Faster shutter speeds, and faster flash synchronization. Will this make you happier? The advantages of an all-electronic shutter are considerable: no shutter noise, and no shutter shake. They sport an extra transistor at each pixel to remember the accumulated electron count, so exposure for the entire sensor can begin and end at the same time-then read out afterwards. CMOS sensors with so-called global shutters are coming down the high-tech highway. Every pixel is exposed simultaneously, or at least as simultaneously as that mechanical shutter will allow.īut here’s good news. So these use a mechanical shutter and clear all the rows at once. Rolling shutter distortions are acceptable for video and cell phone selfies. For moving objects, this delay between top and bottom can produce weird effects. The image is built up by a “rolling shutter” in which the last row is cleared and begins its exposure 1/50 second after the first row. So in practice each row is cleared (initializing exposure) eight microseconds after the row before it. No big deal, but the next row has 16 microseconds extra, the following row 24 microseconds…and the last row is overexposed by a non-negligible 1/50 second. That would mean the second row has eight microseconds additional exposure before you get around to reading it out. No, 1/50 second is the total readout time for all the rows, and it throws a monkey wrench into the proceedings.Ĭonsider: You expose the first row of the sensor, then you spend eight microseconds reading it out. Exposure time can be just about anything you want. That’s not the exposure time, which is determined by how long you allow each row to collect light before reading it out. If there are 2,500 rows (typical for cell phone cameras), then reading out all of them requires 1/50 second. To read out a row might take eight microseconds. Electronics are fast, but not infinitely fast. An electric signal empties a row, and its pixels (“buckets”) are then left alone to collect light during exposure-after which they’re read out one by one. In a CMOS sensor, this is done row by row. To measure this metaphoric rain, you empty all the buckets on your sensor, expose them to the sky, and then see how much water each has collected. This situation is somewhat analogous to having a few million buckets arranged in rows and columns in your backyard, collecting raindrops. These are tallied up during exposure, and converted to a voltage that is simply proportional to the amount of light illuminating the pixel. The more light that hits a pixel, the more electrons. This interaction is known as the photoelectric effect, and in 1921 Albert Einstein bagged a Nobel Prize for explaining it. Incoming light hits a pixel and frees an electron from a silicon atom. Consider the workings of a typical CMOS sensor, the type that is currently in vogue. So why not just electronically turn it on, and then a fraction of a second later turn it off? But today’s digital cameras have a sensor that can be used over and over. So here’s the deal: In film days, shutters were an obvious necessity. That may sound adequate, but not if you make thousands of pix a week as some pros do. How does it manage that, and why doesn’t your expensive DSLR have such a shutter? After all, mechanical shutters wear out-usually after a few hundred thousand shots. Unlike highly evolved mechanical shutters-the spring-loaded clockwork you’ve been using since you were knee-high to a bonsai tree-your cell phone camera is as silent as the dead. No, of the more than a trillion photos taken this year, roughly nine out of 10 are made with a phone. You know: something with a tripod socket. Most photos today aren’t shot with a camera-at least if you define “camera” as hardware solely used to record images. It won’t surprise you, although it could make you uneasy. Fan photographed with a rolling shutter and a global shutter.
0 Comments
Leave a Reply. |