The impact of a buoyancy driven oil droplet with an oil-water interface is investigated using timeresolved Particle Image Velocimetry (PIV) along with a phase discrimination by means of highspeed Laser Induced Fluorescence (LIF). In this paper we focus on the investigation of strategies to optimize the performance of high-speed PIV algorithms. Furthermore this data will be used for validation of numerical simulations of two phase flows. In order to simultaneously measure the flow velocities inside and around the oil droplet by means of PIV the refractive indices of both phases need to be carefully matched. The aqueous phase consists of a mixture of corn syrup and water, which defines the viscosity as well as the refractive index. The disperse phase consists of a mixture of two kinds of mineral oils. The latter are mixed to match the refractive index of the continuous phase. Both phases are seeded with tracer particles required for PIV. A fluorescent dye is added to the dispersed phase to allow discrimination of the PIV signals originating from both phases by means of LIF. The LIF and PIV signals are captured by two aligned, synchronized highspeed cameras, each used for one of the measurement techniques. In order to study the impact of an oil droplet into an oil-water interface, it is necessary to accurately measure a velocity range that is spread over three orders of magnitude. The impact of the droplet is measured in a time series of PIV recordings with a high temporal resolution. This allows to optimize the time interval between two correlated frames within the time series to achieve a high signal-to-noise ratio, while still being able to measure a large velocity dynamic range. An approach to estimate an optimal time separation is presented in this study. The results are compared to the experimental data of Mohamed-Kassim and Longmire [1] and the effectiveness of the introduced optimization on the data quality is discussed. Furthermore the results of this experiment are compared to a numerical simulation developed by Coyajee et al.[2].