Lighting for Digital 5

Dave Montizambert searches for optimum digital exposure

Page 2

Therefore, to capture the optimum exposure in digital it makes sense to reverse Ansel’s rule to read, 'Expose for the highlights and process for the shadows'. Now that is a pretty bold black and white statement that forgets to mention the quarter-tones, mid-tones, and threequartertones, and it really does not fully address the extended capabilities of photographers who physically create lighting for digital capture. It’s all about contrast, more specifically contrast control. Ansel Adams did it with exposure and development because he could not affect the lighting contrast before capture. Studio photographers do it with lighting since they can physically alter the lighting contrast before capture. Therefore I am compelled to make a new big bold black and white statement for those who do control contrast prior to capture, “Place the scene’s tones with lighting and then set this manufactured contrast at the optimum exposure by making a hypothetical grey card read middle grey in the raw capture – 113–118 levels for a 2.2 gamma colour space such as Adobe RGB or sRGB or 93–96 levels for a 1.8 gamma colour space such as Colormatch RGB. Huh, that’s funny, that is pretty much how I did it with film – I lit for tones and then exposed for middle grey.

Bits, Bitmap, and Bits Per Channel

Pixel stands for picture element. Digital cameras and scanners create bitmap data. A bitmap is a representation of an image using pixels as opposed to vector data that is used to represent graphics which it does using a series of lines or curves that are stored as mathematical formulae rather than pixels. Therefore a digital camera or scanner cannot create vector data only pixel data. Pixels are made up of bits, hence the term bitmap meaning that a digitized image is a map of a series of bits. A bit is a single binary digit – 0 for off or 1 for on. This provides a limited choice of tones, white or black. The greyscale image of model, Sarah Madro, on the top left side of Image [5], confirms that an image made up of one-bit pixels cannot create any tones between pure white and pure black and so cannot create the illusion of continuous tone.

Two-bit pixels are strictly that, two bit, pretty well worthless for making images. Looking at the top row, centre image of Sarah in Image [5], I think you will agree that this image is better than the previous but still a long way from looking like continuous tone. A two-bit pixel can only be one of four possible tones; zero and zero for white, zero and one for light grey, one and zero for dark grey, and one and one for black. For any sort of real quality we need greater pixel depth, which means more bits per pixel. As you can see from the bit-depth image progression of Image [5], increasing the pixel depth by one-bit doubles the possible tonal or brightness levels – so from 2 to 3 bits goes from 4 to 8 levels, from 3 to 4 goes from 8 to 16 levels, and so on. Jumping up in bit depth to a more usable level, we find that an image made of 8 bit pixels breaks the greyscale down into 256 brightness levels; 256 levels easily fools the eye into believing that it is seeing continuous tone as in the image of Sarah on the bottom far right side of Image [5]. And so it seems that a greater pixel depth such as 8-bit and up is needed if we are to fool the eye into believing that a mosaic of square pixels is a continuous tone image.

In Image [5] we established that we need at least 8 bits to create a believable looking greyscale image, but what of colour? Creating colour digital images we need at least 8- bits for the red channel, 8-bits for the green channel, and 8-bits for the blue channel. 8+8+8 equals 24-bit colour as seen in top graphic of Image [6]. You will often hear the two interchangeably, 8-bit colour means 8 bits per channel, and 24-bit means the accumulation of 8 bits per channel in red/green/blue composite colour.

With 24-bit image files we end up with 16.8 million possible colours; each channel has 256 possible levels of tone – 256x256x256=16.8 million as seen in the bottom graphic of Image [6]. This is a lot of colour! In fact it is more than enough to create the illusion of the full visible spectrum.

Dave Montizambert lectures internationally on lighting, digital photography and Adobe Photoshop. He is also a published author having written two books on lighting and digital photography (www.montizambert.com) plus numerous magazine articles on these topics in North America, Europe, Russia and Asia. Dave also creates Photoshop tutorial CDs & DVDs for www.software-cinema.com. Dave is available for lectures and workshops in your area and can be reached at montizambert@telus.net or www.montizambert.com. Dave Montizambert owns and operates Montizambert Photography Inc. located in downtown Vancouver. For the past 25 years his company has created photographic images to aid various organisations and companies with their communication needs. He has created images for clients such as: McDonalds Foods, Motorola, Atlanta Scientific/ Nexus Engineering, Toyo Tires, Tri-Star Pictures, Warner Brothers, Constantine Films of Germany, Chevron Canada, Cuervo Tequila, the Canadian Broadcasting Corporation, J&B Scotch, Hong Kong Bank, Chimera Softboxes, B.C. Lottery Corp., Blackcomb & Whistler Mountains, Tsing Tao Brewery of China, B.C. Hot House, Kona Bikes, No Fear Sports Gear, Kodak, and Canada Post. His work has won Georgie, Lotus, Hemlock, Studio Magazine, CAPIC, and Graphex awards.

Page 1 - Page 2

Photo Quote: I tried to keep both arts alive, but the camera won. I found that while the camera does not express the soul, perhaps a photograph can! - Ansel Adams