Consumer vs. Scientific Digital Cameras for the Life Scientist

Brian L. Kuyatt, M.S. Application Support Engineer, Diagnostic Instruments, Inc. Philip Merlo, M.S.B. Vice-President-New Business, Diagnostic Instruments, Inc. As fun as taking pictures may be, it is the professional presentation and analysis of scientific images that is important to scientists.

This article discusses a few key concepts of digital cameras, optics and associated software to illustrate scientific and economic differences between consumer and scientific digital cameras for the life scientist. The scientific digital camera market includes not only the broad spectrum of scientific research but also the documentation markets in the life sciences and medicine (e.g., pathology and forensics).

We have all seen the recent improvements in consumer digital cameras. They have gone from low resolution, curiosities to very capable, high resolution, point and shoot cameras. The prices have dropped dramatically making them even more enticing. Do these cameras have a place in the scientific market? The answer may be "yes" or "no" depending on your intended use. What are the basic characteristics of consumer digital cameras?

The resolution of these new cameras is typically 2 to 5 megapixels or 2 to 5 million total pixels in each image. They are portable and can take pictures without being attached to a computer. Their prices are low, ranging from $300 to $1,500 and they can be purchased through any catalog or electronics store. Why would anyone want to buy a scientific grade camera when these economical choices are available?

Let’s review these features in more detail and see how they apply to the life scientist. After evaluating the scientific issues in camera selection, we will also consider the actual time taken to capture images and the related issue of the true cost of this image capture. These last two constraints will be evaluated at the end after considering the scientific issues.

Resolution
The basic element of an image is the pixel. It is defined as the smallest picture element in an image. In the film medium it is the grain size of the chemicals in the film. In modern video and digital imaging cameras it is the individual, light sensitive, silicon structure within the sensor array. The number of pixels that span the image determine the resolution of the camera. Knowing this, it can generally be said, that the higher the number of pixels that span the image the better the image quality.

The ability to resolve small features with more pixels allows you to enlarge and examine these features in more detail. Maximising the resolution in the final image is an important goal in any image system. It should be noted that each component in your imaging system has the power to degrade the resolution from theoretical maximum.

"More pixels the better" is a generalisation that has constraints. The optimum solution is one that matches the resolving power of each component in the system. Constraints on the resolution in an imaging system are the microscope objective characteristics, optical coupler quality, and CCD chip size in the digital camera. If your intended use is for microscopy, the microscope optics may be the limit to your system’s resolution. Depending on the grade and magnification of your objectives, the resolving power of the system can vary.

Mathematically calculating the resolving power of each microscope objective is an important first step in determination of the critical image resolution needed by the life scientist. Attributes of the microscope objective that restrict resolution are the magnification-numerical aperture limit, the type of objective and the quality of manufacturing.

Magnification and numerical aperture deal with the fundamental physical principals that limit the magnification and resolution possible in light microscopy. The type of objective design influences the maximum feasible NA. Each design allows different amounts of chromatic and geometric aberrations. For light microscopy, these designs commonly include achromat, plan achromat, plan fluorite, and plan apochromat. In general, the achromats are the least optically corrected objectives and the apochromats are the most corrected objectives.

Knowing the NA, objective magnification and optical aberrations associated with each type of microscope objective, we can calculate the resolving power for each microscope objective. Included in these calculations it is also important to include the optical coupler magnification and characteristics. Objective resolving power in light microscopes is the primary determinant of the theoretical limit of resolution capacity. The following table shows the resolving power of certain microscope objectives in the intermediate image for 0.63x and 1.0x optical coupling adapters in combination with a typical 2/3" CCD (8.5 mm x 6.4 mm).

Fig. 1. Camera Resolution Table Determining Microscope Objective Resolving Powers

A closer look at the chart indicates that at high objective magnifications (above 40x with 1.0x adapter or above 25x with 0.63x adapter) the resolution is typically constrained by the microscope optics and at low magnification the resolution is typically constrained by the camera image sensor or CCD resolution. At high objective magnifications it is possible to capture images that are of higher resolution than actually necessary because the resolving power of the objective is limited by the wavelength of light itself in the calculation.

This is commonly referred to as "empty pixel" or just "empty" resolution. Scientists purchasing a digital camera with CCD sizes greater than the resolving power needed with the high magnification objective are literally wasting their money because the resolution of small objects has already reached its limit.

An example of this principle can be demonstrated using a 20x/0.5NA objective and either a 1.0x or 0.63x optical coupler. With the 1.0x coupler, the required camera resolution to obtain the highest resolving power of that objective is a 1275 x 969 pixel CCD resolution.

In the first two digital images (Figure 2) of bone marrow below, the one on the left has only a 1024 x 768 pixel resolution while the one on the right has 2048 x 1536 pixel resolution. At normal digital zoom levels, small objects (red and white blood cells) appear to be resolved properly. When the same images are zoomed to 400% (Figure 3) the cells are resolved only in the image only on the right with the appropriate resolution matched to the resolving power of the objective and coupler.

Fig. 2. Bone Marrow (20x/0.5NA—1.0x coupler); Left Image--1024 x
768 pixels; Right Image—2048 x 1536 pixels; cells appear normally
resolved at normal zoom level.

Fig. 3. Left Image—1024 x 768 pixels (400% Zoom); Right Image—
2048 x 1536 pixels (400% Zoom); at higher zoom, only the image
on the right is properly resolved.

At this point you need to determine the purpose of the magnification and resolution. Typically, low magnification objectives are used for gross morphology that does not require higher resolution. Further, the low power objectives do not have adequate numerical aperture to fulfill the requirements of high resolution provided by the high magnification objectives. Therefore, the maximum resolution that you require for your work should be a deciding factor when purchasing a digital camera.

A good economic strategy for outfitting your imaging
system (for best resolution, on a limited budget) is to determine the one or two magnifications that you will commonly use for image capture and then purchase plan fluorites or plan apochromats for these magnifications along with a digital camera that does not constrain the resolution at these magnifications. Then in the future you can upgrade the other objectives and accessories.

In a related economic issue, resolution affects your image storage capacity and the time required for image manipulation; the larger the resolution of the image captured, the larger the image storage requirements. This is not just a linear relationship, doubling the linear number of rows and columns in an image quadruples the file size (This is an area calculation: rows x columns = area). This can be taxing on your storage capacity and if this resolution does not benefit you in any way, it just fills your hard drive.

For example, a 5 megapixel camera produces a 15 megabyte 24 bit RGB file and 1 Gigabyte of storage can only accommodate 66 of these images. Working with large files should also be considered. The time needed to open an image over a network, manipulate the image in an application and then save the image back out, all increase as the square of the linear resolution. Trying these operations out on your computer equipment is an important test to see if the resolution benefit is worth the time and storage space.

Optical Coupling

Between the microscope and the camera in the image system is the optical coupler. It can be an image quality bottleneck if it is not correctly selected.

The main function of the optical coupler is to properly size the image from the microscope to fit the camera’s image sensor. The optimum magnification coupler will project a large enough image (the circular field of view from the microscope) to cover the rectangular image sensor, but will not enlarge it to the point that only a small section of the center of the image lands on the image sensor.

The first consideration in the selection of a coupler is whether it was designed for high-resolution cameras. Some base line couplers optimise the cost vs. resolution balance for video systems, which have only 1/2 to 1/4 the resolution of modern digital cameras. The result could be an expensive plan apochromatic microscope objective and high-resolution digital camera crippled in resolution by an inappropriate optical coupler. Make sure that the coupler you buy is designed to handle the objective and camera you are purchasing.

Reflections within the optical system reduce contrast and color saturation. For these reasons, use of coated optics throughout the optical system is strongly suggested. Lens coatings work by reducing reflections off lens surfaces. Inexpensive optical adapters with uncoated or poorly coated lenses can only cause the capture of poor grade images. Also, due to the small number of adapters sold for consumer grade digital cameras, the cost of all these adapters tends to be much higher than for the scientific grade cameras.

Depending on the microscope used and quality of the adapter, an adapter for a consumer grade camera can cost as high as $800 to $900; almost as much as some of the cameras themselves.System Cost Let’s look at the system cost for putting a consumer camera on a microscope. A 2 to 3 megapixel camera costs in the range of $300 to $600, with charger, spare battery and larger memory card--$100, a card reader for the memory card--$100, a quality optical coupler to attach it to your microscope--$480 to $680, and software to manipulate your images--$600 (Adobe PhotoShop) to $1000 (Media Cybernetics ImagePro Express).

The total cost of these systems would then range from $1,580 to $2,500. The typical system cost for a scientific camera would be as follows: Camera--$4,000 to $8,000; Coupler-- $80 to $300; Software--$0 to $1,000; this results in a camera cost of $4,080 to $9,300. When considering the purchase price alone, the consumer camera seems like the economical choice. However, this does not take into account the cost of using the camera, that is, the time involved in actually capturing images.

Ease of Use/Productivity

Have you tried focusing and framing your specimens with the consumer cameras? The small LCD screens greatly limit your ability to visualise what you are capturing (especially if the LCD screen is located on the back of the camera). Changing settings on the camera may entail removing the camera from the microscope and running through the menus on the LCD screen. To transfer images to the computer requires unplugging the memory card from the camera and plugging it into the computer and then reading it.

The other method of transferring images requires turning off the camera and connecting the transfer cable then restarting the camera and going into dump mode (many cameras won’t acquire images while tethered to the computer). Often, after all this, you realize the image is not quite what you wanted. The process then starts all over, unplug, turn off camera, plug card into camera, turn on camera, get back into mode you want, frame and focus, capture image, turn off camera, remove memory card, plug into computer, load into application program and view. Time quickly slips by.

The scientific cameras are setup to allow real time viewing right on your computer screen. Focusing, framing, image adjustment, and near instantaneous image acquisition provide a quick, reliable and productive system to work with. Most software packages include powerful features such as auto white balance, auto exposure, image enhancement, annotation, embedding of a calibration mark and measurement tools.

Other packages may also incorporate presentation modes, sequential imaging, image notation, image archiving, report generation, print layout tools, and image compression options. A scientific camera that can capture multiple, high resolution images per minute, easily out performs the consumer camera with its cumbersome interface that can only capture multiple images per hour.

Software with Macro Utility

To further improve the speed and ease of use, users should strongly consider software with a macro utility that allows the user to automate image capture and processing functions. Automation by macros, utilising user-created standard camera setup features, can increase the speed of capturing high quality images by an additional 20% or more.

Most software packages offered with digital cameras are simple and do not have these automated macro functions. Macro functionality for use in digital image capture and other image processing is primarily offered by software companies that charge $1,000 or more for their software but Diagnostic Instruments, Inc. offers free SPOT software package with all their cameras that contains a macro utility for ease of image capture and processing (see macro examples in Figure 4).

Fig. 4. Macro Function in SPOT® Software

The Economic Issue--How much is your time actually worth?

From the scientific standpoint, scientific grade digital cameras are the cameras of choice but what about the pure economic issues in a purchase? From the standpoint of the time involved in capturing images, the differences in price between consumer and scientific digital cameras are diminished even further. If we use a typical example below to illustrate the true cost of image capture. This is typicallyreferred to as one of the price/performance issues in a digital camera purchase.

Example: 48 images to document

Consumer Camera
1 image per 10 minute = 6 images /hour 48 /6 = 8 hours work

Scientific Camera
1 image per 2 minutes = 30 images / hour 48 / 30 = 1.60 hours work

Savings
8 hours - 1.60 hours = 6.40 hours x $ 50 / hour = $ 320/Day
4 days per week x $ 320/ day = $ 1280 week

Maximum Weeks until break even
$9,800 — $1,580 = $8,220/ $1,280 = 6 .4 weeks

It is obvious from this example that it only takes a month and a half to make up the purchase price difference between a consumer and
scientific grade digital camera. The lesson—don’t be short-sighted when purchasing a digital camera based solely on price.

Support

One last point that should be considered is the fact that the scientific digital cameras are supported by your local microscope dealer. These are imaging professionals that provide technical support for the scientific cameras but not for the consumer cameras. Getting setup and capturing images quickly the first time can be invaluable. Unless the scientist has sufficient experience and knowledge to deal with all the issues in a digital camera purchase, wasted time and resources can result.

Summary

In summary, don’t be dazzled by the resolution numbers and the low up front price of consumer cameras. They are great for general point and shoot "picture taking" but when it comes to the specialised needs of the life scientist, the user should purchase a digital camera that addresses the actual resolution requirements, while not compromising the economics and productivity related to scientific research. The term caveat emptor takes on an expanded meaning when considering the purchase of a digital camera for life science research.

Note—Diagnostic Instruments, Inc. (DI) is the leading manufacturer of optical couplers between cameras and microscopes and in the manufacture of boom stands for research. DI also is a major manufacturer of scientific digital cameras (Spot® and Insight® cameras) for the life science and industrial markets