Sensors Part I: The sensor size is very important? The subject of the resolution, sensor size and quality of a photo. Pressing the shutter The technical quality of a photo depends first and foremost of the lens, without a doubt the most expensive element of the camera; second sensor, the electronic component that converts the image captured by the lens into digital information on light and color, third camera processor (each brand has its own), which manipulates the sensor data according programming each brand and model are the factory, and finally the resolution (megapixels) and dynamic range . blog In this article, we discuss the sensor, so make yourself comfortable because there are a lot of bread by slicing.
SENSOR: a light-sensitive element, is in the place that was previously the film, ie in the focal plane. The sensor is responsible for capturing the scene projected by the lens and turn it into a series of digital signals (ones and zeros), which will be converted into a photographic image by the camera processor. The sensor is composed of many sub elements (photosites), for simplicity we will call them pixels, as in a sensor usually be millions of these sub items, the unit used is the megapixel. Each pixel is arranged in a particular RGB color pattern, such as BAYER (the most known and used, but not only), and each pixel has color information and brightness at that point. A CMOS sensor
There are two types of sensors, the CCD (Charge - Coupled Device) and CMOS (Complementary Metal-Oxide - Semiconductor)
each has its advantages and disadvantages , but we will not discuss these issues in this post.
The megapixels are a unit of quantity, not quality, so a camera with a 14-megapixel sensor is not necessarily better quality than one with a 10-megapixel sensor, although those who are responsible for the marketing of the cameras we want convince you otherwise.
It is very common to listen to vendors and customers to discuss the issue of points megapixel camera sales, but is more common to see how they make buying decisions based solely on this parameter, is some people argue that their cell 5 or 8 megapixel camera takes pictures of equal quality to a camera with the same amount of megapixels. Indeed
is desirable to obtain good resolution for better detail in the photos, high resolution is important when printing in large sizes and when you want to cut some photos to enlarge the subject. Obviously the resolution (megapixels) is important, but not the only factor we must look
Others have heard or read that the sensor size is critical to the quality of the photos, and only make decisions based on this parameter, and if there is a camera with a sensor 20 x 20 cm. would rush to buy it.
But everything has a practical limit, both the number of megapixels as the size of the sensor.
The real secret is not the quality or the amount of megapixels or the size of the sensor, found on the size of the photosites, of these sub elements that capture the pixels one by one, they determine the quality of information for each individual pixel, and thus the overall quality, ie photo.
There is a factor that depends directly on the size of the light sensor element, and is
The Signal to Noise Ratio (Signal to Noise Ratio): Signal: is all relevant information about what you want capture, either a picture, sound or any other information.
Noise: is any unwanted information that inevitably captures along with the signal, reducing its capacity to interfere with it.
This signal and noise are also some relative, if you want to hear the conversation between two people and there music something back with the volume up, the conversation is the sign and the music is noise, but if I listen to music and two people speak very highly and / or very near, the music will be the talk signal and noise ... I mean? (The noise is interfering).
For example, if you're in town and want to listen to your radio, tune your favorite station and you're done. But the receiver also produces some unwanted noise, due to its electronics, noise that is totally inevitable and inherent to their status transistorized computer, with chips etc. But one does not hear the noise because the signal (music) is very strong and need not rise much the volume to hear properly, in other words the signal is so high that in comparison with it the noise is very low ...
has a high signal to noise ratio .
Another time, if one is somewhere away from the city and want to tune your favorite station, but the signal is very weak because of the distance, then you have to turn the volume up and realizes that his music is inevitably accompanied by a nasty hiss (noise) that noise is always present, what is happening now is that the signal is lower than the rise volume (amplify), noise can match and even surpass, that is we now have more noise than signal making it difficult or impossible to hear the music, the radio now
has a low signal to noise ratio. If I failed to explain properly on the concept SIGNAL / NOISE, please write me to mail the blog and try to do better.
Just as in the example above, each sensor element produces its own noise, and the higher the signal to noise ratio, the more will the resulting image quality, the secret is to keep the noise as low possible to improve the relationship. The only way to do this is: using larger elements (sensor design itself), and maintaining the ISO VALUE
as low as possible. Fortunately
self-noise level of an electronic element grows very little to increase its size, but not to increase their signal amplification, the ISO value in the case of the sensor.
To understand how it influences the size of the sensor elements in the image quality, let me illustrate the process with some examples:
If you put in the rain some deposits of different sizes, which one would receive or would capture a greater number of drops and thus a greater amount of water? (Assuming all time remain the same in the rain).
What deposit will receive more rain drops?
Obviously the larger container will collect more raindrops, the same way a larger photosite collect as much
photons (light particles). Thus, the fact that a sensor is composed of larger items benefits the quality of a photo.
But that is not the merits, we saw the issue of signal to noise ratio, since each component of the sensor
always produces a certain noise level, see what are the advantages of larger photosites:
Signal to Noise at ISO 100
As noted above, the size of a photosite affects little noise it produces (all electronic devices produce), but their size helps us to improve the signal- noise. In the small photosite photons have 9 and 5 units of noise, our signal to noise ratio will be 1.8; in the medium photosite have one more unit of noise, but we have more than twice as many photons as the previous, so our signal to noise ratio will be 3.67, as we see is higher than in the photosite small, since although noise is a little higher, the number of photons that can be captured is much higher, and now we see the larger photosite, which has a signal to noise ratio of 7, far higher than the other two.
Signal to Noise at ISO 800
In this graph we see what happens when you raise the ISO value, the units are automatically increased noise, now changing the signal to noise ratio in the three photosites, but with much greater IMPACT affecting the smallest element, whose ratio is now 0.56 and there is more noise than signal in the intermediate size we now have 1.8 of SNR, and is that the signal becomes twice the noise, yet is less affected than the smaller element, and finally the largest is now 4.1, SNR, showing his ability to work without a lot of high ISO values impact on the quality of your information.
could say then that the signal to noise ratio is the relationship between useful information and useless. Just as the "noise" appears on a radio with an annoying sound that interferes with the music, in a photo appears as a series of dots or small spots that interfere with the sharpness of the picture.
NOTE: The values \u200b\u200bused are only illustrative, plotted the principle of signal to noise ratio with different ISO settings, but they do represent reality. Then some pictures taken with a pocket camera (point & shoot) with a small 14-megapixel sensor, with very good light and using ISO values \u200b\u200bof 80, 400 and 1,600, how we affect the ISO to the quality of a picture:
So, at first glance did not notice the noise (digital), high amplification product of signal with high ISO values, but look closer and you will realize, how as the ISO value increases, the signal to noise ratio gets worse, becoming louder the noise and reducing the sharpness and detail capacity the photo
Whereas a camera point & shoot, is a photo
with a level of detail acceptable.
By raising the ISO to 400 there is already a noticeable loss of detail.
By raising the ISO to 1,600 loss of detail in this photo
is unacceptable for anyone.
Click on each photo to enlarge
Well, now understood as sensors with larger photosites achieve better images?; Be seen how the larger photosites take more quality values High ISO?, Note that I said the larger sensors, but those with larger items, not necessarily the same thing ... see:
The following graph we have three sensors imaginary three different sizes, all three have the same resolution
(192 pixels), but differ in size from the smaller sensor has an area that is the sixteenth part of the sensor and the fourth largest of the sensor medium.
If we keep constant the resolution, then the size of the photosites is proportional to the areas of sensors, so the smaller sensor will have some photosites that will be the fourth of the sensor photosites medium, and sixteen vece smaller than the larger sensor.
Click to enlarge
Which of these three sensors have the elements (photosites) bigger? ... obviously the 8 x 6 cm., then a resolution equal
the larger sensor will have a better image quality, better performance at high ISO settings, better signal to noise ratio and better dynamic range.
In this case a larger sensor achieves better image quality (because the photosites are larger). Now let's see what happens when you increase the resolution of the sensor 16 times larger, we now have much higher resolution (more megapixels in real life), this is what pushed us to buy in stores and in advertisements ... but I will have more quality with higher resolution?
Click to enlarge
By increasing the resolution without changing the area, we have no alternative but to use smaller elements, which puts the larger sensor in the same situation that the smaller sensor, both have the same size photosites ... realize? ... clear that there are additional considerations, usually the larger sensors found in cameras reflex (DSLR), and these are best small camera lens fixed lens, whereupon the better part of a sensor basic point & Cameras shoot and / or bridge (prosumer) DSLR cameras additionally have more powerful processors, and better resources for digital filters and stuff, that's why we have achieved in a similar situation to get more out of its sensors, even when they do not have huge photosites.
I hope I have clarified some doubts about the size of the sensors, the resolution (the famous megapixels), its impact on picture quality and ISO.
In the next post we will pursue the subject of the resolution and some real life figures obtained from the data of some camera manufacturers.
Until the next post!
UPDATE !!... Sensors Part II is released!
