Photoshop tutorials
Home [www.limephoto.co.za] Photography Galleries About the photographer Photographer's contact details Links Tutorials page Workshops Front page
All text and photographs © Emil von Maltitz 2010

 

High Dynamic Range Images – Part II


Tonal blending using layers and masks

 

IIn the previous article on blending high dynamic ranges images I looked at my personal favourite technique of simply creating large basic selections and blurring the edge of these selection masks to bring through broad areas of tonal value. When done carefully, this technique can mean for very natural looking images that still have tremendous tonal depth. The problem is that for some images these basic masks don’t work. For instance - image have very fine detail that also happen to have a very high tonal range. Here simple selections don’t work and the photographer has to resort to a more complex selection in order to single out shadows or highlights.

It is this style of imagery that the standalone programmes such as Photomatix excel at. Here the tonality of the various images is mapped out in their respective luminance values and then blended together according the range of tonal values that the photographer desires. This is one of the ways in which you can get that ‘HDR look’. I personally am not all that fond of this ‘look’ but recognise that it is a useful tool in the photographers box of tricks. Here is my approach to creating tonal mapped images without the dedicated software (although you will obviously need an editing suite like Photoshop, Elements, Gimp or Corel Paint).

Creating a tonal mask.

The most important part of tonal mapping is actually working out where your tonal values lie. This is actually very easy. Make sure that you have the channels palette open on the screen. Here you will see four different layers: RGB, Red, Green and Blue. Clicking on Red, Green or blue channels turn the image into a greyscale image relating to the respective channel. Hitting ‘ctrl + left click’ hereafter referred to ctrl + click) selects that channel (you will usually see the marching ants appear). If you ctrl + click the RGB channel you select all the tonal values that are brighter than 50% grey. The beauty of the selection is that it is self-feathered. In other words if we were to go into quick mask you will see various shades of mask, meaning that the tones are selected according to their tonal value as well. Creating a mask of this selection would therefore create a negative mask.

highlight layer Shadow layer Base layer

Putting this into practice then:

  • Step 1 Open the 3 images that you have created your HDR with.
  • Step 2 Layer them in a stack so that they are optimised for the highlights at the top of the stack (the darkest image), optimised for the shadows in the middle (lightest image) and your base exposure at the bottom (the average exposure). It’s a good idea to auto-align the images so that they match up perfectly. You can do this by going to Edit>Auto Align Layers... and following the prompts on the dialogue that appears. Auto usually works well.
  • Step 3 There are two ways to go about this next step. Both work, but have subtle differences depending on the image in question. I would recommend trying both and seeing which works better for the particular image. Click on the view layer icons for the highlights and shadows layer so that only the base exposure layer is visible. Now, go to the channels palette of the and ctrl + click the RGB channel of this layer. This creates a selection in the image of all the pixels that are greater than 50% grey in tonal value.
  • Step 4 Now that you have the selection select the top highlights layer and click on the eye to make it visible. Making sure that the top layer of your layer stack (highlight’s image) is selected and that your colour palette is set to default (hit ‘D’ while the brush tool is selected), select the create mask icon at the bottom of the layers palette. This automatically creates mask that retains the tones above 50% grey from the highlights image while all the shadows from the highlights image are hidden. Letting through the layers below.
  • Step 5 Switch off the eyes next to the highlights and shadows layers and select base exposure layer a second time. Go through to the channels palette and ctrl + click the RGB channel. You now need to invert the selection by hitting Shift + ctrl + I. After this, as in step 4, select the create mask icon in the layers palette, but this time while the shadows layer is selected.

Alternative selections:

Rather than using the base layer to select the highlights and shadows you can opt to use the actual highlights layer to select the brighter portions of the image and shadows layer to select the shadow portions of the image. Repeat Step 3 and 4, but select the RGB channel from the highlight and shadow layers respectively. The final result will be an image with an overall lower contrast than using the base layer as your selection layer would have.

What you should be left with is an HDR image created through tonal mapping. However, this approach sometimes requires a little creativity to get it to work. For instance, when the transition is too great a difference in tonality, the transition area becomes very obvious. Here, applying some Gaussian blur, as in High Dynamic Range Images Part I, can mitigate this. See the next two images as an example.

In this tutorial’s example image the final blend lacks contrast, making the whole image look slightly muddy. To increase the contrast and while retaining the values in the sky I’ve opted to go to the Shadows layer with its mask in place and simply drop the opacity to around 25%. This brings some detail into the shadows so that there is no clipping while retaining more contrast.

Layers palette

 

 

 

 

HDR straight HDR with adjusted opacity

By creating actions for the entire process this is the closest one can get to an automated HDR. Once you have your blended image it then comes down to localised adjustments and optimising the photograph to what you originally previsualised.

The important thing to note is that the one technique in blending high dynamic range images is not necessarily better than another different technique. I often use multiple techniques in a single image to blend the exposures together. My ultimate goal is to create an image that does not automatically look like every other HDR image that is out there. Fiddle, play and experiment and you’ll find that once you have mastered selections, blending HDR images opens up a whole new world of creative possibilities.

Completed image

 

Back to Tutorials