Sensor Size Vs. Sensor Resolution

Everyone knows that bigger is better. Kinda… Larger sensors are often associated with lower noise levels and generally higher quality photos while smaller sensors are associated with lower data loads and higher transfer rates but noisier and generally lower quality images. It’s a trade off at this point in technology. What a lot of folks need to remember is that there is a fundamental difference between sensor size and sensor resolution. In particular when it comes to Red cameras and their wonky formats. A lot of people including Red staff will describe lens coverage in regards to a specific resolution such as “4K” or “5K”. That’s great since they pretty much own the names and if someone is asking if a lens covers 4K, they’re usually referring to a Red One or 5K on an Epic. But that’s where things get confusing.

The resolution of the sensor (we’ll use 4K for the remainder of this post) has absolutely nothing to do with the coverage of any given lens. OmniVision gives us a perfect example with a newly announced camera phone sensor that measures a mere 1/2.3-inch but will still be capable of performing at a resolution of 4K (according to their press release). The size of this tiny little camera phone sensor is astonishingly small compared to, say a Canon 5D MkIII with it’s 24x36mm sensor. The 4K sensor used in a Red One is approximately 22.1mm x 12.4mm which requires a 25.4mm image circle to be produced by the lens in order to cover the entire sensor. The image circle required to cover the new camera phone sensor from OmniVision would be about 7.6mm. So we have an image circle requirement of 25.4mm on one sensor and 7.6 on another sensor, yet both are considered 4K sensors. Starting to get the picture?

Let’s pretend for a moment that a 50mm lens is designed for this new OmniVision sensor and to keep costs down, they restrict the image produced by the lens to just barely cover the sensor with a 7.6mm image circle. It’s going to work great and produce a nice image that covers their entire 4K sensor without a problem. Now take that same lens and apply it to the 4K sensor of the Red One and it will still only proceed a 7.6mm image circle onto a sensor that requires a minimum 25.4mm image circle. It’s just going to produce an image as though one is looking through a key-hole in a door. Black all around with a small image in the very center. Now you’re getting the picture. To say that a lens “covers 4K” is fine for now because there are really only a handful of cameras that record in 4K. But once 4K becomes a common image resolution throughout the tech world and the motion picture industry, stating that something covers 4K isn’t going to get you very far.

The solution would be to stick with established formats such as Super35 or Super16 that represent a specific physical dimension. Granted they are archaic names that only exist because of the doomed emulsion film that the industry was built on, but they are organized standards that were established and propagated over and over throughout the industry for decades. Why change it now?