Cut here cut there paste here… enjoy the show!
***
An image sensor is a device that converts an optical image into an electronic signal. It is used mostly in digital cameras and other imaging devices. Early analog sensors were video camera tubes, most currently used are digital charge-coupled device (CCD) or complementary metal–oxide–semiconductor (CMOS) active pixel sensors.
CCD vs CMOS
Today, most digital still cameras use either a CCD image sensor or a CMOS sensor. Both types of sensor accomplish the same task of capturing light and converting it into electrical signals.
A CCD image sensor is an analog device. When light strikes the chip it is held as a small electrical charge in each photo sensor. The charges are converted to voltage one pixel at a time as they are read from the chip. Additional circuitry in the camera converts the voltage into digital information.
A CMOS imaging chip is a type of active pixel sensor made using the CMOS semiconductor process. Extra circuitry next to each photo sensor converts the light energy to a voltage. Additional circuitry on the chip may be included to convert the voltage to digital data.
Neither technology has a clear advantage in image quality. On one hand, CCD sensors are more susceptible to vertical smear from bright light sources when the sensor is overloaded; high-end frame transfer CCDs in turn do not suffer from this problem. On the other hand, CMOS sensors are susceptible to undesired effects that come as a result of rolling shutter.
CMOS can potentially be implemented with fewer components, use less power, and/or provide faster readout than CCDs. CCD is a more mature technologyand is in most respects the equal of CMOS.[1][2] CMOS sensors are less expensive to manufacture than CCD sensors.
Another hybrid CCD/CMOS architecture, sold under the name “sCMOS”, consists of CMOS readout integrated circuits (ROICs) that are bump bonded to a CCD imaging substrate – a technology that was developed for infrared staring arrays and now adapted to silicon-based detector technology.[3] Another approach is to utilize the very fine dimensions available in modern CMOS technology to implement a CCD like structure entirely in CMOS technology. This can be achieved by separating individual poly-silicon gates by a very small gap. These hybrid sensors are still in the research phase, and can potentially harness the benefits of both the CCDs and the CMOS imagers.[4]
What is a Backside Illumination CMOS sensor?
Backside illumination CMOS sensors refer to a new type of structure over previously used CMOS sensors. These sensors change the positioning of a wiring layer to reduce light loss which results in an increase of light sensitivity and a reduction of noise recorded on the image.
The image above shows a conventional CMOS design on the left and a Backside Illuminated CMOS sensor design on the right.
The conventional CMOS design has the Metal Wiring layer (indicated by the Blue square) positioned above the Photodiode layer (indicated by the brown square). Due to the positioning of the Metal wiring layer some light is reflected and therfore is lost
The Backside Illuminated CMOS sensor has the Metal wiring layer positioned below the Photodiode layer which means light is not reflected and lost. Due to this design the Photodiodes receive more light and the sensor is able to produce higher quality images in dark or low light scenes. So in my opinion, if you see the spec of your camera backside illuminated CMOS then it’s a good thing than have the conventional CMOS.
WHY DOES SENSORY IMAGE SIZE MATTER?
The first thing I look for when purchasing a camera is something most aren’t even aware of. It’s not the brand name or the quality of the lens, the touch screen technology or the LCD screen size, and not the array of functions it offers or shooting presets available – it’s the size of the image sensor. As a 20-year pro photographer who’s captured over a million images during my career, I’m the guy who admires the parts of the engine instead of falling in love with the flashy exterior or high-end sound system. The image sensor is where the rubber meets the photosensitive diodes.
In writing my first installment for Primed, I’ll give a few definitions to clear things up a bit when it comes to a camera’s image sensors and size, explain in detail the parts of a sensor, how it alters the photos (or video) you capture, where it came from, and why it’s important to consider its size – I’ll cover the meat and bones, get to the heart of the matter, the nub, the crux, the nuts and bolts, get down to the brass tacks, all while exhausting our thesaurus. Let’s dive in, shall we?
In today’s digital SLR camera, the image sensor is what film was to a 35mm SLR camera. It isn’t a setting you can control in your camera’s menu, yet a specification you may wish to purchase. Known as a solid-state device, it’s a silicon chip of a certain size containing millions of photosensitive diodes called photosites (or sensels) that record light or photons, transforming them into an electric signal displaying color, tone, highlight, and shadow — to convey the moment you just captured. A digital image file is created from this process, which stores the recorded light data as a set of numbers corresponding to the color and brightness of each pixel. (It’s the smallest addressable screen element — also known as a pel standing for “picture element” on a display — just keep clicking the magnifying tool in Photoshop and you’ll meet one face to face.) In a magical way only Doug Henning would be proud of, all of these pixels come together to create a single image, a photograph.
Sensors come in two forms – either a charged-coupled device (CCD) or a complementary metal–oxide–semiconductor (CMOS) – and are predominantly used in digital cameras: everything from your smartphone to a point-and-shoot, a Four Thirds to a DSLR, to a medium format system. They range from 3 to 3,200 megapixels, although the consumer market at the moment lies below the 80MP range.
Image sensors came into the world through the public and private sector – one avenue through the government’s use of digital technology to further a desire in their peeping Tom business of spy satellites. Their work advanced the science of digital imaging and in the 1960s engineer Eugene Lally, working for NASA’s Jet Propulsion Laboratory (JPL), described the use of mosaic photosensors to digitize light signals, which in turn produced still images. NASA followed this discovery over subsequent years developing small, light, powerful image sensors for use in the harsh conditions of space. JPL engineer Frederic Billingsley first used the word pixel in 1965. I wonder if he was related to the Beaver’s mom, Barbara Billingsley?
On the private end, in 1969, Bell Labs’ need for developing a solid-state camera for use in video telephone service also played a part in the invention of the image sensor. George Smith and Willard Boyle, attempting to create a new kind of semiconductor memory for computers, designed a new form of imaging technology, the very first CCD. This got the ball rolling for the development of digital photography, and the Nobel Prize in Physics for Boyle and Smith in 2009.
There’s some dispute as to who created the first digital camera – some say Texas Instruments is to thank. It patented a film-less electronic camera in 1972 – today, 96 percent of its business deals with semiconductors (acquiring National Semiconductor
earlier this year for a cool $6.5 billion). Other reports say Eastman Kodak made the first one in 1975 (developed with NASA’s JPL), then again some believe it was Sony’s 1981 analog electric camera Mavica which was the first consumer SLR camera to use CCD, 100 x 100 pixel, eliminating the need for film – although I’m not sure I’d want my precious family reunion photo created by one of those things.
Nikon put its chips on the table in 1986 with its prototype called SVC, and then in 1990, Kodak introduced the first commercially available fully digital SLR, the DCS-100, a 1.3 megapixel digital camera back that attached to a Nikon F3 SLR film body – this modified drive unit not only had an external storage unit connected via cable and could produce a 5×7-inch digital photo-quality print, but also ran a hefty $30,000 US. Ouch. Needless to say, we’ve come a long way in 21 years. During this period, image sensor quality, efficiently, size and availability increased while prices dropped – the first sign in the consumer market that film might not be dead, but it was dying. Sure, it still remains a favorite medium of hard-core artist-photographers and up-and-comers who never had the chance to try techniques like cross-processing, but the market is small and getting smaller.
Throughout the 1990s, JPL continued to further the advancement of CMOS image sensors. Its goal was to maintain scientific image quality while creating a camera for interplanetary spacecrafts. CMOS sensors appealed to NASA than the more widely used CCD because CMOS sensors were easier and cheaper to manufacture, and offered a slightly different method of recording the signals. This, in turn consumed less power — up to 100 times less — and when it comes to space, battery power is critical. Later, JPL invented the CMOS active-pixel sensor (CMOS-APS), widely used in mid-range DSLR models today, which, through amplification, improved the image quality still with less power demand. CCDs were the original technology for image sensors and are still used today in other applications, but the digital imaging direction seems to have shifted toward CMOS – at least in the consumer market – most cellphone cameras use CMOS sensors as do the most compact and DSLR models.
By 1999, Nikon’s D1 was a fully integrated DSLR offering the use of its manual-focus and auto-focus lenses. Other manufacturers entered the digital market soon after: Fujifilm in 2000, Canon in 2001 with its 4.1 megapixel EOS-1D, not to mention Minolta, Pentax, Olympus, Panasonic, Samsung, Sigma and Sony. Canon continued the push the boundaries, introducing its 6.3 megapixel EOS 300D SLR camera in 2003, with an MSRP under a grand. Since the CCD was originally invented for video, it wasn’t long until Nikon caught on and released the D90 in 2009, the first DSLR to feature video recording.
My first experience with digital was in 1995, in the middle of the Utah desert shooting with Nikon’s E2s digital SLR. While digitally documenting the Eco-Challenge adventure race I transmitted my images over phone lines every evening, watching my filmless photos appear across four columns in the Salt Lake City Tribune the following day – it was a weird and wild experience. Over the past seven to eight years not only has digital technology taken over the market in consumer purchases, but improvements and options added in the majority of DSLR cameras also helped digital match film in quality and detail. I finally eBay-ed my last film camera earlier this year, my beloved medium format system, mainly because I hadn’t used it for a while and didn’t feel like adding the bulky, expensive digital back to an already cumbersome heavy camera. Admittedly, I’ve also been super pleased with the detail obtained from my DSLR system.
Even though you may decide buy a certain brand of camera, you may have a Sony image sensor inside regardless.
Larger, more efficient image sensors continue to be manufactured. Megapixel counts are on the rise, as is high ISO performance, and the minimization of digital ‘noise’ produced by the sensor – which is a major downer, so fixing this would be cool. These new sensors can capture a massive amount of detail, with little noise producing vibrant photos with rich hues and crystal clear detail. Sony is a major player in the digital market, most unaware of its role as a huge manufacturer of camera sensors, supplying many of its competitors. Even though you may decide to buy a certain non-Sonybrand of camera, there’s very possibly a Sony image sensor inside regardless.
Sony NEX-C3 sensor compared to Nikon J1 sensor
Today, many high-end pro-level DSLRs come with full frame image sensors – equal to the size of 35mm film (36 x 24 mm), thus the name “full frame.” Of course, full frame doesn’t come cheap — nowhere near it, actually. High quality equates to high price. However, in 2009 with the release of the Alpha 850 (the first full frame DSLR under $2000), Sony broke that mold by offering an affordable alternative at an amateur-level price. I’m not entirely sure why manufacturers like Nikon have chosen to offer only one DSLR model above 20 megapixels at a current astronomical cost of $8000, while Canon offers two, its lowest at a much more reasonable $2500. Sony’s Alpha 900 rates highest in this 35mm full frame image sensor class with a 24.6-megapixel resolution. Having said that, I hear this may be changing in the upcoming months as whispers of a robust image sensor in Nikon’s new D800 moves through the tech rumor mill… or is it the D4? No definitive news as of yet.
Translucent mirrors in DSLRs have also entered the market, replacing reflex mirrors that flip up and down to expose the sensor – these new versions allow photographers to shoot faster with minimal shake and hesitation between frames, but don’t necessarily offer an advantage when it comes to image sensor quality. Another form of the digital advancement we’re seeing more of this year is mirrorless interchangeable-lens cameras (MILC). Positioned between compact cameras and DSLRs, these cameras have created a new format called Micro Four Thirds (MFT), often without viewfinders, and vary in image sensor size, five to nine times larger than the Four Thirds system, yet smaller than DSLR full frame sensors – the advantage being a larger sensor in a smaller camera.
Now you know what an image sensor is, how it came to be, and why you may need to work a few extra hours to afford this expensive hobby, but digital imaging still remains a mathematical game — not surprising since a photograph is comprised of numbers determining pixel color, placement, intensity, as well as millions that make up a digital image file. When it comes to getting frisky, Dr. Drew might say it’s the motion of the ocean, but when you consider a camera’s image sensor, size matters, and knowing the differences, advantages, and disadvantage becomes critical.
I recently watched an interview NBC’s Brian Williams did with Annie Leibovitz who, when asked what kind of camera one should buy, remarked the iPhone – “that is the snapshot camera of today… it’s the wallet with the family pictures in it.” Although I truly dig the iPhone 4S‘ new 8 megapixel camera and all the revolutionary technology crammed into the smartphone, the Sony-made image sensor is just not large enough to rival images captured with a DSLR – and that’s expected. Apple describes its A5 chip, designed with an image signal processor, as “just as good as the ones found in DSLR cameras” and this might be true, but the image sensor is not – big difference between a signal processor and a sensor. It might allow you to shoot faster, or capture nice color and tonal range, or to use when you don’t have a camera handy, but it can’t match the quality of a larger image sensor that’s comes with a higher-quality lens. Simply put, you can’t squeeze a V8 engine into a moped. Then again, I can’t make a call, text, tweet, Google Map a route, or play Fruit Ninja with my DSLR camera, either. I will say the iPhone makes a great compact portfolio.
To know what to look for, you must first consider an image sensor size comparison between camera types, as seen in Table 1-1 below.
Image Sensor
|
Size
|
Camera types
|
| Medium format |
50.7 x 39 mm |
High-end Pro digital medium format |
| Full frame |
36 x 24 mm |
High-end Pro DSLRs |
| APS-C |
24 x 16 mm |
Prosumer-based DSLRs |
| 4/3″ |
17.3 x 13 mm |
Four Thirds System |
| 1/1.8″ |
7.2 x 5.3 mm |
High-end compact cameras |
| 1/2.5″ |
5.3 x 4.0 mm |
Consumer-based compact cameras and high-end cellphone cameras |
Compact cameras and cellphones start off tiny – anywhere from 5.3 x 4.0mm up to 20.7 x 13.8mm; this goes back to the V8 vs. moped reference – the trade-offs for lightweight and small size equates to lower quality image files.
Amateur and mid-level DSLRs usually house a decently sized APS sensor ranging from 22.2 x 14.8mm to 28.7 x 19mm – larger than a compact and Micro Four Thirds camera, yet smaller than a high-end DSLR. Combine this APS sensor size with a DSLR lens and you add a crop factor of 1.3x to 1.6x (depending on the size of the sensor) changing the length of your lens. If your lens is 100mm, with an APS-sensor camera, it’s now a 130mm with a 1.3x crop factor, or 160mm with a 1.6x crop factor. This isn’t an advantage or disadvantage, just a mathematical fact. Some manufacturers produce sensor-specific modes (such as Nikon’s D3 and D700 DSLRs) to counteract this issue, accommodating different lens formats.
As mentioned earlier, pro-level DSLRs come with full frame image sensors around 36 x 24mm, matching the 3×2 format of 35mm film, offering a larger sensor with no crop factor. Medium format digital cameras go a step further storing the largest image sensors in the consumer market, up to four times larger than full frame, from 50.7 x 39 mm to 53.7 x 40.3mm, producing the highest-quality image files; however, you lose mobility with the size and weight of these bulkier systems, and their costs can match a new car loan. Unless, of course, you’re a pro photographer charging Leibovitz-level rates or were left a nice trust fund from Grandpa Rockefeller. For the majority of consumers, this slice of photographic heaven is out of reach.
It should be noted that the smaller your image sensor is, the more depth of field you acquire for the equivalent field of view and aperture. Take a 100mm lens on a Four Thirds camera, and it’s easier to get everything in focus more so than with a pro-level DSLR with that same focal length. The misconception to some is to see this as an advantage, but I don’t. Having more control over depth of field is the advantage in photography, so larger sensors with more photographic knowledge is the way to go.
With this newfound knowledge of ‘larger equals more money’ when it comes to image sensor versus price, why pay the additional cost? Answer: better image quality in the form of detail, low light performance, reduced noise, and a greater dynamic range. In case you were wondering, here comes the meat and bones.
Detail
Quality and detail go hand in hand. We see this today when we compare a VHS tape to a DVD, or even a DVD to a Blu-ray Disc. The theory follows the same film-based fact the master B&W photographers utilized throughout the 20th century: the larger the film size, the more information they secured. This transfers into sharpness of intricate details within a photograph. It may not seem to make a difference when viewing photos as small JPEGs on the web or your smartphone screen, but shift to a larger high-res display or print enlargement and you begin to see what I mean. Step up to the next level with a 5 x 7 print and again the quality comparison becomes more evident. Once you decide to zoom into the original image file in Photoshop, or blow it up to a 20 x 30 print, major differences rear their ugly heads in the form of fuzzy detail, less dynamic range and digital noise.
Low light performance / High ISO noise
The definition of digital noise in a photograph is randomly spaced, brightly colored pixels – in a sense; it’s interference showing up in your images, often seen in darker sections like shadows or large single-toned areas like a clear sky. The more you raise the ISO in your camera, increasing the image sensor’s sensitivity to light, the more noise shows up – very similar to grain with high ISO films. Just crank up your ISO over 800, or use your auto ISO (one feature I’m not fond of) and every time the camera decides to blast your ISO into noise hell, you begin to see the after-effects.
Besides the amount of pixels increasing, the larger your sensor, the less noise you acquire at higher ISO settings.
Digital noise can replace small features and proper color – lack of richness in shadow areas and less-than-smooth detail –the reason why I recommend using a higher ISO only as a last resort.
However, image sensor size also plays an important role when it comes to noise. Besides the amount of pixels increasing, the larger your sensor, the less noise you acquire at higher ISO settings. This is due to the larger pixel’s ability to receive more light, creating a greater signal-to-noise (S/N) ratio (as with any electrical system, whether sound or image based, the better the signal, the less audible or visual noise you receive). Photosites on the sensor can also be farther apart creating less contamination from electrical signals if they were closer. The result? Smoother photographic renditions on higher ISOs, expanding the range of light a photographer can capture, especially with moving subjects in low light. I think I’m salivating.
How to find out which camera has the best S/N ratio depending on the ISO setting? An independent test lab known as DxOmark rates sensors depending on a camera’s S/N ratio and charts performance at each ISO setting, pointing out unacceptable level of noise, measured in decibels.
Greater dynamic range
The larger and higher quality your image sensor, the greater dynamic range it can cover in a single image file — yet another big advantage in photography. In their tests, DxOmark looks for a dynamic range of at least 9EVs (or 9 stops of light) in each sensor – a measurement of the range from highlight to shadow – the more range covered, the better the image quality. As an expert in exposure, having written a book on the topic, another misconception of most photographers is the ability to capture all detail in all areas of any given scene. This is impossible for most scenes since many fall out of the dynamic range of the image sensor; it was no different with film, and actually the dynamic range was smaller – around 6 stops of light.
DxOmark considers 9EVs or less too limiting for image sensors, showing less than smooth gradations between shades and colors below this level. Leaf — a digital back manufacturer that entered the market in 1992 and partnered with Phase One in 2009 — claims its latest full frame 53.7 x 40.3mm sensor can cover an “unsurpassed 12 f–stop dynamic range.” DxOmark agrees, listing Phase One’s top medium format digital back as the highest image sensor quality in the consumer market with a 91 out of 100 rating; but then again, a $42,000 price tag should get you no less. Nikon’s more reasonable (yet still costly at $5100) D3S ranked highest in the DSLR category, and Pentax’s K5 was tops for APS-C sensor followed closely by Sony’s NEX-7 mirrorless camera.
With changes in digital technology moving at a rapid pace, it’s hard to say which direction image sensors will go in the near future – we may even see a new type of technology take over the market. I would guess organizations like NASA may play a big part in determining what goes into our cameras of tomorrow. Having an interest in securing as much detail about distant stars and galaxies, much of its visual work relies on the quality and size of the image sensor inside space-based telescopes.
DALSA Semiconductor successfully manufactured a 111 megapixel image sensor as early as 2006. Then, in 2009, it was announced the Large Synoptic Survey Telescope (the world’s largest sky-survey telescope partially sponsored by Mr. Microsoft himself, Bill Gates) had a massive 3,200 megapixel camera to cover the universe, delivering near real-time images to the public — that’s right, 3,200 megapixels — space is large, dudes. Fermilab followed this up in 2010 with its 4-ton Dark Energy Camera cranking upwards of 570 megapixels. These sensors may be too large to fit into any current consumer-based digital system, but I’m sure the technology to shrink ’em down is in the works.
The size of a camera and its image sensor should get smaller. That’s to say, manufacturers most likely will be able to fit more onto a sensor with finer detail and better quality, but I imagine the same theory that has been a part of photography the past 180 years will continue to apply. Math rarely changes regardless of the medium. The larger the sensor you have in your camera, the more detail, less noise, and more superior image quality you will obtain – and although you’ll have so much more than you did just a few years earlier, you’ll still want more.
My Conclusion
- Both CCD and CMOS has no clear technology. Choose properly and carefully about what you need. As notes, people reviews says that CMOS is having much better quality in video.
- The Backside Illuminated CMOS sensor is able to produce higher quality images in dark or low light scenes rather than the conventional CMOS sensor.
- Your camera’s sensor size does matter!
This is the example where you check camera size on reviews (dpreview.com)
MP, Sensor Size, and Sensor type of respectively Nikon CoolpixP300, Nikon Coolpix P310, and Nikon Coolpix AW100
Source
http://en.wikipedia.org/wiki/Image_sensor
http://nikonasia-en.custhelp.com/app/answers/detail/a_id/6777/~/what-is-a-backside-illumination-cmos-sensor%3F
http://www.engadget.com/2011/12/16/engadget-primed-why-your-cameras-sensor-size-matters/
Notes . Noted 