The reduction of noise in the spectra are critical to the performance of the spectrometer. The three primary components of noise in a CCD detector system are photon noise, dark noise, and read noise, all of which are temporal components and can be reduced by spectral averaging, which is symbolic of 1/N1/2 reduction in random patterns of noise though averaging multiple scans.   However, some noise components of dark noise are spatial or fixed pattern and cannot be reduced by signal averaging. Photon noise (sometimes referred to as shot noise) results from the inherent statistical variation in the arrival rate of photons incident on the CCD.  Dark noise is statistical variation in the number of electrons thermally generated within the pixel in a photon-independent fashion, and it creates a spatially fixed pattern noise that is minimized by cooling, and short integration times. In general, high-performance CCD sensors exhibit a one-half reduction in dark current for every 5 to 9° C as they are cooled below room temperature, a specification referred to as the doubling temperature.  In similarity to photon noise, dark noise follows a Poisson relationship to dark current, and is equivalent to the square-root of the number of thermal electrons generated within the image exposure time. The major contribution to read noise originates with an on-chip preamplifier, and this noise is added uniformly to every image pixel.  Under low illumination level conditions (and under the best cooling conditions), read noise is greater than photon noise and the image signal is said to be read-noise limited.  It must be considered that cost and size of the spectrometer increases with cooling reduction.  Cooling can also make heat dissipation and condensation in the final design more challenging.

Size or compactness of the spectrometer is dictated by geometry, and cooling electronics. 

Speed is governed by pre (onboard) or post processing (USB/ embedded or external PC). We provide an onboard processing option. 

Sensitivity is important if you have very little light from collected from your sample. A low f/# spectrometer should be considered with the option of double pass geometry, larger slit size, and high QE detector. 

Signal in SNR is dictated by the illumination power of the broadband or laser source, wavelength chosen for Raman and Fluorescence, light collection (low f/#), and the QE efficiency of the detector. Absorption and transmission spectroscopy methods are typically are not light starved, while light collected from the sample tends to be more of an issue with reflective, scattering (Raman) and fluorescence spectroscopy.  The noise in SNR is reduce by cooling, averaging, reduction of stray light, minimizing read noise, and having a high QE detector.

Dynamic range is the difference between the noise floor and the saturation range of the detector. Or the minimum and max signal  (in proportion to concentration) that a detector can read. The signal strength is determined by the full-well capacity (pixel size), and the noise is the sum of dark and read noises.  A short dynamic range is a result of non-cooled detectors. The amount of dark charge collected in each pixel is dependent not only on the device temperature, but also on the integration time and the storage time before readout. 

The dynamic range is greater than the linear range.  The linear range of a detector represents the range of concentrations over which the sensitivity of the detector is constant within a specified variation.