EMCCD vs sCMOS
Compare electron-multiplying CCD with sCMOS for low-light imaging.
EMCCD SNR
2.249
sCMOS SNR
2.857
Winner
sCMOS
ENF²
1.9763
EMCCD: σ = √(F²·(S + D) + (σ_read/G)²), F² = 2(G − 1)·G^(−(N+1)/N) + 1/G
sCMOS: σ = √(S + D + σ_read²)
Model and references
Textbook approximation. A standard approximation from the literature (for example paraxial, plane-wave or Gaussian-beam), valid within the range the page states. About the tiers
Both pixels see Poisson signal and dark charge. The EMCCD multiplies them with the Robbins–Hadwen excess noise of an N-stage register and divides its read noise by the gain; the sCMOS adds its read noise as a Gaussian. Equal QE is assumed (enter signal in electrons), and clock-induced charge, the sCMOS read-noise tail and photon-counting readout are left out, and one dark-charge value is used for both sensors (a deep-cooled EMCCD is usually far lower). EM gain wins only while σ_sCMOS² exceeds about (F² − 1)·S.
- Robbins M. S., Hadwen B. J. (2003). The noise performance of electron multiplying charge-coupled devices. IEEE Trans. Electron Devices 50, 1227–1232. F² = 2(G − 1)·G^(−(N+1)/N) + 1/G. doi:10.1109/TED.2003.813462
- Janesick J. R. (2007). Photon Transfer: DN → λ. SPIE Press. Shot, dark and read noise of a sensor pixel. doi:10.1117/3.725073