HDR – follow up to previous post

In yesterday’s post I mentioned a method to expand the number of images you make when your camera is set on auto bracketing by using aperture priority and then changing the exposure compensation for each set of auto bracketed images.  I tried that last night and learned something unexpected.

For the test my camera settings were on auto-bracket (5 images, +/- 1 EV) to give me an image series like this:

2 stops overexposed

1 stop overexposed

normal exposure

1 stop underexposed

2 stops underexposed

I also made two other series by changing the exposure compensation to +2/3 EV and -2/3 EV.  This resulted in 15 images, all with different shutter speeds (f/8 was my aperture).

The slowest shutter speed gave me this:

ISO 100, f/8, 1 sec.

The fastest shutter speed gave me this:

ISO 100, f/8, 40 sec.

And the “correct” exposure looked like this:

ISO 100, f/8, 8 sec.

Sort of like Goldilocks and her choices.

Well, I dropped all 15 images into NIK HDR Efex Pro and got this out as the default:

Very nice.  Detail in the deepest shadows as well as the lightest parts of the image.  And very little noise in the shadows.  Obviously the software does a good job of picking and choosing the parts of each image that delivers the best exposure.

But why go to that much trouble?  Why not just use the brightest, darkest, and normal exposed images?  Then there’s only three to worry about.  Let’s see what we get:

At first glance this looked pretty good, but then I started examining it more closely.  Turns out more images for the software to chew on ARE better.  Here are some closeups:

15 image HDR

3 image HDR

The first thing I noticed were the color shifts in the 3 image version.  Look at the shadow line on the right – see that green cast in the 3 image that isn’t there in the 15 image?  The green continues onto the carpet at the top of this crop.  Next, look at the details in the door grain on the left.  In the 15 image version the grain is clearly defined with a nice range from light to dark tones.  In the 3 image version that all looks a bit muddy, as if the tones are running together.  Finally, look at the detail in the carpet.  It looks sharper in the 15 image than in the 3 image.  If I print these two at 16×24″ or larger this difference will definitely have an impact on print quality.

So, my thinking turns out to be correct – more images means better blends – but surprisingly for more than just contrast.  Color and details are all built into the data being processed so the more that is available the “finer” the software can perform.  Sure looks worth the extra seconds needed to work on those 15 images.  Now to apply this to landscapes and enjoy the benefits.


Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s