We propose a novel system-level error tolerance approach specifically targeted for multimedia compression algorithms. In particular we focus on the motion estimation process performed by most video encoders. While current manufacturing process classifies fabricated systems into two classes, namely, perfect and imperfect, our proposed scheme employs categories which are based on acceptable/unacceptable performance degradation. By enabling the use of systems that would otherwise have been discarded we seek to increase the overall yield rate in the system fabrication process. To achieve this, we propose testing algorithms that aim at determining if faults in a given chip produce acceptable performance degradation, and we propose a technique which can cancel the effect of those among the acceptable faults that can be compensated.