Quantum computation is expected to outperform classical computation, yet understanding the origins of this advantage remains a fundamental challenge. In this talk, I will focus on the quantum feature, called magic, which can support the quantum advantage. I will introduce a quantum convolution to test and measure magic and discuss the possible future directions in this framework.