1 Study Guide for Complete Digital Photography Chapters 4-5 Chapter 4: Image Transfer Choosing a Computer In general, newer machines will almost certainly have the processing power you need to drive today s digital photography applications. That said, there will be a performance difference between a new, top-of-the-line computer and a lesser model. This will be especially pronounced if you use a camera that has a sensor that packs a lot of pixels 25 megapixels or higher. One thing that s changed a lot in the last few years is that more image editing applications take advantage of the extra processing power that can be provided by a video card. The video card in your computer provides the interface between the computer and your monitor, and these cards often have their own dedicated graphics processing computers onboard. These graphics processing units, or GPUs, can provide an extra processing kick for those programs that exploit them. The latest version of Adobe Photoshop, Lightroom, and Apple Aperture can exploit a GPU to improve performance. RAM (random access memory) is also an important consideration when working with digital images. If scrolling or zooming your images is slow, or if you see images paint in chunks onto the screen, then you might want to consider adding more RAM. Storage Digital images can take up a lot of space. The good news is that hard disk storage is cheap. Simply put, the more storage you have, the better. Backup Hopefully, you haven t already learned this the hard way, but eventually a hard drive will fail. A hard drive is a mechanical device with a lot of moving parts, and after a while, all that disc spinning and drive head moving takes its toll and a mechanical failure occurs. When this happens, your drive will most likely become unusable. While there are services that can attempt to get your data off the drive, they usually charge around $1,000 with no guarantee of success. You can configure multiple hard drives into a RAID, or redundant array of independent disks. The different categories of RAIDs provide different degrees or speed or fault tolerance (the ability to recover data if a drive fails) RAID0-6. RAID0 and RAID1 are the two types you re most likely to encounter. RAID0 uses striping, and RAID1 uses mirroring.
2 Monitors If you ve ever stood in the TV department of an appliance store, you ve seen firsthand how different an image can look on different screens. Monitor profiling and calibration is the process of tuning your system so that your printed output better matches what you see on-screen, and so that the same image will appeared consistent from one screen to another. Just as we talked about mixing the three additive primaries of light (RGB) can be mixed to create white light, the three subtractive primary colors, RGB, can be mixed to create black. Two Approaches to Workflow Copy images to your computer Organize your images by placing them in folders. Using the file manager on your computer, you can find, move, copy, and rename images, and perform any other organizational tasks. Use browsing software to view thumbnails and previews of all of the images in a folder. Launch images from your browser or open images from your file manager into an image editing application. Output your images and then back up and archive the files using backup software or your program s file manager. Adobe Bridge, which has shipped with Photoshop since version CS2, offers all the browser features you d expect, wrapped up in an interface that integrates seamlessly with the other apps in the adobe creative suite. Camera Bits Photo Mechanic provides most of the features you d want in a browser, including capturing, renaming, raw conversion, metadata for editing, and much more. In addition, the program provides batch processing features, the ability to import from multiple cards simultaneously or have older hardware. Image Editing Applications You can use an image editing program like Photoshop to make wild changes to an image. Of course, what you ll use it for the most are less obvious edits. Cropping and tone and color adjustments, and maybe the occasional retouching are the editing tasks you ll perform the most. Photoshop CC, Lightroom, Photoshop Elements, Nikon Capture NX, Aviary, GIMP, Paint.net, and Corel Draw X7 are some of the more popular image editors. Nondestructive Editing
3 Earlier, you learned that an image is composed of a number of pixels represented by numeric values. In most image editors, when you make an adjustment, the color values of the adjusted pixels are altered to reflect your changes. In a nondestructive editing system, actual pixel values are never altered. Instead, a description of any edit that you make is stored in a list. Any time the computer need to display, output, or print the image, the original pixel values are processed according to the instructions stored in the list of edits. Available for both Mac and Windows, Lightroom offers importing, sorting, comparing, organizing, keywording, editing, raw conversion, Web page output, and printing, all in a single application that provides different virtual rooms [modules]. Lightroom s strongest advantages are an entirely nondestructive editing approach and the fact that it s built around Photoshop CC for raw conversion. If you plan on using raw, then you ll need to be sure that your raw conversion or workflow application supports your particular camera. A raw coverter must have a special profile for each specific type of raw-capable camera. If your application doesn t specifically say that it works with the raw files from your camera of choice, then you ll need to consider or consider a different program. Most software vendors are diligent about releasing updates to support camers as they re released, but it can sometimes take a while for these updates to appear. Card Readers Depending on the type of reader, a media card reader can be faster than plugging your camera into your computer, and it doesn t drain your camera battery. Also, most card readers support lots of different formats, so if you have more than one camera, then you need to carry only one card reader and a cable. Handling an Unreadable Card Sometimes, you might find that a card that has always worked fine in your card reader suddenly doesn t read. If this happens, put the card back in the camera and see whether you can view images on the camera s screen. If you can, then you know that the camera can read the card just fine. Plug the camera into your computer and try transferring the images that way. It will usually work. When you re done, format the card. This will usually get the card working with your card reader again. Transferring Images Manually Using Windows or a Mac 1. To transfer images manually: 2. Plug in the camera or card reader. 3. When the camera or card reader icon appears on your desktop, open it up. You should see a folder inside called DCIM. There might be some other folders, but DCIM is the only one you need to worry about. It is the standard location that camera vendors have agreed upon for storing images.
4 4. Inside the DCIM folder, you will find additional folders, usually named with some combination of numbers and the make of your camera (e.g. 100NIKON ). Depending on how many images you ve shot, there may be more than one. Open each folder and copy its contents to your desired location(s). Chapter 5: Image Sensors The real world in which we live is an analog world. Light and sound come to us as continuous analog waves that our senses interpret. Storing a series of numbers, on the other hand, is much simpler. Therefore, if you can find a way to represent something in the real world as a series of numbers, you can store those numbers very easily using your chose recording medium. The process of converting something into numbers (or digits) is called digitizing, hence the term, digital. The first step in digitizing is to divide your subject into distinct units. In the case of digital cameras, these units are called picture elements, or pixels. Your camera s image sensor is divided into a grid of pixels. When you take a picture, the sensor is exposed to light, and the light is sampled at each pixel in the grid. How fine your grid is (that is, what resolution it has) varies depending on the sophistication of your equipment. Next, each pixel in the grid is analyzed to determine its content. Each sample is measured to determine how full it is; that is, a corresponding numeric value is assigned that represents that pixel s contents which means, a representation composed entirely of digits or numbers. Finally, these numeric values are stored on some type of storage medium. If we want to record more than simple black and white images, we need to be able to specify more levels that is, we need to have more choices than just 0 or 1. By going to a higher bit depth such as 8-bits, which allows for 256 different values, we can record more information. The two factors that determine the quality of a digitizing process are the number of pixels you capture and the dynamic range (how wide a range of levels you have for each pixel). These days, most higher-end SLRS use CMOS chips. Both types of sensors register light in the same way, and for the sake of this discussion, the two technologies are interchangeable. The image sensor in your digital camera is a silicon chip that is covered with a grid of small electrodes called photosites, one for each pixel. Each photosite contains a photodiode and a capacitor along with other electronic components, depending on the type of sensor. When you turn your camera on, it places a uniform charge, or voltage, onto the capacitor at each photosite of its image sensor. When light strikes a particular photosite, it causes the photodiode to drain some of the charge from the capacitors. The amount that the charge is lowered is directly proportional to the number of photons that strikes the photodiode.
5 By measuring the reduction in voltage at a particular photosite, your camera is able to determine how much light hit that particular site during the exposure. Most cameras use either a 12-bit or 14-bit analog-to-digital converter, which means that the value from each photosite is converted into a 12 or 14 bit number. Note that an analog-to-digital converter with a higher bit depth doesn t give your sensor a larger dynamic range. Photosites are only sensitive to how much light they receive; they know nothing about color. To achieve color, some additional computing is required. The first successful color printing process, discovered in in 1903 by the Lumière brothers, was called autochrome, and used red, green, and blue dyes to color grains of starch that could be applied to glass plates to create color images. This process was dubbed autochrome. This story is not just a trivial history lesson. Understanding that your full color images are composed of separate channels will come in very handy later when you start editing. Very often, you ll correct color casts and adjust your images by viewing and manipulating individual color channels. You say black, I saw grayscale Each color model has a particular gamut, or range, of colors it can display. Some gamuts are more appropriate to certain tasks than others are, and all are smaller than the range of colors you eye can perceive. Each photosite on your camera s image sensor is covered by a filter red, green, or blue. This combination of filters is called a color filter array, and most images sensors use a filter pattern called the Bayer Pattern. With these filters, the image sensor can produce separate, incomplete red, green, and blue images. The images are incomplete because the red image, for example, is missing all of the pixels that were covered with a blue filter, and the blue filter is missing all of the pixels that were covered with a red filter. Both the red and blue images are missing the vast number of green-filtered pixels. The camera can calculate the color of any given pixel by analyzing all of the adjacent pixels. This process of interpolating is called demosaicing, and different vendors employ different approaches to the demosaicing process. By packing more and more photosites onto an image sensor, chipmakers can increase the sensor s resolution. However, there is a price to pay for this. To pack more photosites onto the surface of the chip, the individual sites have to be made much smaller. As each site gets smaller, its capability to collect light is comprised because it simply doesn t have as much physical space to catch passing photons. This limitation results in a chip with poor signal-to-noise ratio, that is, the amount of good date the chip is collection the signal is muddied by the amount of noise noise from the camera s electronics, noise from the other nearby electrical sources, etc.
6 In your final image, this signal-to-noise confusion can manifest as grainy patterns in your images. Turning Data into an Image Before the light from your lens ever strikes the sensor, it is passed through several filters. These include an infrared filter and low-pass filters. The infrared filter prevents false colors, which can be caused by infrared light that is not visible to the human eye. The low=pass filter helps prevent artifacts (such as moiré patterns) from being created during the digitizing process. Color Space conversion A color space is simply a mathematical model that can be used to represent colors. Some color spaces are larger than others, and so might allow for more variation in a particular color. Gamma Correction In an imaging chip, when twice as much light hits a single pixel, twice as much change in voltage is produced. In other words, the response of the pixels to light is linear---increase the light and you get a linear change in voltage. Your eyes don t work this way. When you increase brightness, your eyes register a logarithmic increase, rather than a linear increase. So a doubling of light does not make you perceive a scene as being twice as bright. The practical upshot of all this is that your eyes are able to see a lot of really fine detail in shadow and highlight areas. Because of their nonlinear nature, your eyes tend to expand the shadow and highlight areas, registering more subtle changes and allowing you to see more detail. In order to get accurate brightness values to expand the highlights and shadows so they appear more like what you eye is used to your camera applies a mathematical curve to all its brightness values. This is called a gamma curve, and can also be referred to as gamma correction. White Balance and Image Processing A digital image sensor cannot automatically adapt to different types of light. Instead, the image data from the sensor much be calibrated to the type of light you re shooting in, a process called white balance. Sharpening and Noise Reduction Many cameras perform some type of noise reduction algorithm to reduce unwanted noise in your image, and almost all cameras perform some type of sharpening in order to compensate for the softening caused by the low-pass filter. JPEG Compression and Saving
7 At this point in the process, your camera has processed the image and is ready to save it to the camera s media card. Before saving, it applies JPEG compression, to save space and to speed the write-time. All this process takes place as soon as your image has been shot, and as you might imagine, it can take a while to perform all these calculations. By take a while, I mean take a while in computer terms, but even though your camera s processor is very fast, it can still be overwhelmed with image processing. To counter this, most cameras include extra memory buffer that allows them to cache a few images for processing, freeing up the camera for immediate shooting. When saving in JPEG mode, the camera first coverts the image data from its original 12- or 14-bit format down to an 8-bit format, reducing the range of brightness levels from 4,096 or 16,384 all the way down to 256. Once the data is in 8-bit mode, the camera is ready to start compressing. Most cameras offer two forms of JPEG compression: a low-quality option that visibly degrades an image but offers 20:1 compression, and a high-quality option that performs a good amount of compression usually around 4:1 but without severely degrading the image. JPEG compression works by exploiting the fact that human vision is more sensitive to changes in brightness than to changes in color. To JPEG-compress an image, your camera first converts the image to into a color space where each pixel is expressed using a chrominance (color) value and a luminance (brightness) value. Next, the chrominance values are analyzed in blocks of 8x8 pixels. The color in each 64-pixel area is average so that any slight (and hopefully imperceptible) change in color is removed, a process known as quantization. Note that because the averaging is being performed only on the chrominance channel, all the luminance information in the image the information your eye is most sensitive to is preserved. After quantization, a nonglossy compression algorithm is applied to the entire image. In the very simplest terms, a nonlossy compression scheme works something like this: rather than encoding AAAAAABBBBBCCC, you simply encode 6A5B3C. After quantization, the chrominance information in your image will be more uniform, and larger chunks of similar data will be available, meaning that this final compression step will be more effective.