The debate concerning whether neural information is primarily transmitted via firing rate or precise spike timing remains central to neuroscience. Here, we tried to address this fundamental ambiguity in the primary motor cortex (MI) using a novel Brain-Machine Interface (BMI) task where rats were trained to voluntarily modulate neuronal population synchrony for reward. This operant conditioning paradigm allowed for the quantitative decomposition of the neural control signal into its coarse- and fine-timescale components. Behavioral analysis confirmed robust and stable control, marked by a synchronized elevation in population firing rate within the 300 ms pre-reward window. Temporal perturbation testing demonstrated that the coarse-timescale rate modulation alone provides a robust, near-threshold signal, confirming the capacity of firing rate for reliable information transmission. Crucially, we quantified the Rate-Independent Coincidence Spiking Pair count (RI-NCSP) —a metric designed to filter out chance coincidences driven by elevated firing rates—revealing a statistically significant Excess Synchrony that emerged exclusively during successful trials. Our finding proves that the precise spike timing contributes information independently of the firing rate increase. Finally, we utilized a multi-class Support Vector Machine (SVM) to assess feature utility. Decoding accuracy achieved with the combined feature of both components was significantly superior to that of either component alone. These results imply a Multiscale Hybrid Code architecture for neural coding, where the coarse rate component provides robustness and the precise timing component offers critical informational refinement. This framework is essential for the future design of BMI systems that leverage multiscale feature fusion for optimizing performance and efficiency.