Each filename pre-pended with the figure number of the main text that it contributes to. All files relevant to each figure are included. For experimental data, this includes the unprocessed measurement results and the Mathematica notebook used for their analysis. In the case of theory results the Mathematica script used to generate the numeric data for the plot is included. All numeric data is presented in comma separated variable (.csv) or tab-separated text files (.txt) which can be opened with standard spreadsheet or text editing programs. Mathematica notebook scripts (.nb) are known to work with Mathematica v12.0. Compatibility with older or more recent versions is untested. SEM images of the sample are provided as pictures in .tif format that can be opened by most standard image viewing software. ------------------------------------------------------------------------------------------------- Fig1: The figure is generated as the first plot in Mathematica notebook Fig 1.nb. Klyschko efficiency values for the desired lines on the plot are input manually into the plotting equation (currently set at values of 0.6, 0.7, and 0.8) ------------------------------------------------------------------------------------------------- Fig2b: Raw measurement data is split across two types of files - “Fig2b_APDxx.csv” and “Fig2b_CAMxx.csv” data with their corresponding numbers (xx). The Fig2b_CAMxx files give the total raw counts output from the Andor camera (a subset of pixels from the camera are summed in hardware to obtain the total intensity of light falling on the camera – every cell in the CAM file corresponds to total number of photons measured by the camera in a particular measurement). The Fig2b_APDxx files give the total number of counts measured by the APD detector in the same measurement window. Each pair of .csv files sharing the same numeric value in the filename corresponds to a unique set level of sample of loss (NB. the actual numeric value does not correspond to the value of loss). Each of the two types of .csv file contains 40x13 cells, with each one corresponding to a single experimental trial of 1 second acquisition time, at a level of applied sample loss set unique to that pair of files. The ordering of the cells is consistent between the two file types, such that the counts data in each cell of a Fig2b_CAMxx file is correlated with the counts data of the same cell in the corresponding APDxx. The data in each of these cell pairings are collected simultaneously in the same experimental trial. As an example, A1 of the Fig2b_CAM99 file corresponds to the same experimental trial as the counts data given in cell A1 of Fig2b_APD99. This pair of files will be processed in the Mathematica script to generate a single data point on the graph and its error bars. The raw data for each pair of CSV files is processed using the Mathematica script Fig2b.nb. The only parameters in the script that need to be input are the full file path names to the two files, Fig2b_CAMxx.csv and Fig2b_APDxx.csv. The Mathematica script outputs the systems precision ratio calculated according to equation 3 of the manuscript. The data is processed for each grouping of 40 measurments (each row), and the variance between the 13 separate data sets is then further used to determine the variance of the absorption estimate (indicated by error bars on the data points in the figure) The theoretical fit line for Fig2b is calculated by the same procedure as for Fig1, with the equation parameters determined by the experimental measurements. ------------------------------------------------------------------------------------------------- Fig2c: SEM image showing resolution test structures (Platinum deposition strips) ------------------------------------------------------------------------------------------------- Fig2d: Raw data for the figure is given in text file Fig2d.txt. In order, columns correspond to: -Position of X-Translation (mm); -Position of Y-Translation (mm); -Measured power on probe beam path, after sample (W) -Measured power on reference beam path (W) Data is processed using Mathematica script Fig2d.nb, using a similar approach to that described below for Fig3a. ------------------------------------------------------------------------------------------------- Fig3a: Raw data is given in text file Fig3a.txt. In order, columns correspond to: -Position of X-Translation (mm); -Position of Y-Translation (mm); -Measured power on probe beam path, after sample (W) -Measured power on reference beam path (W) Data is processed using Mathematica script Fig3a.nb (setting correct file path to data file at the top of the script). The script constructs fig3a using the ratio between the optical signal of the reference and probe beams. Note that the parameter BSR2 is experimentally determined (without a sample in place), is related to the reflection coefficient of the beamsplitter used and difference in relative losses of the two arms, and takes a value of 0.74. ------------------------------------------------------------------------------------------------- Fig3b: Raw data is given by the two text files Fig3bAPD.txt and Fig3bCCD.txt. In order, columns correspond to: -Counting index of the X-Translation stage position -Counting index of the Y-Translation stage position -Actual position of X-Translation stage (mm); -Actual position of X-Translation stage (mm); -Fifth column of the file, in the case of Fig3bCCD.txt, corresponds to total counts collected by the Andor camera within the 1 second acquisition time by a preset region of pixels, and in the case of Fig3bAPD.txt corresponds to total counts detected by the APD detector within the same acquisition window. Data for Fig3b is processed using Mathematica script Fig3b.nb (setting correct file path to data file at the top of the script). This script constructs the Figure according to equation 1 from the manuscript The global parameters in the script are: \[Eta]T =0.9 is the reference Klyshko efficiency DC =-100 are the dark counts on the reference arm detector \[Eta]M = 0.53 is the maximum klyshko efficiency \[Eta]Le = 0.98 is the transmittance after the sample ------------------------------------------------------------------------------------------------- Figure3c: Data is collected in a similar approach to Fig3a (differntial absorption measurement with a classical laser), but using the Andor camera as a detector rather than two powermeters. Raw data is given in Fig3c.csv, with columns corresponding to: -Counting index of the X-Translation stage position -Counting index of the Y-Translation stage position -Actual position of X-Translation stage (mm); -Actual position of X-Translation stage (mm); - Total counts detected in reference beam - Total counts detected in sample probe beam The data is processed by Mathematica script Fig3c.nb to generate the figure. The global parameter R is related to the reflectivity of the beam-splitter used to produce the reference power beam for the differential measurement, in addition to different losses for the probe and reference arms. Its value was determined experimentally to be R=0.975/0.28 ------------------------------------------------------------------------------------------------- Fig4: Data is collected and organised similarly to Fig3b, with the difference that the measurement at each pixel of the image is repeated 80 times, in order to evaluate the variance of the absorption estimate. For each of the two raw data files, Fig4apd and Fig4CCD, columns correspond to -Counting index of the X-Translation stage position -Counting index of the Y-Translation stage position -Actual position of X-Translation stage (mm); -Actual position of X-Translation stage (mm); -The remaining 80 columns each correspond to total counts collected in individual experimental trials for each pixel in the image (i.e. measurement is repeated 80 times). Similarly to the data from Fig2b, although the data is collected in two files, count rates total counts for the APD and CCD are collected simultaneously for each experimental trial, so data sets from the same columns in the two different files are correlated. Data is processed using Fig4.nb (again, setting the correct filepath near the top of the script for the two raw .csv data files). Experimentally determined parameters \[Eta]T, DC, \[Eta]M, and \[Eta]Le are the same as for Fig3b. The script generates the figure and the histogram using equation 1 from the manuscript. The precision ratio is obtained using equation 3 from the manuscript.