Merge branch 'p-digitizer' into 'master'

        LOG(TRACE) << "Creating output plots";

        // Plot axis are in kilo electrons - convert from framework units!

        int maximum = static_cast<int>(Units::convert(config_.get<int>("output_plots_scale"), "ke"));

        auto maximum = static_cast<double>(Units::convert(config_.get<double>("output_plots_scale"), "ke"));

        auto nbins = config_.get<int>("output_plots_bins");

        // Create histograms if needed

        h_gain = CreateHistogram<TH1D>("gain", "applied gain; gain factor;events", 40, -20, 20);

        h_pxq_gain = CreateHistogram<TH1D>(

            "pixelcharge_gain", "pixel charge w/ gain applied;pixel charge [ke];pixels", nbins, 0, maximum);

        h_thr = CreateHistogram<TH1D>("threshold", "applied threshold; threshold [ke];events", maximum, 0, maximum / 10);

        h_thr = CreateHistogram<TH1D>(

            "threshold", "applied threshold; threshold [ke];events", static_cast<int>(maximum), 0, maximum / 10);

        h_pxq_sat = CreateHistogram<TH1D>(

            "pixelcharge_saturation", "pixel charge with front-end saturation;pixel charge [ke];pixels", nbins, 0, maximum);

        h_pxq_thr = CreateHistogram<TH1D>(

                CreateHistogram<TH1D>("pixelcharge_adc", "final pixel charge;pixel charge [ke];pixels", nbins, 0, maximum);

        }

        int time_maximum = static_cast<int>(Units::convert(config_.get<int>("output_plots_timescale"), "ns"));

        h_px_toa = CreateHistogram<TH1D>("pixel_toa", "pixel time-of-arrival;pixel ToA [ns];pixels", nbins, 0, maximum);

        auto time_maximum = static_cast<double>(Units::convert(config_.get<double>("output_plots_timescale"), "ns"));

        h_px_toa = CreateHistogram<TH1D>("pixel_toa", "pixel time-of-arrival;pixel ToA [ns];pixels", nbins, 0, time_maximum);

        // Create time-of-arrival plot with different axis, depending on whether TDC simulation is enabled or not

        if(tdc_resolution_ > 0) {

        double integrated_charge = 0;

        for(; bin != pulse.end(); bin++) {

            integrated_charge += *bin;

            if(integrated_charge >= threshold) {

            if(std::abs(integrated_charge) >= threshold) {

                break;

            }

        }

* A Gaussian noise is added to the input charge value in order to simulate input noise to the preamplifier circuit.

* The preamplifier is simulated by applying a gain function to the input charge, or by multiplying the input charge with a defined gain factor.

* An optional, very simplistic, front-end saturation can be simulated which replaces the measured pixel charge with a value drawn from a Gaussian distribution with the configured saturation mean and width if the charge measured is larger than the calculated saturation value. This follows the approach taken in \[[@holmestad]\]. The pixel charge is compared to the smeared saturation value in order to generate a smooth transition rather than an edge in the spectrum.

* A charge threshold is applied. Only if the threshold is surpassed, the pixel is accounted for - for all values below the threshold, the pixel charge is discarded. The actually applied threshold is smeared with a Gaussian distribution on an event-by-event basis allowing for simulating fluctuations of the threshold level.

* A charge threshold is applied. Only if the threshold is surpassed, the pixel is accounted for - for all values below the threshold, the pixel charge is discarded. The actually applied threshold is smeared with a Gaussian distribution on an event-by-event basis allowing for simulating fluctuations of the threshold level. It should be noted that only positive threshold values are possible, and that this threshold will be compared to the absolute of the charge. This therefore both works for positive and negative inputs.

* A charge-to-digital converter (QDC) with configurable resolution, given in bit, can be simulated. For this, first an inaccuracy of the QDC is simulated using an additional Gaussian smearing which allows to take QDC noise into account. Then, the charge is converted into QDC units using the `qdc_slope` and `qdc_offset` parameters provided. Finally, the calculated value is clamped to be contained within the QDC resolution, over- and underflows are treated as saturation.

The QDC implementation also allows to simulate ToT (time-over-threshold) devices by setting the `qdc_offset` parameter to the negative `threshold`. Then, the QDC only converts charge above threshold.

* A time-to-digital converter (TDC) with configurable resolution, given in bit, can be simulated if pulse information is available from the input data. If the necessary pulse information is available from the input data, e.g. by using the PulseTransfer module to generate PixelCharge objects, this module calculates the time-of-arrival (ToA) as the time when the integrated input charge crosses the threshold.

* A time-to-digital converter (TDC) with configurable resolution, given in bit, can be simulated if pulse information is available from the input data. If the necessary pulse information is available from the input data, e.g. by using the PulseTransfer module to generate PixelCharge objects, this module calculates the time-of-arrival (ToA) as the time when the integrated input charge crosses the threshold. Also here, the absolute of the integrated charge is compared to a positive threshold value to be independent of the signal polarity.

First, the time from the start of the event until the first crossing of the charge threshold is calculated. It should be noted that this calculation does not take into account charge noise simulated in the QDC. The resulting ToA is smeared with a Gaussian distribution which allows to take TDC fluctuations into account. Then, the ToA is converted into TDC units using the `tdc_slope` and `tdc_offset` parameters provided. Finally, the calculated value is clamped to be contained within the TDC resolution, over- and underflows are treated as saturation.

If no time information is available from the input data, a time stamp of 0 is stored.