Wednesday, August 6, 2014

Using HDR with NightScape Exposures

A "false" HDR exposure of the Milky Way stars reflecting in Oxbox Bend, Grand Teton N.P. ~ © Royce Bair
(Canon 5D Mark III, Rokinon 24mm f/1.4 @ f/2.0, 13 seconds, ISO 6400)
Exposure for sky
When doing my "NightScape" style of starry night landscape photography, I will often do a much longer second exposure (often at 4X longer) and blend the foreground portion with the original "sky" exposure. This enables me to give more detail to mountains, trees, and water features —otherwise, they are are usually just a silhouette against the night sky (like the image to the right).

Second Exposure Problems: Making a longer exposure second exposure is very time consuming because I usually turn on the camera's "Long Exposure Noise Reduction" feature which also takes a "black" exposure for the same amount of time, compares the two images, and cancels out the hot/noisy pixels in your final image. I gladly take this additional time for the increase in quality; however, images with water will also reflect cloud changes (between the exposures) and star movement in the longer, second exposure. Water automatically magnifies the brightest stars and diminishes the lesser stars. It also has a tendency to make each bright star produce a "long" reflection —similar to the way a rising moon reflects on the water. My longer (4X) second exposure greatly exaggerated those lengthy star reflections, causing them to look weird and out of character. To avoid this, and still get better foreground detail I turned to False HDR: False High Dynamic Range Imaging (sometimes called Fake HDR).

Standard HDR vs. False HDR: In regular HDR imaging, one takes at least three exposures: one at the normal exposure value, one at a +1.0 EV (or higher), and one at a -1.0 EV (or lower). These three images are blended, "fused", or tone-mapped together using HDR software, i.e. Photomatix —with the final image extending the dynamic range of the scene.

In False HDR, one manipulates the original exposure to obtain the other two exposure values. This is best accomplished using the Adobe Camera Raw Converter.

In the Adobe Camera Raw Converter, use the "Exposure" slider to obtain your 3 Exposure Values
Save one image at the neutral or "0.00" EV. Save the second image at the "+1.00" EV, and the
third image at the "-1.00" EV (16-bit RAW images can be converted to EV's up to +/- 2.00 stops).
Here are my three converted images: Normal exposure, +1.00 EV, and -1.00 EV.
These are saved as 16-bit TIF files.
HDR Processing: The above three exposure values were save as 16-bit TIF files and imported into Photomatix for HDR processing. I typically output a "Natural Fusion" image that just blends the three images and extends the dynamic range of the scene. I then output a tone-mapped version that gives additional local contrast, texture, and shadow detail. The Natural Fusion and tone-mapped versions are then blended using Photoshop's Adjustment Layers, varying the opacity of each layer to taste. Additional contrast is then applied to the final image via an S-curve to bring out the latent colors in the sky.

5 comments:

  1. Nice article!

    May i ask how do you blend the two final images together?
    Layer masks in Photoshop and brushing?

    ReplyDelete
  2. Intersting approach. Does creating the 3 different exposures in raw converter improve the final quality that much, rather than just opening the RAW file directly in Photomatix and creating a faux HDR image that way?

    ReplyDelete
  3. This comment has been removed by the author.

    ReplyDelete