7

Recall $\text{sinc}(x)=\frac{\sin x}x$. It's a familiar exercise that $\int_0^{\infty}\text{sinc}(x)\,dx=\frac{\pi}2$.

But, at present, I wish to ask about the following claim on a "sinc-ing" product which is supported by extensive numerical computations.

Question. Is it true that $$\int_0^{\infty}dx\prod_{n=1}^{\infty}\text{sinc}\left(\frac{x}{2n-1}\right) =2\int_0^{\infty}dx\prod_{n=1}^{\infty}\cos\left(\frac{x}n\right)\,\,?$$

T. Amdeberhan
  • 41,802
  • 2
    an answer to this question was given as a part of an answer to an earlier of your questions https://mathoverflow.net/questions/259049/this-is-not-a-dyadic-cosine-product . You forgot it, right? Like you forgot the integral from your own paper https://mathoverflow.net/questions/263397/is-there-a-transformation-or-a-proof-for-these-integrals ? – Martin Nicholson Aug 24 '21 at 13:30
  • 1
    Meta question: https://meta.mathoverflow.net/questions/5116/ – Stefan Kohl Aug 24 '21 at 18:07

1 Answers1

20

We can start with a substitution $x=2y$ $$\int_0^{\infty}dx\prod_{n=1}^{\infty}\text{sinc}\left(\frac{x}{2n-1}\right)=2\int_0^{\infty}dy\prod_{n=1}^{\infty}\text{sinc}\left(\frac{2y}{2n-1}\right).$$ Now the double angle formula $$\text{sinc}(2a)= \text{sinc}\left(a\right)\cos(a)$$ can be iterated to arrive to the familiar formula for $\text{sinc}$ $$\text{sinc}(2a)=\prod_{i=0}^{\infty}\cos\left(\frac{a}{2^i}\right).$$ This means that $$\prod_{n=1}^{\infty}\text{sinc}\left(\frac{2y}{2n-1}\right)=\prod_{m= 1}^{\infty}\cos\left(\frac{y}{m}\right)$$ because every $m\in \mathbb N$ can uniquely be written as $2^i(2n-1)$ for some $i\geq 0$ and $n\geq 1$. And this, in turn, proves the integral identity.

Gjergji Zaimi
  • 85,056