How can I reduce the noise in my spectra?
There are several ways that you can possibly reduce the noise in your spectra.
- TEC: If you have a spectrometer that utilizes a thermoelectric cooler such as a NIRQuest or QE spectrometer, please be sure that the TEC and fan are enabled and it is set to the recommended setpoint.
- Unprocessed mode intensity: The best measurements are made when the unprocessed mode intensity in counts is between 80 and 90% of the number of counts where the detector becomes saturated. Please check the region in which you are seeing the noise to see if the counts are between 80 and 90% of where the detector becomes saturated (we use 85% as the “recommended peak value”). If the counts are much lower than this, please increase the integration time to get the signal close to or within this level. Please note that increasing integration time increases the number of counts through the entire spectrum and depending on the grating in your spectrometer and the light being measured, there may be some region that becomes saturated when the integration time is increased. Any pixels that are saturated will not yield good data. The data is scaled so that a spectrometer’s detector becomes saturated at 2 to the power of the number of bits of the A/D converter minus 1.
- Scans to average: The Scans to average default is 1, which means that the spectrometer does one scan and that is your acquisition. The more scans to average you use, the greater the signal to noise ratio. Please note that your total acquisition time will be the number of scans to average multiplied by the integration time.
- Boxcars: Using a small number of boxcars will increase the signal to noise ratio; however, it will also flatten the peaks. It is good to use between 1-5 boxcars.
- Integration time:As you increase integration time, the percentage of noise in the signal will increase. We define the maximum integration time of our spectrometers as that integration time whereby the noise comprises 50% of the total signal. If you can use a larger diameter fiber, etc. to be able to use a shorter integration time.
- Detector Temperature: Noise increases with increases to detector temperature. If your spectrometer does not have a TEC, try to work in a cooler temperature environment.