In this paper, it is proposed that the spectral shape of combustion noise is the same as the similarity spectrum of the noise from the large turbulence structures of high-speed jets. This is true regardless of the turbulence level in the combustion process, the equivalence ratio of the fuel or whether the fuel is a gas or a liquid as long as it is hydrocarbon. At the present time, the mechanism by which combustion noise is generated is still not known. So, it is not possible to prove the correctness of the proposal theoretically or computationally. Here, an empirical approach is followed. Comparisons between the similarity spectrum and many sets of combustion noise spectra are made. The spectra are from open flames, low speed jets, can-type combustor, auxiliary power units and turbofan engines noise measurements. Good agreements are found offering favorable support for the proposal. This paper is written in honor of Dr. Marvin E. Goldstein, a world-class aeroacoustician, an accomplished applied mathematician and a much-respected long time friend.