Of course, my solution is not the most elegant, but I think it is quite reasonable.
We define the independence of the integral from

, for which we use the fact that the function of the form
}{x})
up to

constant is the Fourier transform of unit step height in the interval
![[-a,\,a]](http://latex.codecogs.com/png.latex?[-a,\,a])
. Consider this integral as the value at

inverse Fourier transform of the product of the functions
}{x^{n + 1}}\,\prod\limits_{k = 1}^n{\sin(kx)})
, but the contraction of the first n steps will be of compact support on the interval
![\left[-\frac{n(n +1)}{2},\,\frac{n(n +1)}{2}\right]](http://latex.codecogs.com/png.latex?\left[-\frac{n(n +1)}{2},\,\frac{n(n +1)}{2}\right])
. Therefore, if
}{2})
, when calculating the convolution of this function with the last step, the width of

at point

, we simply calculate the area under the graph of the convolution of the first

steps, but this area does not depend on
}{2})
.
We divide this integral
=I_1+I_2)
into two intervals
![[0,\,a] \wedge [a,\,\infty)](http://latex.codecogs.com/png.latex?[0,\,a] \wedge [a,\,\infty))
. Parameter
)
we will suggest later, based on the conditions:
\to 0\,\text{ if }\,t\to\infty)
. Now perform the substitution

and estimate this integral on the interval
![[0,\,a]](http://latex.codecogs.com/png.latex?[0,\,a])
:
Now we impose a condition on the parameter: a

.
Now we estimate the integral on the interval
)
:
From this representation

and given the inequality

, we obtain an estimate

. And now we impose the condition

. Let
 = t^{-\alpha},\,\alpha > 0)
. Then the condition

that means

, but the condition

means

. Thus, when

, all conditions are met.
So, on the basis of
)
, it follows that
 = \frac{\pi n!}{2}.)