The technology in complementary metal-oxide semiconductors (CMOS) sensors has been around for a long time. These sensors are inexpensive to manufacture and have made their way into the many imaging devices found in your pockets (phones, cameras, MP3 players, etc.).
As a technology, CMOS is not merely relegated to funny cat pictures or a video of your buddy’s failed attempt at an American Ninja Warrior course. Instead, CMOS has evolved from a simple sensor design to a niche high-speed camera, all the way to a robust technology that benefits a wide variety of microscopy applications as diverse as time-lapse applications, to cell trafficking, to light sheet microscopy.
The coming of age of CMOS happened a few years ago with the launch of a new Fairchild sensor design incorporated into cameras such as the Hamamatsu Flash 4 v2, pco.edge, and Andor Neo/Zyla. What has been coined scientific CMOS (sCMOS), in many arenas, has overtaken CCD as the gold standard for fluorescence imaging.
Although the new generation of CCD sensors has its place alongside sCMOS, as noted in a previous blog post, sCMOS technology has eroded, if not completely replaced, CCD’s position as the preferred sensor technology for advanced imaging applications.
Just in time for the 45th meeting of the Society of Neuroscience in Chicago, IL, Hamamatsu, pco, and Andor have announced major technological advancements in their respective sCMOS cameras. Photometrics announced its entry into the sCMOS market with very unique shot noise reduction technology.
For Hamamatsu, pco, and Andor, the release of the new cameras include sensors that employ improved microlenses. The microlens improvement increases the overall QE of the sensor by 5-10% across the visible light spectrum. This offers users the ability to capture more light in less time, increasing signal-to-noise ratio, shortening exposure times, and increasing frame rates.
Although three companies have announced new cameras, only Hamamatsu has access to the new chips now, where other companies will not have access for several more months.
Without a sCMOS camera to offer microscopists until last week, Photometrics has tossed an 80 yard touchdown-scoring bomb into the sCMOS market. Named the Photometrics PRIME, the standard 4.2 megapixel sCMOS sensor used in the vast majority of cameras in this class has been juiced with noise and data-reducing algorithms. These advanced features are unique and stay with the tradition of advanced technology in flagship cameras from Photometrics.
In low light fluorescence imaging, one of the most important aspects of the detector is signal-to-noise ratio (SNR). Although the equation to calculate signal-to-noise ratio (SNR) is complicated, the concept is simple: How well does the camera sensor read out signal above the level of electronic noise?
Smart engineers from the various companies have been working on improving SNR and appear to have addressed it in force. The new 82% QE sensors offer an incremental improvement. However, with Photometrics Prime Enhance technology, Photometrics reports an improvement of 3X-5X in SNR. The data provided in Photometrics technical notes provide a glimpse at what is possible. And what is possible is amazing!
A characteristic of sCMOS that has always been attractive is high frame rates. If you thought a base, full resolution frame rate of 100FPS was impressive, the Hamamatsu Flash 4.0 v2 can achieve over 25,000 FPS with ROI. These frame rates, however, are exposure-time limited. What is the most common cause of limited exposure time? Photons.
With a higher QE in the Hamamatsu Flash 4.0 v2 PLUS and the Prime Enhance technology from Photometrics, sCMOS cameras are set to challenge the conventional wisdom that EMCCD’s are required for low light, high speed imaging.
As documented in the technical note, Photometrics Prime Enhance technology can generate equivalent data in 100 milliseconds for what would normally be 800 milliseconds in standard operation! Take that, EMCCD!
It is worth noting, however, that although Prime Enhance generates clean and beautiful data, the same cannot be said for the visual image. The Prime Enhance algorithms reduce noise in the pixel gray values but, because noise is reduced by factoring in neighboring pixels, the final result is an image with a Photoshop Palette Knife appearance. This is most noticeable as signal decreases to “near noise” levels, but incredibly, the grey level histogram still looks good. If beautiful images are what you are after, fear not, Prime Enhance can be turned off, exposure time extended, and a beautiful image will result. However, if you want to go fast in low light, Prime Enhance makes it possible!
At 4.2 megapixels, 65,536 gray levels (16 bit depth), and 100 FPS, the current generation of sCMOS cameras generates a lot of data! Localization-based super resolution systems are already using sCOMS cameras, which is why the Photometrics PRIME has two more tricks up its sleeve: Prime Locate and Multi-ROI.
Prime Locate allows the data transfer of only the pixels which register a grey value in localization-based super resolution systems. Considering many of these systems generate 60,000 – 100,000 images before building the journal cover-worthy super resolution image, the data savings will be tremendous. This technology also increases frame rates, lowers file size, and reduces storage concerns.
The Multi ROI function in the Photometrics Prime also allows users to capture multiple regions of interest (ROI) in a single field of view. So if the user has two small features in one huge field of view, leave the empty data on the microscope and only acquire the ROIs. Reducing file size and collecting more data, what could be better?
With the introduction of new sCMOS sensors and sCMOS sensor technology, the market is rapidly changing. The new 82% QE sCMOS sensors have brought more than just high performance. There are now several variants that differ on price, cooling, triggering, resolution, shutter, interface, and now quantum efficiency.
While the new Flash 4 V2 Plus is sure to be priced at a premium, companies such as pco have released USB3 cameras at a much lower price, which are based on the current and popular sensor. Whatever your priority, there is a sCMOS camera to step up to the challenge.
Although sCMOS has firmly positioned itself as the premier technology for high-end fluorescence imaging, it does not cover the entire range of scientific imaging – particularly when price and application are considered. Many investigators are not going to be able to justify the price premium on sCMOS cameras for features that will be rarely used on, for example, a routine fluorescent stereo microscope.
Although CCD has fallen out of favor for high-end widefield acquisition, it still has its place on microscopes. Recently QImaging released the Retiga R1, R3, and R6 cameras at shockingly low prices. With the Retiga EXi, what used to cost $12,000 three months ago has bottomed-out at less than $5,000 with the introduction of the new Retiga R1. In addition, the new R1 has deeper cooling, higher QE, and a higher frame rate for live cell imaging!
Interestingly, this opens the door to simultaneous, multi-channel, imaging applications that require several cameras and employ the Multi-Cam from Cairn Research. What used to be a $40,000-$60,000 investment now costs a fraction of the price because of QImaging’s new CCD cameras!
Choosing a camera can be intimidating, but identifying the needs for the application is the first step in making a smart decision. When evaluating technology for your lab, which will be used for many years into the future, it’s important to consider the latest products and technology advancements.
The Imaging Specialists at W. Nuhsbaum, Inc have seen cameras evolve over the years and can provide perspective on the latest technology to arrive on the market. Trust the experience of W. Nuhsbuam, Inc to weather the technology winds of change and advise on the proper technology for your experiments.