All PHT raw data undergo the same processing steps to obtain the mean signal per chopper plateau or mean signal per raster point. These steps are described in detail in Chapter 7.
Single pointing photometry AOTs include one FCS measurement for each detector used. The FCS measurement was taken after completion of the filter sequence for a given detector (see Section 3.10).
Assuming that the derived responsivity applies to all filters of the same detector, the signal from each filter is converted to an in-band power. The powers are converted to a flux density in Jy or MJy/sr using equations in Sections 5.2.5 and 5.3.
For AOTs in mapping mode each raster measurement is bracketed by two
FCS measurements with identical FCS heating power. For the calibration
it is assumed that the responsivity does not vary with time. The average
responsivity of the two FCS measurements is used for each raster point.
Since the map plus FCS measurements are obtained for one filter at a
time, no systematic uncertainties due to 
are involved.
During a sparse map concatenated chain (PHT17/18/19 for PHT-P or PHT37/38/39 for PHT-C) two FCS measurements were obtained, the first measurement after the first sky measurement in PHT17/PHT37 and the second one after the last sky measurement in PHT19/PHT39. The FCS in-band powers are adapted to the flux levels as specified in the start and end AOTs and can therefore be different FCS heating power settings. The average responsivity of the two FCS measurements is used for the responsivity of all measurements in the chain.
The multi-filter option can be used in the sparse map AOTs. In such case the two FCS measurements can be of different filter band and a transfer of the calibration is required to the filters for which no FCS calibrations were obtained.
In this mode one FCS measurement was collected for the last (largest) aperture in the sequence. An important source of systematic photometric uncertainty is the accuracy in the aperture scaling (see Section 5.4). Uncertainties in the beam profile can be minimized by comparison with a similar observation on a point source.
For PHT05 and PHT25 only one filter band could be selected and the FCS measurement has the same integration time as the source measurement. The observer could therefore minimize statistical and transient uncertainties. Uncertainties due to dark signal and FCS zero signal level could be minimized by including dedicated dark and cold FCS measurements in the observation. Since the absolute photometry AOTs were single filter observations, the systematic photometric accuracy only depends on the uncertainties in the FCS power calibration tables and either in the illumination matrix for PHT-C or the aperture scaling for PHT-P (both for FCS and sky).
The instrument set-up of a PHT40 observation is always the same. After dark signal subtraction, the signal of each pixel is directly converted to a flux density in units of Jy. The PHT-S spectral response function was determined from observations of calibration stars (Section 5.2.6). In-orbit beam profile measurements were used to determine the conversion from a point source flux to an extended source flux.
Each target measurement with PHT40 is preceded by a `pseudo' measurement of 32 seconds in dark instrument configuration. This measurement offers a qualitative assessment whether the responsivity of the PHT-S detectors is affected by a transient introduced by a preceding PHT40 observation.
For example, in case the preceding PHT40 observation involved a source with bright spectral features, the pixels that detected the features could still be in a transient while the pseudo dark was taken. It is also found that different pixels in the PHT-S array show different transient behaviour. This can cause artifacts in the spectrum, especially after observing a strong continuum source.
Possible features in the dark spectrum at a certain wavelength should caution the observer that features in the same pixels of the actual source spectrum could have an instrumental cause. The dark measurement in PHT40 is not intended for subtraction from the source spectrum. Instead, an orbit dependent dark signal from dedicated calibration measurements is subtracted.