Aries 6504 Pro
The Aries 6504 Pro is a 4-million-pixel sCMOS camera with 6.5-micron pixels and a peak QE of 95%.
On paper, it's the most sensitive sCMOS camera. But is it?
The answer is, as always, it depends.
Below are some example images and information explaining how, for low light imaging where the highest SNR (Signal-to-Noise) for detection is the aim, we believe it is.
What is sensitivity?
When we think about sensitivity we think Quantum Efficiency and Noise (Dark Current and Read). However, these don't address how users think about sensitivity in relation to their application. We find there are really 3 types of customer-identified sensitivity: Total Signal, Visibility and Detection
Total sensitivity requires a particular signal level. For example, I need 10,000 electrons of signal (SNR >20:1). Visibility is much more about low signals being imaged with statistical confidence (SNR 5:1). The last being detection where you need to see your target well enough to know whether it was there or not, and clearly see shapes, features, and structures.
We will be focusing our sensitivity analysis and comparison in the detection regime.
Max Detection
sCMOS has been able to deliver read noises of 1.1 electrons for over a decade using a technique called CMS (Correlated multisampling). In this case, 2 gain amplifiers are used, not to create a HDR image, but instead simply to measure the same signal twice, thereby reducing the noise. This slows the camera & reduces the bit depth and full well capacity. All of these trade-offs combined have meant existing sCMOS never truly compared to EMCCD for detection until now.
The Aries 6504 Pro achieves 0.4 electron read noise in 16-bit and at much faster speeds.
Sensitivity at Speed
By reducing data to 11 or even 8 bits, some sensors allowed customers to image at up to 100 fps at full frame, and faster with regions of interest.
The Aries 6504 Pro can achieve 4 X lower read noise for the same 100 fps, or you can use the high speed mode which delivers 297 fps full frame with 0.8 electrons of read noise.
The 297 fps high speed mode is available on both the Aries 6504 and our exclusive Aries 6504 Pro.
Low Noise & Reduced Patterns
Several cameras achieve sub-electron read noise by slow scanning. Opposite, we see a direct comparison with a competitor's 5.3 MP sCMOS featuring the same size pixel (6.5 micron) with 95% peak QE.
Not only is the SNR of the Aries higher and the frame rate is 15 X faster, we are also not seeing the high levels of pattern noise which distract from the aim of detection and visibility.
Detection @ 5 Photons
Not all wavelengths are equal and this is why sensors have a spectral response shown by the QE curve.
So what would happen if we calculated the response to a low signal, with various cameras at differing wavelengths? We've selected a signal of 5 photons as it should allow us to achieve detection level sensitivity (SNR 2:1) equivalent to these shown in the examples.
Let's start by looking at responses to some typical emissions in fluorescence microscopy for blue, green, red, and far red using DAPI, Alexa Flour 588, 994, and 647.
This following assessment will model the detection sensitivity across the spectrum for general fluorescence microscopy.
Fluorescence Microscopy SNR Response
The expected SNR response shows the Aries 6504 Pro achieving higher SNR than all other 6.5 micron pixel sCMOS cameras available on the market at the noted wavelengths.
- Dark Blue – Aries 6504 Pro
- Orange – All other 4 MP 95% sCMOS
- Green – Aries 6510 and other 10 MP 95% QE sCMOS
- Light Blue – 5.3 MP 95% QE sCMOS
Comparison to qCMOS
In recent years, we've seen a new technology self titled as qCMOS. This achieves some really exciting read noise values and some excellent statistical data relevant to measurements of single counts. However, these sensors actually achieve lower QE in the fluorescence microscope emissions selected.
Even though lower read noise is possible in slow scan modes, the weight of the QE is enough for the Aries 6504 to deliver higher SNR.
Dark blue – Aries 6504 Pro
Light blue – Other qCMOS
Pixel Size Impact
Pixel Size does not make a camera more sensitive. The 2 main drivers for a camera's sensitivity are QE and noise (dark current and read noise). However, pixel size can have a massive impact on system sensitivity; getting more light into a pixel helps drive higher SNR, albeit at the risk of reducing achievable resolution.
When changing between cameras of different pixel size on a microscope system, the observation is clear.
Here we can see the real SNR difference between the Aries 6504 Pro and the qCMOS when pixel size is taken into account.
The larger pixel delivers more light. More light means a higher SNR, and a higher SNR equals more detection sensitivity.
What about EMCCD?
Although EMCCD has a similar peak QE for most applications, in fluorescence microscopy it drives much of its system sensitivity from pixel size being 4-6 X larger than a 6.5-micron pixel, depending on the model. To make this a more reasonable comparison, let's try to match all the pixels to the same size using binning.
- Dark Blue – Aries 6504 Pro (Bin 2)
- Red – EMCCD (13 micron Bin 1)
- Green – qCMOS (Bin 3)
The lower read noise of the Aries 6504 Pro, combined with the fact the EMCCD suffers excess noise, means the Aries 6504 Pro clearly wins on SNR. The smaller pixel size of qCMOS means additional binning is required to achieve the same pixel size, increasing its read noise.
Outside Fluorescence Microscopy
All comparisons discussed so far focus on fluorescence microscopy, and, for simplicity, we've only covered a few typical wavelengths. But what about sensitivity in other applications? Well, as noted at the start, it depends.
Here we can see the quantum efficiency curves of the Aries 6504 and qCMOS cameras along with the major emissions for atom imaging. And, as you can see, sometimes you win, sometimes you lose.
That's why we always say "the most sensitive camera?" The question mark is the key; cameras fit applications, but no camera always wins.
Summary
The Aries 6504 Pro sCMOS camera delivers the highest peak QE in association with the lowest read noise.
The Aries 6504 drives higher SNR and so outperforms all other 6.5-micron pixel sCMOS for "Detection Sensitivity" in the example wavelengths shown.
We think this is the ideal camera for detection sensitivity, but that does not instantly make it right for you. We have larger sensor formats, faster cameras, global shutter devices, and TDI cameras.
Always try to speak to an imaging consultant.