Lanczos resampling formula/algorithm is defined as
Lanczos kernel- $$L(x)=\begin{cases}\operatorname{sinc}(x)\operatorname{sinc}(x/a), & \text{if $-a\lt x \lt a$}\\0, & \text{otherwise}\end{cases}\\ \text{where } \operatorname{sinc}(x)=\frac{\sin(\pi x)}{\pi x}.$$ $a$ is a positive integer which determines the size of the kernel.
Interpolation formula-$$S(x)=\sum_{i=\lfloor x \rfloor - a+1}^{\lfloor x \rfloor + a }s_iL(x-i)$$ where $s_i$ are 1D sample points.
How can I have a deeper understanding of it? I wanted to know that where does the intuitive idea of this formula come from?