The resolution and sensitivity were estimated using phantom measurements. For this purpose, a phantom consisting of a 20-mm cylindrical agar–milk matrix (made from a solution with 1.8 g agar and 40 ml water to which 2 ml milk was added) was built that contained cylindrical holes with different sizes, one copper wire and a black human hair. The diameters of the holes, which had circular cross sections, were 1, 2, 3, and 5 mm, and they were filled with black gelatin (1 g gelatin in 10 ml water and black ink), water to which black ink was added, transparent gelatin (same concentration as before), and rapeseed oil, respectively. The SOS of the various materials was evaluated separately by means of TOF measurements. Their values were for the agar matrix, for gelatin, for rapeseed oil, and for water. A photograph of the phantom is shown in Fig. 3(a), where the dimensions of the insets and the corresponding SOS values are also given. This configuration of materials and size distributions was chosen for determining the limit regarding the size of the reconstructed SOS distributions. As the phantom had a constant cross section in the -direction, it could be regarded as a 2-D object. Therefore, scanning in the vertical direction was not necessary and just one selected plane was imaged. For dual-mode PA and pulse-echo LUS imaging, the cylindrical sample was rotated five times and then the data were averaged prior to image reconstruction. As dual-mode and transmission imaging were performed sequentially, an inclined polymer plate was placed in front of the staircase-shaped polymer target during data acquisition of the PA and LUS signals. This way, US waves reflected at the staircase-shaped target that would arrive at the detector at the same time as the pulse-echo LUS signals traveling via the acoustic mirror were deflected away from the sensor. For SOS measurements, only the 20 external absorbers arranged in front of the sample were illuminated, the other light sources were switched off and no averaging was used.