Peak optimization failing if manual peak picking

[lpatiny]

When peaks and selected manually the optimization fails.

[jobo322]

it is pretty strange, the reason of the wrong optimization is at the moment of convert the width to PPM the fwhm looks to be divided twice by frequency it only happens in the manual peak detected

[jobo322]

here is possibly the source of this:

• https://github.com/cheminfo/nmrium/blob/main/src/component/reducer/actions/PeaksActions.ts#L99-L106