ABSTRACT
This section mainly describes my experience with digital photography underwater. I present a number of techniques and tips which I think may be of value to others. Disclaimer: This section is not meant as a photography course. The tips are based on my own experience. They reflect the way I operate. I do not read books on photography, so I do not know if my tips are common practice. See for yourself if you can use it.

Equipment

• Take your time preparing the camera. If possible, take preparations at home and not at the dive site. Good preparation reduces the chance of flooding the housing
• When diving cold water, flush the housing with air from your cylinder. Cylinder air is (should be) dry. This reduces the chance of condensation (which of coarse takes place on face plate before the camera lens).
• Having expensive equipment is no guarantee for good photos. State of the art equipment makes taking photos easier. Making good photos is an ability (no, an art) of the photographer.
• Using a wide angle lens enables you to get closer to your object. Getting closer means less dust between the lens and the object. I am almost always using a wide angle.

Technique and Technology

Pixels, colors and histograms

A digital photograph on your computer is made up of pixels, the smallest image elements. A pixel contains the color value that has been recorded by the corresponding CCD element of your camera. Usually, the color of a pixel is represented in the computer by three values corresponding to the red, green and blue component of the color. Red, green and blue are the primary colors. By mixing these colors in different proportions all other colors can be made. Each color component value can range from 0 (black) to 255 (full red/green/blue). So for each color component there are 256 levels or nuances. The total number of colors that can be created is 256 red nuances * 256 green nuances * 256 blue nuances = 16.8 million colors.

This enlarged cut-out shows the pixels in the photo

A photo...

and its histograms for primary colors

A histogram is a graph in which for each color or nuance the amount of pixels is counted and plotted. The nuance is presented on the x-axis from dark (left) to full illumination (right). The number of pixels for each nuance is presented as a vertical (black) bar (the count value is reflected by the y-axis). On the left hand side a photograph is presented together with the histograms for the primary color components. The photo was taken using natural illumination. It appears blue and a bit dark. The histogram for the red color component shows a peak at the left hand side, at dark red nuances. This reflects the fact that red light is attenuated by water. Most pixels have almost no red component. The histogram of the blue component on the other hand shows pixels distributed along all blue nuances. The combination of low red nuances and apparent blue nuances makes the photo look blueish. On all three histograms the most pixels are on the left hand side (dark side) of the graph. This reflects the fact that the photograph is a bit dark. The histogram of an over illuminated photo would show most pixels at the right hand side. So on 'regular' photographs each of the primary color histograms should show the pixels distributed over the entire range. This is only a very rough rule of thumb: of course the histograms highly depend on colors and contrast of the object. Anyway, histograms show a lot of color information of the photo.

Manual white balancing

If your (digital!) camera supports manual white balancing you're lucky. This is a feature extremely useful for underwater photography when using natural illumination.

White balancing is the way a whites object appears on your photo. You might think 'What's the fuzz about this? White is white'. However, this is not trivial. How white appears on your photo depends on the color of the illumination. If you take a blue light and shine on a white object, the object appears blue. It will be blue on your photo. A bit more subtle is illumination using TL fluorescent light: this gives your photos a greenish appearance. This is because TL light contains other color components than for example sun light. Your eye corrects very well for this: your partner does not appear to you like the Hulk under TL illumination. For photos however, you have to correct the white balance.

On most digital cameras white balance can be chosen from a number of presets. Depending on your illumination type you chose the right preset. For example, my Oly C-4000 Zoom has 6 presents: sunny, cloudy, tungsten illumination and 3 presets for fluorescent light. Fortunately for the ignorant photographer, factory default setting for white balance is 'auto', which means your camera makes a best guess for itself which white balance preset to choose: (most) photos just look alright!

For underwater photography natural illumination become so extreme that your camera generally cannot cope. As diver you know that after a few meters depth the red component in the light becomes attenuated to very low levels. The result is that everything looks blue, especially on photos. You're probably familiar with relatives buying a cheap disposable underwater cam without flash and showing you blue fish on a blue background. For this situation manual white balancing comes in handy. You 'teach' your camera what is white by pointing it at a white object and push a button. Now the camera knows what is white and which correction to apply to make it appear white on the photograph.

The sequence below show a glass paperweight at the bottom of the Kardinge swimming pool (ok, not really exiting). The tiles at the bottom are white, so the entire scene basically is white. The photos are presented together with the histogram for the three color components red, green and blue.

White balancing is set to automatic. The scene is illuminated using a flash. This basically is white light at short distance. So the picture looks as it should be.	White balancing is set to automatic. The 'natural' illumination of pool lamps is used (you can see the images formed by the paperweight). Since the lights are 5-10 m away, the red component of the light is attenuated giving the picture a blueish appearance.	White balancing is set to manual and has been corrected using a white plastic bag as reference. The scene is still illuminated by the pool lights. However, the colors are shown correctly as if the scene was illuminated with white light: the white tiles appear white.
The histograms show for all primary color components nicely distributed pixels	The histogram for red shows a shift to the left, corresponding to the attenuation of the red by water. The green component is the strongest.	The histogram clearly shows the proper white balancing: each primary color is equally strong, which is to be expected for a basically white scene

How about Photoshop for restoring the white balance? Yes, that is possible. However, in the original picture almost all red information is lost. You have to blow up the red signal to get a natural image. However, you cannot blow up something which is not there.

Normally, the entire range of red nuances consists of 256 red nuances. When the red component is attenuated by the water only the lowest part of the range is used. Say the attenuated red levels range for example from 0 to 24. This means only the lowest 25 of 255 levels are used. Blowing up the red component means the the levels are equally spread amongst the 0-255 scale. However, after spreading there are still no more than 25 discrete levels (=nuances). For the image white balanced by the camera however all red levels are used. (I actually don't know how this is arranged in the camera. I guess the analog gain level for each primary color is increased before converting to digital) The difference is shown in the photos and histograms below.

The original photo, automatically white balanced by the camera.	The Photoshopped version of photo. The correction applied was simply applying the 'Auto Color' feature of Photoshop.	For comparison, the manually white balanced photo is repeated here.
The red component histogram shows the shift to the left, indicating attenuation of the red primary color component.	As can be seen from the histograms, each color component is now equally strong. However, certain color nuances are simply not present, especially in the red component histogram. This shows up as empty vertical bands. In the photo this appears as a grainy appearance, because color transitions are not smooth but take place at discrete steps	The histograms show equally strong primary colors and presence of all nuances

The art of photography

• Flash! Or not...
  No only use your strobe, consider natural illumination as well. This makes a hell of difference. The photos below show the same fish in different illumination.

  A pike perched. Flashed. Results in highlights on the body.

  The water was shallow, the sun was shining. Hence, lots of natural light. Using it gives a complete other picture.

• Into the blue
  Using no strobe often leads to blue or green colored photos. This is not necessarily wrong. Personally, I like such pictures, because they reflect the sphere of diving. I am very proud of the picture below, showing the diver on the Marcus wreck in the Red Sea. The photo was taken using natural illumination
  

  One of my best pictures. The diver swims from the dark interior of the wreck into the light above. The blue color definitely adds to the sphere.

• To photoshop or not to photoshop...
  Photoshop can do a lot with your photo. However, be limited using it. Try to take good photos instead of enhancing rubbish using Photoshop. Use Photoshop as the finishing touch of your photos. Photos tend to become unnatural when retouching to much.
• Think different
  Be creative! Think different! Always search for alternatives. Let me give you an example and verbose the way a dragon-fly photo was taken.

  1. Yeah! A dragon-fly ! One of my favorite photo objects. Let's take a picture before the beast flies away. Click. Wow, awful picture.... Where is the animal anyway?		2. Nice animal. It is still there. Lets make the background fuzzy by playing with diaphragm and exposure time. Click. Oof. Not really what I want.
  3. Let's try a different background: the sky. Click. Ahhh... It is becoming better. But the wings, I do not see them.		4. Geee... The animal must be stoned or so. It is still there with me around annoying it. Ok then! Lets try under illuminating it. Click. Yes! Now we are getting somewhere!
  	5. Let's fiddle around with composition a bit. Click. Kick ass!!! This is it!!! Thank you little friend for allowing me to harass you! Another photographer goes home happily with a nice picture. See the photo gallery for a larger version of it.

• Different angles
  Try different angles

  1. Snail		2. Snail and anemone may make a nicer composition
  3. From the other side it becomes even better		4. Shifting a bit to get a good composition
  5. Rotating the camera (and flash) changes dimensions and illumination		6. Shifting a bit... Voila!

• Just one more...
  Take more than one picture. Slight changes may make the difference!

  1. Two pictures, almost the same

  2. This one is better. Why? I don't know. Perhaps the head of the snail and the anemone are more free from the background.

• Use the environment
  Just taking pictures of fish is not really exciting. Literally thousands of photographers took the same photograph of the same type of fish. Buying a book is more efficient when you're interested in photos of fish. Try using the scene to make photos more exciting. Or just choose an alternative illumination. See the examples below.

  Alternative way of showing a moray eel. It just peeps around a rock, exaggerated by putting it in the corner of the photograph. Again, the blue color adds up to the sphere.

  A lucky shot of a fast swimming pike-perch. It swims right into the spotlight. Only its front side is illuminated, drawing attention to its head and eyes and giving it a bit of an aggressive look. I am very glad with this one.

• Screw the rules
  Don't let your creativity be limited by rules. For example: there is a rule stating that you should not take photos into the sun. I present you three photos taken into the sun I am happy with:
• Strobe upside down
  Fish and other underwater animals often hide in small caves or underneath rocks. Using your strobe means your strobe has to shine right into the hide out. If not, walls and roofs cast shadows on the object. You can rearrange your strobes. This takes time and may be difficult on fixed camera-strobe assemblies. Simply rotating or repositioning your assembly might solve the problem. It can result in nice photos to use unnatural illumination, like the example below.

  This photo was taken holding the camera-strobe assembly upside down. The pike is illuminated from below. This is unnatural, since normally you would expect light coming from above.

Presentations

• Show your photos in a presentation. Lots of tools are available. Use for example Proshow. You can even use MS Powerpoint if you have nothing else available.
• Select the photos carefully. Do not show to much photos. Show the exciting photos. Keep your audience attracted.
• Music does it! Put music to your show. If possible, synchronize transitions with the music. Different types of music give entirely different presentations. For under water photos ambient, new age usually is ideal. Generally, use instrumental music. Lyrics distract the audience from the images.
• Fluent fading transitions usually are most appropriate. Fancy transitions are annoying and distract from the content.