Fourier Transform negative discrepancy

55 Views Asked by At

My teacher wrote on the board for Fourier Transforms for $f(x)$ from $-\infty<x<\infty$:

$$F(\mu)=\int_{-\infty}^{\infty}f(x)e^{i\mu x}dx$$ and for the inverse transform $$f(x)=\int_{-\infty}^{\infty}F(\mu)e^{-i\mu x}\frac{d\mu}{2\pi}$$

Notice where I have the negatives in the exponents... My confusion is that I see the following all over online:

$$F(\mu)=\int_{-\infty}^{\infty}f(x)e^{-i\mu x}dx$$ and for the inverse transform $$f(x)=\int_{-\infty}^{\infty}F(\mu)e^{i\mu x}\frac{d\mu}{2\pi}$$ with the non-negative in the inverse transform. But I can't find the version that my teacher wrote.

1) I am trying to figure out if they are equivalent somehow or if there was a mistake in class.

2) Also, if there was not a mistake, would both forms give the same answer in the end of problems? Or is there some kind of change of variable that happens?

Thank you.

1

There are 1 best solutions below

0
On

The two forms are equivalent.

But the second form is indeed the preferred one at 99.9 %. I see at least two reasons for that :

  • It can be seen as the hermitian (dot) product (usual for complex-valued functions) :

$$\langle f,g \rangle=\int_{-\infty}^{\infty}f(x)\overline{g(x)}dx$$ where $g(x):=e^{i\mu x}$.