We would like to think these are frequently asked questions (FAQS), but let’s be honest: how many people really ask about this stuff? So we prefer Theoretically Asked Questions (TAQs), which is pronounced “tax”. These questions / issues do come up now and then, and perhaps you are among those who have asked these questions.
CMOS and CCD Sensors We’ve Used
Early on, we used frame transfer CCD technology to acquire images quickly without excessive intra-frame blurring. These provided pristine images, especially when compared to the early CMOS devices, which some would describe as “swill”. However, in the past 5-10 years, CMOS devices have displaced the CCD in all but a few applications, owing to the clever correction of KTC noise, among others. The improvements in quantum efficiency sensors have included frame transfer CCD and multi-transistor CMOS focal plane arrays having high dynamic range. Typical performance for our cameras includes 100-500 Hz frame rates, 12-bit digital data with 1 LSB noise, and formats ranging from 256×512 to 1024×1280.
What Constitutes “High Speed”?
TSG has delivered several variants of high speed cameras as OEM components for products requiring large format, high speed, and low noise. For us, high speed has meant 300-500 frames per second at full frame (the same devices can frame at many kHz if partial frames are allowed). When coupled with low noise, which for us means 1 LSB (of 12 bits) or better, a very useful performance regime is reached for surveillance applications.
Why Use Fiber Faceplates ?
We are tempted to say “because they are cool”. But the real reason is that a faceplate allows for contact imaging. Why is that important? Because contact imaging helps capture all the photons emitted by a phosphor, such as one might find on an electron tube device, much like night vision googles. So very low light phenomena can be imaged, even if only 1-2 ms are available for capturing the image. There are a good number of companies that know how to bond fiber faceplates. We are one of the few that makes the effort to do micron scale alignment and “live” MTF optimization across the focal plane, all while assuring optimized performance across shock, vibe, and wide temperature spans.
What is a Smart Camera?
Many DSLR manufacturers pack their cameras with really helpful features that involve some image processing and thereby dub them “smart cameras”. That’s not what we mean when we use those terms. For us, a camera becomes smart when it has an accessible operating system (OS) and can be programmed using a commercially available language to answer high level image questions such as “is there a person standing over there?”, or even “is that flashing light an LED or is that a glint from the sun?” Commercial languages we like to use include C, C++, Python, R, and OCaml.