In this section, I will describe my general HDR workflow. I will not go into the details as these will be discussed in other sections. Once you have shot the photos out in the field, the HDR work really only starts. I know what you think: He’s right! You still have to run the shots through an HDR program to get the final image. But this is actually only a tiny fraction of the complete HDR workflow that (at least) I execute before an HDR photo is finished.
The first rule of thumb is:
Creating the HDR and doing the tone mapping takes minutes, the work required in Photoshop after that can take hours!
In the Internet, you can see tons of HDR photos that were not post-processed in any way. The usual result is a picture that looks flat with little contrast.
The Complete HDR Workflow
In summary, my personal HDR workflow goes like this:
- Create the source shots using the exposures that are adequate for the dynamic range of the subject. This may be the standard three shots (+2, 0, -2 EV). It may also be 5 or even 9 shots. By all means shot in RAW format. This provides the best quality and flexibility. I have shot in JPEG for a long time, and I wish I used RAW right from the start. Especially in post-processing, the difference is huge!
- Take your time – I discovered that directly after a shoot, I am not able to get the most out of a shot. It is better to let them linger on you disk for a few days. Then, when you look at them again, you are more objective in your judgment about what are the good shots and what are the bad ones.
- Select the best shot (bracketed series) and start creating the HDR image.
- Adobe Camera RAW (ACR)
- If your lens tends to create chromatic aberration (CA – color fringes on high-contrast edges), load the RAW shots into Adobe Camera RAW (ACR) and correct that. If you don’t, then these color fringes will be amplified in the HDR process, leading to very ugly-looking images. ACR does a very good job at removing those artifacts. Photomatix (at least until version 4.0), on the other hand, is very poor at doing this!
- You should also check if all source images have the same color temperature (white balance). If this is not the case, set the White Balance for each image to the same value. This is particularly important if you are working on an HDR panorama.
- Save the corrected photos as 16-bit TIFs.
- Early noise reduction: Experiments show that applying NR at this early stage, before feeding the source exposures into Photomatix, produces cleaner and sharper images. This would be the point in your HDR workflow that is ideal for early NR.
- Make sure you turn off any NR and sharpening in ACR.
- Load the TIFFs produced by ACR into Photoshop and apply your favorite NR software to each of the exposures. Make sure you read my tips on this method.
- Save the results back to TIFFs and use those in the next step as input to Photomatix.
- Load the (possibly corrected photos) into Photomatix. If they are already CA-corrected, do not choose the respective option in Photomatix.
- Avoid any correction of ghosting if at all possible. Using this option normally reduces image quality. In Photomatix Pro 4.0 and above, there is a semi-automatic deghosting option that can be used to get rid of ghosts in specific areas without contaminating the entire picture with deghosting artifacts.
- Create and save the HDR
- Go to the tone mapping dialog and define your tone mapping options. Once you have found settings that suit your style, this usually takes only 10-15 of minutes. Change the settings in a small region around your preferred settings to see whether the image needs any particular adjustments. My most important advice about the tone mapping is: Be conservative! Extreme settings will result in ugly-looking pictures. Excessive halos and contrast inversions (e.g. the sky being darker than the building in the foreground) are usually the results.
- Click on Process and save the result as 16-bit TIF.
- Load the tone mapped image in Photoshop to start the post-processing.
- Make necessary transformations to correct the image (e.g. perspective correction and/or rotation to straighten the horizon). Do this before you do anything else!
- If you are using Topaz Adjust or a similar tool to enhance local contrast and colors, I recommend to apply this first: Make a copy of the background layer and apply the filter of your choice.
- Apply some noise reduction. The tone mapping process tends to increase the noise in the image such that noise reduction is a must for any HDR photo. You should apply it early in the whole process in order not to amplify the noise with your other processing steps. I recommend Topaz DeNoise. It is simple and very effective. Make a copy of the layer treated with Topaz Adjust and apply Topaz Denoise to it. NOTE: If you applied NR already before the HDR merging and tone-mapping, you may not need this step. Look out for noise that sneaked back into your image and remove it selectively.
- Standard global adjustments may be the most obvious things to improve the image and should be applied in any case. These include Levels (or Curves) adjustments to correct the contrast and brightness and saturation adjustments to enhance the colors. Use adjustment layers instead of the dialogs found under “Image > Adjustments” wherever possible. This allows for a very flexible editing where you can change, turn on, and turn off the adjustments at any time.
- However, in order to really get the most out of your photo, you need to apply specific adjustments selectively to individual parts of the image. For example, if an image contains sky, it it may be desirable to enhance the blue tones in this part of the image without affecting the other parts of the image. E.g. the warm tones of a Mediterranean house will suffer from increasing the blue tones globally. Such a selective adjustment requires sophisticated masking abilities. My experience is that masking is the single most important technique required for good post-processing results.
- Apply some sharpening to the final image. When you are done with the other post-processing steps, your image will be dispersed over multiple layers, each of which has a mask to reveal only some parts of the image. The most frequently used sharpening method – the unsharp mask – is not well suited to sharpen such an image because it modifies the bits of a specific layer directly. I recommend using the high-pass filter for sharpening. This filter isolates only the edges of the image.
- Apply a vignetting effect if it suits the image. Vignetting darkens the edges of the image and focuses the attention of the viewer on the middle of the image. Some subjects can really be improved using this technique. It can increase the depth perceived by the viewer considerably.
- Create a frame. Whether you do this depends on your own personal preferences and on how you wish to present the final image. When you present the image via the Internet, a frame can improve the visual impression considerably.
- Change the mode from 16 bits to 8 bits and save as a JPEG. Again, this depends on the way you want to present your work. For the Internet you usually need a JPEG.
- Let the finished file sit on your hard disk for a day or two. Don’t post it right away. I found that when you get back to it a day later, you will discover things that you might want to change, things that you did not discover while you were working on the details of the image very intensely. If you let you mind do other things and some time passes, you will look at the image more objectively, more like your viewers will. If you’re content with the image now, go ahead and post it.
- Set important EXIF data. EXIF is a standard for image meta data. Sometimes, this data gets stripped from the image during post-processing because some programs simply discard it. Use a tool like exifool to copy the EXIF data from the original 0 ev image to the final JPEG.
- Set the location data in the image. Tools like Geosetter allow you to write a location (longitude and latitude) where the image was taken into the meta data of the image. They take a GPS track that you may have recorded separately or you can manually define a location on a map and write it into the image.
- Upload to your favorite photo sharing platform and do not forget to describe the image properly.
You can see that there is a lot of work involved with creating an excellent HDR image. You may think that this is boring and unnecessary. However, this is exactly what makes the difference between yet another photo and a real eye catcher. Of course, I assume that you got all the other photographic stuff right during the shooting (interesting perspective and composition, unique subject, perfect focus, uncluttered background, etc, etc, etc.). I am not going to write anything about this basic stuff since there are thousands of books on this subject.