Spectral Processing

Processing control is responsible for real-time processing and manipulation on data that have been integrated by Integration component. Processing specifies:

• If and how 3D data are reduced to 2D (pooled);
• What calculations are performed on data;
• Reference data that are used in calculations.

Mode tab

Mode tab specifies general data processing parameters, including data pooling and calculation mode.

Data pooling

Data pooling applies only when 3D data are arriving from the Integration. For example, when CCD is operated in an Image mode (as opposed to Spectra mode ) it produces data matrix – 3D dataset. During Processing step this array can be either left as is, in which case calculations specified by Processing will be applied to the entire matrix, or matrix can reduced to a 2D array before further processing.

Data pooling is the first step in data processing. Data pooling, if specified, occurs before any calculations are done. Type and dimension of reference data (such as Dark signal etc.) must be identical to pooled data. Hence, any changes to Data pooling settings results in invalidation of reference data.

Simplest form of Data pooling is pixel averaging, where data values in each column of data matrix are averaged together. While this is similar to data binning by imaging Detectors, pooling is a purely digital manipulation and does not limit well depth of the detector.

Processing mode

In principle, post-processing calculations on data may involve any imaginable manipulation. In practice, however, there is only one Processing mode that is really used – calculation of light absorption (optical density).    

Scope mode

Scope mode is the simplest data processing mode (or lack of thereof) which results in raw, unprocessed data. Exact meaning of dimensionless intensity count depends on the nature of the detector.

In spectral measurements Scope mode represents raw count of photoelectrons produced by the detector. Number of such photoelectrons is proportional to the intensity of light multiplied by the quantum yield of the detector at that wavelength. Some measurements, such as in Raman spectroscopy, for example, operate only in Scope mode. In other cases intensity in the Scope mode may correspond to potential, resistance, accumulated charge etc…

Less Dark mode

Dark current is signal that is produced by the detector without any probe light. There are several sources of dark current, including thermal noise of the detector, stray light that enters spectrometer etc. There are two properties of dark current that should be taken into account regardless of their source:

• Dark current should be kept constant during measurements. For example, keeping ambient illumination (room light etc.) will avoid changes in dark current due to stray light.
• Dark current must be subtracted from measured signal during typical calculations. Whether data will be presented in transmission, absorption, or any other form, unaccounted static background is likely to cause distortion in signal shape.

In spectral measurements, contribution of dark current is strongest in the areas of intense background absorption where amount of transmitted light is small. Dark current is the key factor that defines the upper limit of dynamic range even when it property subtracted.

Less Dark mode is the basic mode that allows subtracting reference Dark spectrum from current data. Less Dark mode is disabled until a valid Dark spectrum is measured.

Absorption mode

Absorption mode is the workhorse of spectral transmission measurements. By definition, absorption is negative logarithm of ratio of transmitted over incident light. Because both the sample and the reference spectra contain dark current when they are measured, dark current must be subtracted for proper absorption calculation:

O.D. = -log10[(Sample – Dark) / (White – Dark)]

Absorption mode remains disabled until valid White and Dark data are measured or loaded. 

Reference data

Before Absorption mode in Processing step can be enabled, both dark current and white (reference) data must be measured. The order, in which these data are recoded, does not matter, but measurement parameters, conditions, and pooling settings should be as close as possible to subsequent sample measurement.  Dark and White data are acquired by pressing measurement buttons on corresponding tabs in Processing. These acquisitions are treated identically to single sample measurements except that resulting data are stored internally in Processing and are never delivered to any Destination document.

An alternative to measuring reference directly from the Detector is to load reference data from an opened data document, although this option is currently disabled. When loading data from a document, one must ensure that document contains raw (unprocessed) data in correct format and dimension – just as would be acquired from the Detector.

Duration of reference measurements

As with any measurement in Single mode, duration of Dark and White measurements is controlled by Average parameter in the Integration rather than total duration value in Acquisition control.

It is often beneficial to set Average parameter for reference measurements to a value several fold higher than will be used during subsequent sample measurements. This applies to most, if not all, detector modes. Longer accumulation of reference data often can improve S/N ratio during absorption measurements by reducing noise in calculations. This is especially relevant for fast, time-based measurement where longer accumulation generally means lower time resolution.

When reference data are valid

For proper calculations, format and dimensions of reference data must be identical to that of subsequent measurements. To avoid erroneous results, reference data are invalidated and purged whenever key Detector or Processing parameters change. In addition, it is necessary to re-acquire reference data every time integration time or any optical properties (such as slit, area etc) change.   

Information

White and Dark tabs provide two information areas:

• Accumulation parameters reports timing and repetition parameters used in acquisition. Note that currently this information does not account for the possible repetition within the Detector itself, only duration of single acquisition and the Average parameter from Integration.
• Source of reference data: either >>detector<<, if data were directly acquired, or the name of the document, if data loaded.