This study proposes different symbol-level transmit processing methods for diverse multiuser MIMO setups. First, two symbol-level precoders are developed considering a strict per-antenna power constraint and PSK modulation for perfect and imperfect channel state information. Then, a large-scale MIMO setup is considered where the energy consumption of the radio frequency front ends yields a bottleneck for realizing energy efficient MIMO systems. With this, power reduction features such as constant envelope signaling and low-resolution quantization are applied to enable low-cost deployments, with low environmental impact, and better coverage. In this context, the minimum symbol-error probability formulation is con sidered as the design criterion for the case of QPSK data symbols, and, for arbitrary PSK modulation, the study proposes the novel minimum union bound symbol-error probability formulation. Based on these criteria the study proposes different low-resolution symbol-level precoders based on the partial greedy search method and the proposed quality-of-service branch and bound algorithm. Finally, a virtual multiuser MIMO system with PSK modulation realized via the reconfigurable intelligent surface-based passive transmitter setup is considered. Under this framework, this study considers both high-resolution and discrete phase shift reconfigurable intelligent surface models. With these frameworks, the study derives symbol-level power minimization problems under quality of service constraints. Both the symbol-error probability and union-bound symbol-error probability are considered for the quality of service formulation. The problems are solved by utilizing a bisection method and a branch-and-bound method for high and low-resolution reflecting elements, respectively.