Medium-Format Cameras

Achieving 50MP and higher resolution numbers requires a type of camera called a medium-format camera. Before digital photography, the term medium-format cameras referred to cameras that used film wider than 35mm film.

When using a medium-format camera for digital photography, you typically have a couple of options.

Digital camera back. You sometimes can adapt a medium-format film camera by adding a digital camera back, which contains the sensors and electronics required for digital photography. The digital camera back often replaces the film holder on the camera. Obviously, having the ability to switch back and forth between film and digital photography with one camera body is an advantage to the professional photographer.

Integrated system. You also can purchase a new medium-format digital camera that’s integrated in a complete system, containing the sensor and electronics required as part of the camera body. With an integrated solution, a professional photographer can have a camera designed to work as one system, theoretically improving speed and functionality.

Regardless of which medium-format solution you choose, the CCD sensor that records the image is the key to the amount of resolution you’ll have.

The CCD sensors in medium-format digital cameras and in digital camera backs appear in different sizes and with different numbers of pixels, achieving different resolutions. For example, the Kodak 50MP sensor measures around 49.1 by 36.8mm (1.93 by 1.45 inches) and is much larger than those in consumer-level cameras. (Some professionals use large-format cameras, which have film sizes or digital sensors measuring 4 by 5 inches or larger and offering up to several hundred megapixels of resolution.)

The New Standards

At the time of Kodak’s announcement of its 50MP sensor in July 2008, the new sensor replaced the medium-format camera industry’s previous highest resolution sensor, a 39MP unit, also from Kodak. A couple of weeks later, Phase One surpassed Kodak with a 60.5MP sensor it developed with DALSA Semiconductor. Leaf later announced a 56MP wide-frame sensor, again developed with DALSA Semiconductor.

The flurry of high-resolution sensors and related cameras equaling and surpassing the 50MP barrier thrust medium-format digital cameras into the spotlight.

“State of the art [was] at 39 million pixels,” Kodak marketing manager for Image Sensor Solutions Michael DeLuca wrote in his blog at the time of Kodak’s 50MP sensor announcement. “But in this market, image is everything. So, having the pixels and the performance you need is critical to capturing exactly the shot you want. So, now, 50 is the new 39.”

Kodak KAF-50100 and Hasselblad H3DII-50. A few days after Kodak announced its 50MP CCD sensor, Hasselblad announced that it would use the new sensor in its H3DII-50 medium-format digital camera, which will carry a price tag of about $40,000.

Kodak’s sensor offers an 8,176 x 6,132 pixel array. The KAF-50100 features pixels measuring 6 x 6 microns, smaller than the 6.8-micron pixels found in Kodak’s 39MP sensor.

At September 2008’s Photokina trade show in Germany, Hasselblad announced plans to release an H3DII-60 medium-format digital camera that will offer 60MP of resolution sometime this year. (During the initial announcement, details concerning the 60MP sensor, including the manufacturer, were not available.)

Phase One P 65+. Phase One’s 60.5MP sensor, called the P 65+, measures 53.9 x 40.4mm (about 2.1 x 1.6 inches). The P 65+ has an active pixel array of 8,984 by 6,732. Phase One will offer the P 65+ in both a digital camera back (about $40,000) and in a digital camera system (about $42,000).

“Photographers need real reasons to upgrade past 39 megapixel digital backs,” said Henrik Håkonsson, CEO of Phase One. “Real value includes higher resolution but also requires new functions, faster operation, higher quality through expanded sensitivity, increased dynamic range, better results in the studio or on location, and a better longer-term investment. I believe we are able to achieve this and more.”

At Photokina, Phase One also announced a marketing deal with Leica, which, at the same time, announced plans for a 37.5MP digital camera that it calls a hybrid between digital SLR and medium-format. The Leica S2 contains a 45 x 30mm sensor from Kodak.

Leaf AFi-II 10. Leaf’s AFi-II 10 medium-format camera system incorporates a 56MP CCD sensor. Leaf has two other AFi-II camera systems: The 7 (33MP) and the 6 (28MP). The company offers corresponding digital camera backs, called the Aptus-II 10, Aptus-II 7, and Aptus-II 6, all of which feature similar sensor resolutions.

The CCD sensor in the AFi-II 10 measures 56 x 36mm (about 2.2 by 1.4 inches) with a pixel array of 9,288 x 6,000. With its wider sensor, the AFi-II 10 offers a wider image capture than most medium-format cameras can achieve.

The AFi-II 10 will cost about $43,700.

Although the 56MP of resolution in the Leaf/DALSA sensor draws initial attention to the AFi-II 10, Leaf Director of Marketing Seth Greenberg says the camera’s other features keep customers’ attention.

“Fifty-six megapixels is the first thing people ask about,” he says. “But, for us, it’s just something that we can do. We don’t want to get embroiled in a megapixel war. What really sets us apart are the features and what we do with that resolution.”

Finding Fifty

For professional photographers, the recent flurry of digital cameras and sensors with 50-plus megapixels of resolution was welcome news. Sony announced recently the largest resolution for a D-SLR camera at 24.6 megapixels, which remains less than half the resolution available in the newest digital camera backs and medium-format cameras. Leaf’s Greenberg says he was impressed with the detail shown in the various prints displayed at the Photokina trade show from the Leaf medium-format cameras.

“The subtleties of light and shadow create some amazing images,” he says. “It’s an image that our [medium-format] cameras can achieve. You can’t achieve it with the mega-D-SLRs.”

As far as when less expensive cameras might see some of the technologies found in the new medium-format sensors. . . well, let’s just say unless you can fast-forward several years, you’re going to have to wait a while.

In fact, Kodak’s Antonio Ciccarelli, the company’s worldwide marketing manager for Image Sensor Solutions, says the primary technology, the sensor, probably will not trickle down to consumer-level cameras anytime soon because medium-format digital cameras are using CCD technology, and consumer-level and D-SLR cameras are transitioning to CMOS technology in the sensors. CCD’s superior image quality over CMOS means medium-format cameras probably will continue using CCD for several years. One advantage of CMOS over CCD—processing speed—isn’t as important in medium-format cameras. CMOS might eventually replace CCD in many types of cameras, but not for medium-format cameras, at least in the near future.

“CCD remains in very niche-type applications that require the utmost quality,” Ciccarelli says. “CMOS has made some great strides.”

In the end, some technologies found in the top-of-the-line models will trickle down to lower-priced cameras, giving photographers more features for less money.

That doesn’t mean you’ll find 50MP of resolution in a pocket-sized camera for $400 in a decade, but it does mean the advancements in digital camera technologies are likely to continue, which is good news for all levels of photographers.

by Kyle Schurman

Kodak's Pixel Kodak’s 50MP sensor makes use of a smaller pixel, the 6.0-micron Truesense pixel, than the company’s previous generations of sensors (39MP and 31MP). However, the 6.8-micron pixels in the 39MP sensors used a similar basic design, as shown here. In the illustration at left, you can see a top view of a 6.8-micron pixel. Each pixel contains two sections, an ITO gate electrode and a polysilicon (Poly) gate electrode. Either gate can collect light, although the ITO gate is more sensitive to shorter light wavelengths. B1, B2, and B3 represent barrier regions around the buried channel (Bch). The buried channel confines the signal in the electrodes during exposure and readout. The channel stop (Chst) area isolates the pixel horizontally, and the lateral overflow drain (LOD) carries any excessive signal away from the electrodes. With smaller pixels, as used in the 39MP and 50MP sensors, the LOD is important for allowing larger charge capacities without affecting blooming control. (Blooming control refers to the pixel’s ability to record a bright light source that saturates the pixel without bleeding electrons to nearby pixels that would negatively affect image quality.) The illustration at left shows a 3D representation of the pixel in terms of voltage potential. V1 marks the polysilicon gate electrode, and V2 marks the ITO gate electrode. As compared to the 6.8-micron pixel shown here, the 6.0 Truesense pixels offer advantages in several feature areas, according to Kodak’s DeLuca, including a four-channel output design for faster frame rates, lower power requirements, shorter lag time between the shutter click and the image capture, and use of a new red filter pigment for truer color accuracy. "The new [50MP] devices are different because they are built using the Kodak Truesense 6.0-micron full-frame CCD platform," DeLuca says. Source: Kodak

Kodak's Four-Channel Design The new Truesense 6.0-micron Full-Frame CCD Platform technology from Kodak offers several interesting features for the 50MP Kodak sensor, beyond its 6.0-micron pixels. With Truesense 6.0, Kodak sought to increase the maximum resolution its sensor could provide without negatively affecting already existing features, especially in the area of dynamic range (signal-to-noise ratio). Because the 6.0-micron pixel in the 50MP sensor is almost 12% smaller than the 6.8-micron pixels used in Kodak’s 39MP sensor, each pixel in the 50MP sensor provides less signal than each pixel in the 39MP sensor. Consequently, Kodak had to reduce the noise to retain the desired dynamic range. The 39MP sensor used a two-channel readout, while the 50MP sensor uses a four-channel readout. The four-channel setup gives Kodak more bandwidth with which to move data, allowing the sensor’s amplifiers to run at a slower pace (18MHz with the 50MP sensor, vs. 24MHz with the 39MP sensor), which reduces noise. In the four-channel readout, two rows can reach the output register at one time, with half of the data going right and half going left. In the two-channel readout, only one row can reach the output register at one time. The 50MP and 39MP sensors are essentially the same physical size, 49.1 by 36.8mm (1.93 by 1.45 inches). Source: Kodak

Development Of Digital Camera Backs Leaf developed the first digital camera back for medium-format cameras in 1991, and it offered 4MP of resolution. Because the technology didn't yet exist to move such large amounts of data efficiently, the early digital camera backs often used an external cable to connect to a computer for data download. Early models only shot in black and white, too, requiring the use of color filters and a trio of images to create color prints. Digital camera back technology has advanced significantly in the past several years, especially in terms of resolution. As you can see from this example, the digital camera back (the area containing the display screen) fits onto the back of this medium-format camera. Some of the milestones in development of digital camera backs include those listed here. 1991. First digital camera back, produced by Leaf, offers 4MP of resolution. It’s called the DCBI (Digital Camera Back I). 1992. MegaVision introduces the T2 digital camera back, which offers 4MP of resolution. 1993. Phase One is founded, and the company begins selling digital camera backs within a few years. 1994. Leaf introduces DCBII, which remains at 4MP of resolution but offers other improvements. 1998. Leaf offers the 6MP Volare model. 2003. Kodak introduces the 16MP Pro Back Plus. Leaf quickly followed with the 22MP Valeo. 2004. Mamiya begins offering the Mamiya ZD, which is an all-digital medium-format camera that doesn’t require a digital camera back. 2004. Kodak purchases Leaf, making Leaf a subsidiary of the larger company. 2005. Digital camera backs offering up to 39MP of resolution begin appearing, including the P 45+ from Phase One (pictured here). 2008. A few different manufacturers announce plans to offer 50MP or more of resolution in digital camera backs and in medium-format digital camera systems. Sources: Phase One, Leaf, Wikipedia

CCD vs. CMOS The medium-format image sensors discussed here all use CCD technology. CMOS technology is becoming more popular with consumer-level and digital SLR cameras, however. Both types of image sensors consist of pixelated metal oxide semiconductors. Each pixel (orange squares in these illustrations) accumulates charge proportional to the intensity of the light that strikes the sensor. CCD CCD. Bell Labs developed CCDs (charge-coupled devices) in the late 1960s. The original intent for CCDs involved temporary data storage, but RAM chips became the industry’s preferred method. CCDs received a new assignment, proving to be a good choice for recording digital images. CCDs initially appeared in scanners and TV cameras in the mid-1970s. Digital cameras featuring CCD sensors first appeared in the late 1980s. Although CCDs yield tremendous image quality, they can be more expensive to create than CMOS sensors because they use a different manufacturing process than other computer chips. CCD sensors also require more chips and electronics than CMOS sensors, causing them to use more power, to require more space, and to cost more in manufacturing. Finally, CCDs operate more slowly than CMOS sensors. Despite some drawbacks, CCD image quality is outstanding, and CCD sensors appear as though they’ll remain the only technology used with medium-format digital cameras. For professional photographers creating extremely high resolution prints, image quality is far more important than speed or sensor cost. In the illustration here, the CCD image sensor handles the photon-to-electron conversion and electron-to-voltage conversion processes. The charge packet from each pixel shifts from square to square before reaching the output register (bottom row of orange squares). The output register passes the data to the circuit board. This passing of data is one reason why CCD operates a little more slowly than CMOS. Other functions then take place on the circuit board to finalize the image data. CMOS CMOS. CMOS (complementary metal oxide semiconductor) is the technology behind the manufacturing process used to make the sensors. Because of advances in manufacturing technologies in recent years, it has become easier to make CMOS sensors in volume. Rather than requiring extra chips and electronics to perform various functions, manufacturers can incorporate those items directly into CMOS sensors. Because they cost less, provide good speed, and generate decent image quality, CMOS sensors are appearing with more frequency in all consumer-level digital cameras, especially high-end digital SLR (D-SLR) cameras. In the illustration here, each pixel in a CMOS sensor performs the full photon-to-electron-to-voltage conversion itself. The analog-to-digital conversion takes place on the CMOS sensor as well. With all of the image functions on the sensor, CMOS has a speed advantage over CCD. Sony announced its development of a 24.8MP full-frame, 35mm format CMOS sensor early in 2008. Sony has aimed its sensor at digital SLR models. Canon also has surpassed the 20MP barrier for a CMOS digital SLR sensor with its 21.1MP unit. Source: DALSA Semiconductor