Update: Please also see 7 unique ways that luminosity masks can enhance your image beyond increasing dynamic range.
Few words seem to polarize photographers more than HDR. Throw a couple of exposures at an algorithm, turn a few knobs, crank it up to 11, and out comes… something garish. But with some judicious use, HDR can create some beautiful images. On the flip side, luminosity masking tends to be associated with both beautiful images and something approaching quantum physics. But luminosity masking doesn’t always result in beautiful images and isn’t necessarily as challenging as it can seem. So is one technique better than the other? Like a lot of things, the answer is “it depends”. I use both.
When do I use HDR?
There are three primary scenarios where I use HDR. First is when the dynamic range of the image is varying considerably in numerous small parts of the image (such as can often happen with a cityscape image), I find that HDR can be a simpler approach than trying to manually blend exposures with luminosity masks (as the masks get more complex when the dynamic range varies considerably in numerous parts of the image). There are many great HDR software packages, but I prefer the look and feel I get from Photomatix.
Second is sort of a variation on the first where I want to avoid HDR artifacts (halos, noise, etc). In this situation, I actually use HDR to do exposure blending, and then often finish the image with luminosity masks. I use Photoshop to merge the exposures into a 32-bit TIFF and then tone map the image in Camera Raw. The Photoshop HDR approach creates a subtle and very photo-realistic result. One downside of this approach is that the 32-bit TIFF files Photoshop creates are massive (>1GB!) So, as much as I like to use a non-destructive workflow, I’ll often convert the image down to 16 bits, finish with luminosity masking or other techniques, and delete the 32 bit file.
And finally, there are also times when you want to really punch up the texture in an image for creative reasons, and HDR is awesome for that (Nik Color Efex also has some great filters and there are other ways to pull out texture in Photoshop as well). I almost always finish the HDR by blending in some of the original exposures to achieve a nice, balanced final look. I like to get a result that’s just a little surreal to really express the way I felt the scene when I took the photo – but I definitely don’t want the image to look over the top.
When do I use luminosity masks?
HDR tends to create artifacts that need to be managed, especially noise and halos. Both of those tend to get much worse at the interface of large areas of vastly different tones (such as the edge where a mountain and a beautiful sunset meet). Thankfully, that’s exactly where luminosity masks excel. So luminosity masks tend to be my “go to” tool for a lot of natural landscapes where I need to blend 2 or more images to capture the full range of color and tones in a dynamic scene.
Luminosity masks are also great when you want to make precision adjustments to specific parts of the image. By its nature, HDR creates an entirely new image. You could apply parts of that image using layers, but that will end of being a lot of work. And, since HDR can shift the color, tone, and grain of an image – it might be difficult to blend in just part of the HDR in a natural way. Whereas with luminosity masking, you can easily select just the tones you want to impact, and then further restrict that to just the parts of the image where you want to adjust those pixels. This video shows how luminosity masking can be easily used to adjust tones and saturation in specific parts of the image.
Luminosity masks can also be used WITH HDR. If HDR is the best way for me to create a vibrant cityscape, I might use HDR and then use luminosity masks to control halos and noise in the sky.
When should I use something else?
I skip HDR and luminosity masking whenever I can. They take a lot of time to create, the process has some destructive steps, and the resulting files are often pretty huge. There are also some situations where it’s nearly impossible to blend multiple exposures (often scenes with moving pieces or action scenes where you can’t get multiple exposures of the exact same subject). Adobe Lightroom (or Photoshop camera raw) has gotten so good that I’m often able to create my final image completely within Lightroom. The key is shooting high quality images (low ISO RAW) and understanding the basic tonal controls and local editing tools (the gradient and brush tools). Here’s a great example of a cityscape that many people might throw into Photomatix, but I think the final result looks amazing straight out of Lightroom (and it only 20 minutes!) Of course, there are some scenes that just have so much dynamic range you need to use HDR or luminosity masks to get the final image you want.