What is photography? There’s another basic photography class here at Berkeley, a freshman-sophomoreseminar listed as CS39J. The professor of that class, Brian Barsky of the CS department, subtitles his class“Drawing with light”, which is where the term ‘photography’ comes from: photo, photons, light, and graphy, drawing… drawing with light.The ultimate goal of this course is to teach you how to produce good photographs. For the most part, however, we won’t be dealing with the photographs themselves, or what a good photograph is. That’s all mostly subjective stuff, and we’ll leave it to visual arts and your own taste to decide that. Instead, thisclass will dive right into the process of making photographs – skills used by photographers to take photos, a comprehensive look at the way camera technology works to produce the images, and post-processing techniques to enhance and fix images after they come out of the camera.Let’s first consider the definition of a camera. [Ask class for definition of camera]. A camera is a photographic device, which means it’s a device that “draws with light”. The input we give it is light. The output we receive is an image.Light (photons) ->Image (film negative, image file, print, etc)For today, we’ll take a simple view of how this photons-to-image transformation process works.[Diagram of camera, with photon trajectories (Slide 2-3)]All cameras work in the same basic way – you have a system of lenses, or in some cases a pinhole, which concentrates light onto some sort of photosensitive plane at the back of the camera, which records the image. In the days of film, there was a photosensitive piece of film located on this plane. These days with digital, there is an electronic image sensor which reads all of the light information.The Imaging SensorWe’ll talk about lenses later on in the course, but for today we’ll focus on the center of every digital camera: the image sensor. Now if we think back to our definition of a camera – a device which transforms light into an image – it’s the image sensor which plays the biggest role in this.What is an image sensor? In general, it is a device that converts light particles (photons), into electrical signals. You can shoot photons at an image sensor, and it will output electrons, which can collect to give you an accumulated voltage telling you just how much light input it received. [Slide 6] The process in which image sensors do this is known as the “photoelectric effect”.[Slide 7][Draw diagram][Horizontal line, representing silicon substrate][Valence electrons floating around][Photon coming in and hitting the silicon][Electrons knocked loose into collector]The basic operation of the image sensor is for photons to knock out electrons from our photosensitive material, which we can then collect as a measurable charge. For those of you who remember yourchemistry, you may remember that electrons exist in quantized bands of energy, or energy levels. Given enough energy, an electron is able to move up an energy level, and it can also move down an energy level by releasing energy, in the form of light or radiation. For the purposes of our sensor, the most important energy bands are the valence band and the conduction band. Valence electrons in this band are those in the outermost layer of the atom, which are still “attached” in a sense to the atom’s nucleus. The conduction band is a band of energy above this, in which electrons are “unattached” from the nucleus, and are somewhat free to roam around and, by definition, conduct electricity.For conductors, like metals, the valence band and conduction band overlap – most electrons that are valence electrons are also somewhat free to roam around and conduct electricity. For semiconductors and insulators, however, the valence band and the conduction band constitute discreet energy levels, and the energy in between the valence and conduction band is known as the “band gap”. In order for a valence electron to detach itself from the atom and move into the conduction band, it has to receive enough energy equal to or exceeding the band gap, to essentially “jump” the band gap into the conduction band.When we expose our sensor to light, what we are really doing is letting that light, which consists of photons, come in contact with our sensor. Photons have energy of their own, and if one happens to smash into one of our electrons with enough energy, that electron gets knocked out of the valence band,and into the conductance band.We can collect that electron into a big charge collector, known as a photo-well. As more photons hit our sensor and collide with electrons, more and more get knocked loose, and we collect all of these photons into our photo well [Slide 7]. So now we’ve finished our exposure, we close the shutter, stop exposing the sensor, and we’ve got a photo well with some number of electrons, which means we’ve got a certain amount of corresponding charge. Now we can just measure the resulting voltage from this charge, and presto!, we know how many photons we received, and thereby how bright our image was. [Slide 8] If our photo-well didn’t have many electrons and we didn’t measure much voltage as a result, that means it didn’t receive all that many photons, so the sensor records a dark image. If we stuffed our photo-well with lots of electrons, that means we get a really bright image. If our photo well didn’t have any electrons, and we didn’t measure any voltage, we should just get pure black. And if we maxed out the capacity of our photo-well, we should get a pure white image. Whatever the case is, the brightness of the image that we’re going to produce is going to correspond directly to that amount of charge we measured, which corresponds directly to how many photons hit our sensor.After we actually measure our voltage, we feed the measurement into an Analog-to-Digital converter, which converts that raw measurement into the 1’s and 0’s of binary data that our computers can read the image information from [Slide 9].The example I’ve just showed you is a single-pixel image sensor. All the photons hit one spot (even if it isa big spot), and all the electrons that are knocked out collect into one photo-well, which outputs into a single charge measurement, telling us the brightness information for one single pixel. For actual digital cameras, which have