Lessons in DSLR Workflow:High Dynamic Range (HDR)
High Dynamic Range (HDR)
High Dynamic Range (HDR) imaging permits photographers to capture a larger range of tonal detail from a series of image captures, rather than using the limited range of a single image capture. A digital SLR camera is capable of capturing a limited amount of tones in a single image. Characteristically, we surrender elements in an image when we capture it. Example: A scene with bright, backlit clouds in the background and some landform in shadow is beyond the dynamic range of capture. If you expose for the clouds, the landform becomes a dark silhouette. If you set the exposure to capture detail in the landform, the brighter clouds are blown out, and detail is lost. This is because our eyes are much more sensitive to the range of light within a scene than the digital sensor is.
The solution is to take more than one capture and bracket the exposures. The merge-to-HDR Lightroom-to-Photoshop feature enables the photographer to combine a series of bracketed exposures into a single image which encompasses the tonal detail of the entire series.
Warning: The bracketing technique does not work with subjects in motion and is primarily for still-life and landscape photography. Any motion will become blurred, and ghosting will occur. The registration of the bracketed images can become a problem for this type of image processing.
Start with the camera on a steady tripod. Set the camera for Raw file format capture and bracket five stops on the capture. One normal exposure, two on the underexposed side (−1, −2) and the two on the overexposed side (+1, +2). This should give you plenty of room for merging all the tones in the scene. Bracketing unleashes a whole new range of lighting possibilities, which many have steered clear of in the past for technical reasons. You can bracket exposures manually, but many digital SLRs also have an auto feature to do this function. Check your camera manual for auto bracket specifics.
Beware: There is a bit of a tradeoff with bracketing. Trying to expand the tonal range will come at the expense of decreased contrast in some tonal areas as well as reduced saturation of the image. Still, learning to use merge-to-HDR in Photoshop CS3 can help you make the most of your dynamic range under some problematic lighting situations while still balancing this tradeoff with contrast.
Photoshop generates an HDR file by using the EXIF information (camera metadata) from each of the bracketed exposures to determine the shutter speed, aperture, and ISO settings. It then uses this data to evaluate how much light came from each image area. Because this light has different intensities, Photoshop creates the HDR file using 32 bits (creating a large file) to describe each color channel as opposed to the usual 8 or 16 bits (see nearby sidebar on bit depth). Down sampling from the large 32-bit image to the 16- or 8-bit image is like getting better results from a broader survey of information. The more samples in a survey, the better the data collected.
Processing HDR Images
First, select the series of bracketed exposure images in the Lightroom Library Module Grid view. Make sure you’ve taken at least three exposures, although as I mentioned five or more is recommended for the best possible precision. More exposures allow the HDR algorithm to better approximate how your camera deciphers light into digital values.
In Lightroom, click Photo > Edit in > Merge to HDR and click OK (FIGURE 13.11).
Your images will appear in the Photoshop Merge to HDR dialog with the exposure values (EV) listed with the thumbnails on the left-hand Filmstrip. (You uncheck a box to eliminate individual images from the merge.) The merged results appear in a 32-bit image. You can adjust the overall tones with the Set White Point Preview slider. Small nudges on the control slider go a long way, so don’t be too aggressive with this adjustment. I look to adjust the white point on the Histogram to where the white highlights have a bit of detail. You will find that the image becomes darker when maintaining the detail in the highlights (FIGURE 13.12).
Click OK to merge the selected, bracketed images and white point into the 32-bit image. This image will appear in Photoshop after processing with good highlight detail, but may look really dark and lack contrast over all.
In Photoshop, click Image > Mode > 16 bit (or 8 bit) (FIGURE 13.13).
The process of converting the 32-bit image to 16 bits brings up the HDR Conversion dialog. This dialog has four elements: Exposure and Gamma, Highlight Compression, Equalize Histogram, and Local Adaptation (FIGURE 13.14). Depending on how the image is capture and processed, you may not need to use all the controls to adjust the image. The Local Adaptation Method of adjustment is the one I use most often when experimenting with such image processing.
Additional information about FIGURE 13.14:
- Exposure and Gamma is the default option. Adjust the exposure to get the desired brightness and gamma. If you prefer an image with more contrast, lower the gamma. For less contrast raise the gamma.
- Change the HDR Conversion method to Highlight Compression. Just by selecting this choice, the image will darken and the highlights will exhibit detail.
- Change the HDR Conversion method to Equalize Histogram, allowing Photoshop HDR to do the math to average out the Histogram values. Some images undergo a significant improvement in contrast when this is applied.
- Change the HDR Conversion Method to Local Adaptation. This is where you can adjust the tone mapping, placing points on the curve and moving these to season the image to taste. Try a visually pleasing S-shaped curve to enhance the image. Once you’re happy with the curve, adjust the radius and threshold sliders to make sure there are no halos in the photo.
The radius controls the mask blur, whereas the threshold decides what gets blurred and what doesn’t. (Poorly converted HDR images have a glow around the contrast transition edges areas of contrast.)
When you finish the conversion, your disappointment that the image information lacks contrast and color saturation will be realized. But move on the next step and click OK to convert.
Now use the normal adjustment layers to improve the contrast with levels or curves and saturation globally as well as locally on the processed file (see Chapter 14 for more). Again, the object here is to deal with extreme dynamic ranges that cannot be recorded in a single frame, using an automated solution provided by Photoshop (FIGURE 13.15). I have found many photographers using this in creative ways as well because it provides a new element to the toolbox that was not available in the past. Overprocessing the file in HDR can lead to some interesting effects.