Hank Hogan, Contributing Editor
The goal of all microscopists is simple: to see what they want in
a sample. This leads to different camera requirements for different applications,
but advances in the technology have it covered.
There is more going on in microscope cameras than meets the eye. Digital interfaces, as
much as the demands of the applications, promise to deliver higher camera resolution
and user-friendly operation. New CCD sensors under development should yield cameras
of greater sensitivity and lower cost. CMOS sensors, on the other hand, still seem
to be the technology of tomorrow — or maybe the day after.
Exploiting the advances in CCDs, vendors are crafting
better cameras for the scientific microscopy market. And manufacturers are creating
lower-cost, lower-performance and higher-frame-rate cameras for industrial applications.
General trends
“There has been a whole new wave of cameras
this past year that have broken that 10-frame-a-second barrier, and that’s
made a huge difference,” said Brian Craze, manager of electronic imaging at
A.G. Heinze Inc., a distributor in Lake Forest, Calif. Heinze markets microscopes
from Nikon Inc. of Melville, N.Y., as well as imaging systems from a number of
other companies. The advent of digital cameras that have near-video frame rates,
he said, has led to significant changes in microscopy.
Digital microscopy on
the factor floor requires much less from the camera. A lower list price and ease
of use will beat out more advanced (and expensive) cameras. Viewed under polarized
light at 50x, cast commercial-grade titanium looks like modern art. Courtesy of
Diagnostic Instruments Inc.
Increasingly, industrial users are
moving away from analog, video-based systems and associated printers to the digital
camera and image-archival solutions now available to them. Craze cited recent products
from Nikon, Diagnostic Instruments Inc., Leica Microsystems AG and Optronics as
examples of industrial systems. Diagnostic Instruments also makes cameras for the
scientific market. Others in this arena include Hamamatsu Corp., Roper Scientific
Inc. and The Cooke Corp.
Lab researchers still demand less noise,
better resolution and wider wavelength coverage from their cameras to see what they
want. Here, fluorescence imaging reveals microtubules (green), actin filaments (red)
and cell nuclei (blue) in bovine cells. Courtesy of Diagnostic Instruments.
Like Diagnostic Instruments, Sony Corp.
makes microscope cameras for both industrial and scientific applications. But unlike
many camera companies, Sony makes the sensors that are at the heart of digital imaging.
Indeed, Sony’s CCDs show up in many companies’ cameras. As such, Sony
has a unique perspective on the general market demands.
Advances in CCD technology are leading to the development of better
microscope cameras for scientific applications and to their adoption on the factory
floor, albeit at a lower level of sophistication. Courtesy of The Cooke Corp.
“The trend is not so much increasing
the resolution but taking the usable resolutions now and building cameras at a lower
cost,” said Paul Kasparian, a marketing manager at Sony Broadcast and Professional
Co. in Park Ridge, N.J. Many applications demanded sensors with 1 million picture
elements. Beyond the megapixel barrier, however, it is unclear whether adding elements
enables the cameras to produce better images.
Reducing camera cost can come about
in a variety of ways, said Patrick Merlo, president and head of research and development
at Diagnostic Instruments in Sterling Heights, Mich.
“Larger, 1-in.-format CCDs allow
the use of microscope couplers without lenses, since the image does not have to
be reduced to fit a small CCD,” Merlo said. “This results in a lower-cost,
more compact coupler.”
Better ins …
Further improvements to the CCD sensor and to
the camera interface promise to affect cost, resolution and other camera characteristics.
On the sensor side, some semiconductors under development could have a significant
effect.
Mark Christenson, senior scientist
in life sciences at Roper Scientific in Trenton, N.J., said that work on new sensor
materials is under way at Texas Instruments Inc. in the US and Marconi plc in
the UK. These separate product developments aim to amplify the CCD’s signal
without using an external intensifier. The amplification could be on the order of
100 to 1000, depending on a variety of factors.
For microscope camera makers, such
an improvement would enable a number of product enhancements, the least of which
is higher resolution. Christenson noted that intensifiers add cost to the camera
system and are a source of failure, such as when they are exposed to too much light.
“They can burn out and get burn
spots on them … eventually, whereas CCDs typically don’t have the same
kind of lifetime restrictions,” he said. Moreover, the use of an intensifier
limits the available imaging formats.
But a number of problems must be solved
before these CCDs appear in commercial products. For example, the pixels on the
current crop of these sensors must be reduced from 20 μm or more on a side
to less than 10 μm. Although such chips are in the pipeline, camera makers
can’t start incorporating them in their products until they are real, available
hardware. It will be at least a year or two before cameras based on the new CCDs
hit the market.
s for CMOS, there is a great deal
of interest — but little action. The consensus seems to be that CMOS sensors
offer some potentially important improvements over CCDs, but that their image quality
and other characteristics render them undesirable at present.
Karl Kilborn is co-president of system
integrator Intelligent Imaging Innovations Inc. in Santa Monica, Calif. Like many
others, Kilborn said that his company has not used CMOS technology. “We are,
however, very interested in the technology and look forward to improvements,”
he said. “The potential for random access of individual pixels is particularly
interesting.”
… And outs
Regarding the connection between a camera and
a host computer, the trend is toward the elimination of old standards and the adoption
of Apple Computer Inc.’s FireWire (known generically as IEEE-1394) and/or
the USB 2.0 interface. The 400-Mb/s FireWire has generated much interest and is
available on a number of components. At 480 Mb/s, USB 2.0 has a slightly higher
throughput, and it may end up being just as popular, particularly because it supports
first-generation USB devices. For camera makers, this may mean that they must support
both standards.
The microscope camera industry is just
entering this interface debate, said Jerry Fife, product manager for industrial
and scientific cameras at Sony. Currently, Sony sells far more analog than digital
cameras for microscopy. Analog cameras require frame grabbers for transferring images
to a computer, and they may be limited to 640 x 480-pixel video resolution —
much less than most microscopy applications need.
A faster interface would lead to increases
in camera resolution. With higher transfer rates and decreasing storage costs, users
will be able to manage more data, and some cost and performance barriers to higher
resolution may disappear.
Until then, camera resolution is subject
as much to what users can handle as it is to what they need for their applications.
“The bulk of the volume is still in the half-a-megapixel to the one-and-a-half-
to two-megapixel range, just because [there is] too much data to move around,”
Fife said.
Eight is enough
A consequence of the advent of digital cameras,
Heinze’s Craze said, has been the deployment of the systems in industrial
microscopy applications. In many cases, they are replacing color photography, taken
as part of in-line quality control during the manufacture of semiconductors, print
heads, precision machined parts, electronic components or other equipment.
It’s an application with very
specific requirements. “These people simply want a high-resolution, megapixel
digital camera,” Craze said. “They use PCs exclusively. They want to
capture an image and put it in a report. That’s about all they want to do.
Every once in a while they want to do a point-to-point measurement.”
The industrial market requires megapixel
or greater resolution, image rates in the range of 30 fps and full color, but it
does not need a camera with a wide dynamic range. Eight bits is often enough, compared
with 12 bits or more for scientific applications.
In the past, the only electronic replacement
to the photograph for industrial applications involved video cameras, a television
monitor and a video printer. Today, digital cameras — with their simplified
interface and, at most, two button controls — are being increasingly used.
It is a market that has attracted Heinze’s attention within the past few years
and is catching the attention of camera vendors.
“Our future line of lower-end
cameras would probably be more applicable to the industrial market, since this market
is not as demanding when it comes to sensitivity or higher than 10 to 12 bits of
dynamic range,” said Murad Karmali, sales manager at Cooke’s imaging
division in Auburn Hills, Mich.
Merlo calls industry “the qualitative
market” because of its need for pretty pictures. He, too, described the trend
away from video cameras and toward digital cameras with real-time modes. He noted
that the camera requirements are not exceptionally demanding in a technical sense
but that the pressures on price and ease of operation can be quite high.
“This marketplace benefits greatly
from increased spatial resolution combined with lower list prices,” Merlo
said. “This marketplace requires intuitive, easy-to-use software.”
Consequently, the technically superior
product may lose in the industrial arena to one that is less expensive and easier
to operate. Moreover, the interface to the system is secondary to these applications
— the interface to the human operating the camera is of overriding importance.
In the lab, more of the same
In scientific applications, however, users ask
for more of what they have: reduced noise, higher quantum efficiency, higher speed,
wider or multiple wavelength coverage, and better resolution.
Roper offers products for these applications
that feature 6.45-μm pixels for higher resolution. Other cameras in the company’s
product line combine higher sensitivity with higher frame rates by focusing the
image sequentially on different parts of the CCD.
Diagnostic Instruments has a number
of new cameras for the scientific market, as does Cooke in its SensiCam line. Cooke’s
12-bit cameras display low noise, thanks to cooling, extensive binning, on-chip
region-of-interest selection capability and the use of a relatively new enhanced-quantum-efficiency
CCD from Sony.
For its part, Hamamatsu has introduced
a line of electron-bombardment cameras for scientific applications that offer signal
amplification of 1000 times or more. The process starts with the electrons that
are generated when a photon strikes a photocathode. A voltage accelerates these
electrons and drives them into a back-thinned CCD.
Electron-bombardment cameras offer signal
amplification of up to 1200 times for scientific applications. An applied voltage
accelerates the electrons released from the photocathode into a back-thinned CCD.
Courtesy of Hamamatsu Corp.
Fundamentals never change
No matter the technology improvements, the fundamental
aim of microscopy will not change — users need to see what they are looking
for.
James Lechleiter is an associate professor
of cellular and structural biology at the University of Texas’ health science
center in San Antonio and the director of the center’s optical imaging facility.
Much of the work there is done using photomultiplier tubes; some involves microscopes
and cameras.
Lechleiter and microscopists of all
stripes are looking for a signal in a welter of noise. Success comes down to being
able to identify and to focus on the desired object.
“The more signal that I have,
the faster I can go, the higher resolution I can go,” Lechleiter said. “So
that’s really what it comes down to.”
“And I want to do it cheaply, too,” he added.