Photography has advanced considerably from its conception into the digital age.
Along with such advancements come additional complications - different sensor types, a variety of sensor sizes, formats and files! So what does it all mean, and why does it even matter?
A bit of background
Early optical or photographic devices (such as the camera obscura) were little more than a simple opening - sometimes with an equally simple lens - that would allow light to pass through, the goal being to generate an image circle broad enough to cover the area required.
Notice 'image circle' - given the physical properties of light, when it passes through an aperture (i.e. opening) it will reform the image as a circular image element; we simply crop this down to a rectangular image for convenience, partly because the image circle will not be sharp or bright to the very edges, nor are circular formats consistantly useful:
|
Image Circles
Within this formed image circle a capture can be made onto a light sensitive material, whether a sheet of coated glass, a sheet of film, or - most likely today - a digital sensor. Charge-coupled device (CCD) sensor size types (i.e. 1/3, 2/3) are nowhere near that supposed size - neither diagonally nor along the longest edge. The description comes from old TV camera tubes, where the diameter of the tube (equivalent of an image circle) is used as a reference.
The useable image area within such an image circle tends to be around two thirds of the full diameter of that image circle, hence the sensors being considerably smaller than their counterpart names and known by 'type' as opposed to their actual size(s):
|
|
Optics
Different optics will create a different sized image circle. Tiny compacts have equally small digital sensors, and therefore are designed with the sort of simpler optics required to generate coverage for such small sensor sizes.
Lenses are designed to refract light, and without going into the complex physics of how they work, different lenses allow for different focal lengths (either fixed or zoom), an array of apertures, and other optical devices.
However, different lenses are designed to be capable of different coverage - i.e. the image circle they will produce will only be large enough to cover a designated sensor size - or, indeed, of smaller sizes.
Crop factor
If you ever see the term "crop factor" or "35mm equivalent" then this effectively references the amount of image cropped (that never even lands on the sensor) when using a 35mm lens.
This, essentially, comes from the traditions of film formats, where a different physical size of film would require more significant lenses (and physical distance/size) in order to successfully cover the required area.
For example a 'standard' lens in 35mm would be 50mm, as compared to 80mm in medium format (6x7cms) and 150mm in large format (5x4").
Many DSLR cameras use a smaller sensor than the traditional 35mm film size. A common size is APS-C - also known as DX - which has roughly 2/3rds the surface area of a 35mm frame (typically slightly smaller than 15x24mm, against 24x36mm frame in a 35mm camera - or a camera with a 35mm sized 'full frame' sensor).
As a consequence, you need to apply a 1.5x multiplier to compare the two - to judge how a given lens on DX would look in 'full frame' equivalent, multiply its focal length by 1.5 (so a 35mm lens on a DX camera approximates the performance of a 50mm lens on a 35mm camera); to work out what DX sized lens is require to simulate a given lens on 35mm, divide its focal length by 1.5 (so if you want the look of a 105mm lens on 35mm, you need to look for a 70mm lens on DX).
Other popular sensor size calculations include Canon's 1.3x multiplier (used on the 1D mkIII and previous models), it's 1.6x multiplier (used on the 350D, 400D and 40D) and the 2x multiplier Four Thirds sensor used by Leica, Olympus and Panasonic.
The properties of the lens itself don't change; depth of field for example will remain the same irrespective of whether that 150mm lens is on a 35mm camera on a DSLR with a 'cropped sensor'.
Sensor sizes
The current market has cameras with a wide array of sensor sizes and types:
Imagine shooting a subject with a compact with a 1/3 CCD sized sensor and a full frame (36x24mm) DSLR. Let's say both shoot at six-megapixel resolution - so the same.
You wish to create an 8x10 print from both - the smaller sensor requires considerably more enlargement than the larger sensor to achieve the same size of print. So whilst they have the same number of megapixels, those pixels are actually physically smaller and crammed together - this causes a number of problems:
When pixels are close together their spatial difference is higher - this is known as spatial frequency and represents the number of pixel lines found per, in this case, mm.
A smaller sensor leads to a higher spatial frequency
High spatial frequency limits Modulation Transfer Function - this is a measurement of ability to transfer contrast from a source and display on a new plane at a specific resolution (as per the way optical devices/cameras operate)
Therefore higher pixel count on a smaller area leads to worse resolution
This may sound complicated - but challenges the idea that 'bigger is better' and instead suggests that there's no point in cramming too many pixels onto smaller sensor sizes!
Cameras are often mis-represented by their pixel counts, that a 12 megapixel camera is better than an eight megapixel camera is not necessarily true.
Not forgetting that the more pixels that are crammed into the fixed size of a particular sensor, the less able light can be cleanly or accurately gathered by the pixels - which brings the likelihood of increased noise in the capture.
What about CMOS or CCD?
There are two main sensor types for digital photography - CCD (charge-coupled device) or CMOS (complementary metal oxide semiconductor) sensor.
Both types of imagers convert light into electric charge and process it into electronic signals, which is essentially what becomes your capture.
In a CCD sensor, every pixel charge is transferred through a very limited number of output nodes to be converted to a voltage, and then sent off as an analog signal.
All of the pixel is devoted to light capture.
In a CMOS sensor, each pixel has its own charge to voltage conversion, so the sensor also includes amplifiers and noise-correction - the chip outputs digital bits.
These other functions increase the design complexity and reduce the area available for light capture.
Many may question which is better - honestly, it's not particularly significant. Don't let the acronymic techno-waffle confuse you. In short sensors are designed to do a job and do it well.
So, what is right for you?
It all depends on what you want to output. Knowing in advance your goals should make all the difference for your choice of camera.
If you need a small compact to take snaps in auto mode of your mates on a night out to pop into Facebook - then small and cheap will cover everything that you need.
If you're looking for pro-grade sharp large format prints then look towards big sensors in DSLR (maybe even digital medium format) with high quality lenses - bigger is better, but it is, of course, physically larger and will take a much bigger chunk out of your wallet. Plus, it may not always be suitable for the situation.
If it's DSLR sensor sizes that are confusing you, then could be an idea to keep your eyes on Four Thirds systems - this is the only universal, interchangeable DSLR type, meaning that a variety of lenses and bodies from multiple manufacturers can be used in any combination. The sensors aren't too large either, so cameras such as the Olympus E-410 DSLR for example are the smallest and lightest DSLR systems on the market today, and won't murder your bank balance.
Finally, let's not to forget that some places won't let anything resembling pro kit into certain places - so next time you're contemplating taking a DSLR to a gig, concert, or some major tourist attractions; it may be worth checking in advance whether such kit is permitted.
Share this article:
Del.icio.us this
