To Avoid Being Buried Alive!
To Avoid Being Buried Alive!
Everyone is always interested in the MP rating of the newest camera that’s introduced.
Photographers turn up their noses at a 12 MP camera as if it can’t possibly be worth anything once a new 16 MP camera has been introduced. But, honestly, this MP race is beginning to drive me nuts! Why does it drive me nuts you ask? Well, a jpeg from my 12 MP camera is between 4 and 6 MB. Choosing the midpoint (5 MB) between the two, means about 200 images takes up about one Gigabyte of storage on one of my 1 Terabyte hard drives. If I shoot a compressed raw file with a 12 MP camera each image is between 9 and 11 MB. Again, choosing the midpoint (10 MB), means about 100 images takes up about one Gigabyte of storage on one of my 1 Terabyte hard drives! If, as I do on every assignment, I shoot a raw and a jpeg of each image, 200 images shot that way takes up 3 GB of storage space. So what? Well, I’ll tell you what! One Terabyte (1024 Gigs) divided by 3 GB for 200 pictures is approximately 341, and 341 multiplied by 200 images is only 68,000 images (approximately). Since I shoot between 1000 and 1500 images on every assignment and taking a midpoint of 1250 images, I fill a Terabyte hard drive every 55 assignments! This is a sobering thought.
Want a more sobering thought? I put all my assignments on two (yes two!) mirror image 1 Terabyte hard drives! So what you ask again? Well, not counting the three 1 Terabyte hard drives that are already filled and stored away (I just counted them), I have a stack of seven (yes, seven!) 1 Terabyte hard drives sitting next to my current big Mac (see photo just below this paragraph to prove my point). Heaven help me if I switch to a 24 MP camera! Eventually, will I end up with a stack of one Petabyte (1024 GB) drives? Or, eventually a stack of one Exabyte (1024 PB) drives? Or, eventually a stack of one Zettabyte (1024 EB) drives? After a Zettabyte it’s a Yottabyte (1024 ZB) drive! Forgive the pun but that’s a yotta images! Obviously, as the MP rating of cameras continues to rise, this can become a major problem. But wait, it gets worse. Choose to save a jpeg as a tiff or a PSD format for an advertising image and that file is about 35 MB instead of 4 to 6 MB. Add in a few layers to make it glow (as an example) and a single high-resolution image can approach a Gigabyte! This problem isn’t only about storage space but consider the processing power you need to crunch big files too. It was a joke in the old days (a relatively short time ago) of Photoshop V.1, that if you wanted to add a gausian blur to a scanned image of a 4X5 chrome when working on a Mac Quadra 840 (the biggest, baddest, and fastest Mac offered at that time) you chose gausian blur and then went out to lunch while your Mac crunched the pixels accordingly. Is there any way out of this maddening MP, storage space, and processor speed rat race? Maybe there is.
Canon (and maybe the other manufacturers too) has started to become aware of this. The last two versions of Canon’s great “G” series cameras, the G11 and the G12 are down to 10 MP from the G10’s 14.7 MP. In fact, the more I shoot digitally, the more convinced I am that there are other metrics of image quality that are at least as, if not more, important than MP count when searching for ultimate image quality.
The choice of which sensor size to invest in is a critical step in determining which DSLR camera will work best for you. Given equal pixel count, a larger sensor will have larger pixels, and this is a definite advantage. To grasp this concept this let’s simplicate the numbers to make things easier to understand. Photons are individual beams of light; they are really teeny tiny and now picture in your mind’s eye that an individual pixel is a little bucket into which a huge number of photons fall. First an apology: I first read the “bucket” analogy somewhere on the internet, and I think it’s perfect, but I can’t for the life of me remember where I read it so my apologies for not attributing the “bucket” analogy to the original person who wrote it. Now for the simplicating part: Let’s say a single pixel in a point and shoot digital camera can accept 10 photons. But, a few of the photons bounce off the edge of the bucket and bounce into the pixels next to it. A pixel into which the errant photons bounce has its information degraded by the bouncing photons. Obviously, if a pixel can accept only 10 photons, and three of the photons are bouncing in from hitting the edge of adjoining pixels the image that pixel can produce is degraded by a large percentage. Now, let’s say a DSLR sensor, with much larger pixels, can accept 100 photons in each pixel. If the same three photons bounce into a pixel from adjoining pixels they have a much smaller effect on the pixel’s image making capabilities. The term for what all the photons falling into the pixel makes is often called the “signal” and the term for the errant photons that bounce into an adjoining pixel is what “noise” is. The end result is; larger pixels create a cleaner signal, and a cleaner signal means less noise, which in turn, offers the ability to capture images at higher ISOs and/or have the ability to be upsized by a process called interpolation without creating artifacts.
There are two metrics by which to measure pixel size. One way to compare a camera’s pixel size to each another camera’s pixel size is via a metric called MP/cm² (megapixels per square centimeter). DPreview, under their “specifications” pull down menu lists each camera’s MP/cm² (megapixels per square centimeter). Looking under any camera manufacturer you’ll see that their smallest, most pocket sized, digital camera might have 50 MP/cm² while their DSLR with the biggest sensor might have 1.4 MP/cm². The lower the number before the MP/cm² metric is, the bigger its pixels are and the better the camera’s image quality will be.
The other metric that can be used to compare the size of one camera’s individual pixels to another camera’s can be found at dxomark and looking under their “camera sensor/compare sensors” menu. This website offers the size of each pixel measured in microns (the symbol for a micron is “µm” and it equals 1 millionth of a meter) and in this case the bigger the µm number the better. For example, when comparing two 12 MP cameras; the Nikon D300 has pixels that are 5.42 µm while the Nikon D3 has 8.4 µm sized pixels. I don’t know whether the measurement is the diagonal of a square pixel, the diameter of a circular pixel, or some other measurement of a hexagonal pixel on a Fuji sensor nor do I think it matters. Regardless, I do know that a larger µm number holds the promise of better image quality, a cleaner signal, and less noise.
Almost always, given any two cameras with an equal MP pixel count, the lower the MP/cm² number is or the higher the µm number is, the better the camera’s image quality will be and that’s a fact!
