutter release and mirror lockup to ensure that there was no shake at all.

In post, I struggled to get this where I wanted. Straight out of the camera, the tonal range was extremely compressed. Expanding it to make the black night sky black, without losing all the detail in the support, and keeping everything looking coherent, was a real challenge.

I ended up turning to HDREfex Pro make it happen. Yes, I used HDR software with a single exposure, so I could access the tone mapping tools. It did exactly what I wanted, and a little more. It punched up the green color more than I had intended, but I decided I really liked the look.

Tuesday, November 23, 2010

Look Ma, No Flash!

Sunlight Streaks

Nothing but afternoon light broken up by a wooden structure, a beautiful woman, and 100mm lens at f/3.2, really tight.

Thursday, November 11, 2010

Why Cheap UV/Protective Filters Aren't Worth It

Lots of photographers recommend putting a UV/Skylight filter on every lens you have, to protect the front element from blowing sand, salt, and other mishaps.

This isn't bad advice. After all, once you damage that front element enough, you have to replace the lens, which will cost you a lot more than replacing even the most expensive UV filter. There are plenty of stories out there about thousand dollar lenses being saved by their "protective" UV filters.

What gets overlooked too often, though, is that putting cheap glass in front of your good lens can noticeably degrade the image quality.

The image below demonstrates the deleterious effects of a cheap filter on image quality. Lens is a Tokina 12-24mm f/4, at 12mm, on a Canon 60D. The scene was chose not for its artistic merit, but because it featured the sun just outside the field of view of the lens, a situation that, especially with wide-angle lens, is a recipe for lens flare.

I shot first with the $17 UV filter on, and then with it off. For the second shot, I even changed the perspective slightly, in the direction of the sun, creating an even tougher flare test. The difference between the two photos is striking.

Why Cheap UV Filters Aren't Worth It

Sunday, November 7, 2010

Canon 60D First Impressions from a 50D user

My Canon 60D arrived Friday night. I spent some time with it Friday and Saturday, and here are some quick impressions, and photos.

Overall, it handles well, and for someone coming from a 50D, it take minimal getting used to. I'm mostly impressed. The early review is below.

The Good

The image quality is fantastic. At ISO 3200 and 6400, it's far,far better than my 50D.

It's pretty damn good at lower ISOs as well.

Autofocus is speedy and accurate. Not better than my 50D when it's working (50D is currently in the shop with AF issues), but not worse. It's not a 7D, but that's why it costs less. So, far it's done OK in low light.

As a side note,I was really worried about the lack of Micro Focus Adjustment- the ability to adjust the camera/lens focusing to correct for consistent front or back focusing errors. My 50D has it, and I had to do it all of my telephotos. So far, the 60D has been spot on, although I haven't taken it out with the 300 yet. It did prove to me that both my 50mm f/1.8 lenses have issues, so it's time to replace them with something else.

I have large hands and AF on button fits perfectly under my right thumb.

The articulated LCD is a huge improvement. It's far easier to use for making setting adjustments when the camera is on a tripod, and either low, or high, or turned vertical.

Best of all, you can close it so the screen faces the camera. Not only does this protect it from nose grease, it makes me work faster by curbing my tendency to chimp after every shot.

The metering system has doe a fantastic job with everything I've thrown at it so far, including backlighting and crappy compact fluorescent lighting.

Auto-bracketing now can cover plus/minus 3 stops, not just two.

The Bad

Canon changed the remote shutter switch to the same one used on the Rebel series. So I have to go buy a new remote shutter release, as the two that I have, and that worked with the 10D, 30D and 50D, will not work with the 60D.

Even worse, one of those releases is the expensive TC80N3 - Canon's intervalometer model for time lapse and extremely long shutter speed photography. And guess what Canon doesn't make? That's right, an intervalometer for use with the Digital Rebel series.

While I realize that Canon needs to differentiate between the 7D and the 60D, I figured the 7D's much better AF, faster frame rate and higher end build quality were enough. I don't think it's ridiculous to anticipate that buyers of a $1,000 camera would expect to be able to do time lapse photography with it.

Auto-bracketing still limited to only 3 exposures ( 1 under, 1 on target, 1 over ). For HDR, I want 5.

The Takes Getting Used To

The joystick from the back of the 50D has been changed into an 8-way rocker switch in the middle of the command dial on the back of the camera. The command dial itself is smaller and further down the back than on the 10D, 30D, 50D.

This works and it doesn't. It mostly works for setting the AF point, but when working through the menus, for example, turning mirror lockup on and off, it's easy to jump to the wrong place. Since I'm not used to the command dial's revised location yet, I find myself searching for it with my thumb. I assume that will improve with practice.

Annoyance that isn't Canon's Fault

I use Aperture to manage my photos. Apple hasn't updated Aperture to understand the 60D RAW files yet, and it even chokes trying to import them, so I've been forced into the following convoluted work flow:

1) Use Canon's Image Browser Utility to import the files from my SD card.

2) Use Adobe's DNG Converter to convert the RAW files into DNG format (yes, Adobe has updated DNG Converter, Adobe Camera RAW, and Lightroom 3 to read 60D RAW - go Adobe!).

3) Import the DNG files into Aperture.

Come on Apple, get with the program. (Update: Apple updated Aperture to read 60D RAW files shortly after I published this initially).

And now, some sample images:

ISO 800, Canon 100mm F/2.8 L IS Macro

ISO 100, Canon 100mm F/2.8 L IS Macro

Sunset Sweetgum Leaf

ISO 100, Canon 100mm F/2.8 L IS Macro

ISO 100, Tokina 12-24 F/4 (3 Shot HDR using autobracketing )

To really appreciate this shot, you have to view it large - larger than this layout allows.

Lonely Bush At Sunset

Saturday, November 6, 2010

Canon 60D - High ISO Noise Wow

As a 50D user, my first reaction to seeing these on my computer screen after the first round of test shots was: "Holy shit, that's amazing!"

I cannot think of a single example of my 50D's pics looking this good at ISO 3200 or 6400. In fact, if I have to go to 3200 with the 50D, I generally don't shoot. With the 60D, I think I can shoot at 3200 and even 6400.

You'll need to click on the image to pixel-peep, but here are a series of 100% crops of images in Adobe Camera Raw, with and without noise reduction applied. This way you can see the noise as it is straight from the camera, and then how nicely it cleans up.

Canon 60D High ISO Noise Example

These were all shot under compact fluorescent lighting that registers about the same as tungsten for color temperature. In other words, the type of lighting that screws with digital cameras. Well, at least it did with my 10D, 30D and 50D.

It's also the type of lighting that you'll curse when shooting indoor sports. I'm actually looking forward to trying the 60D out on some indoor tennis, and maybe even basketball and wrestling.

No comparison pics for the 50D because mine is in the shop.

More first impressions from my hands-on with my 60D tomorrow.