Peaks between harmonics in a signal

I am hoping someone is used to signal processing to explain why I have a problem. I am trying to reconstruct a signal using a fourier synthesis but I have noticed that my original signal has peaks between the harmonics having compketed a fft I was wondering why this would be but also how would I deal with his when reconstructing my signal. Would I include them in my synthesis or disregard them and only use the harmonics

[Star Strider]

A bit more detail would help.

What did you do?

How did you deconstruct the signal you now want to reconstruct?

Did you do anything to it (like filter it) in the interim?

How are you reconstructing the signal? Are you using the ifft function or your own code?

An image of the original signal and your reconstructed signal would help also.

Answer 1

[Image Analyst]

Why do you find it confusing that a waveform would have frequencies other than integer multiples of its harmonic frequency (if there even is one that's dominant, which there may not be)? A signal can be comprised of frequencies of any value.

If you see spikes in your Fourier transform then it means that there is a lot of energy in that/those particular frequencies. But of course there can be energy at other frequencies too. In general if you try to reproduce a signal with only the "main" frequencies, it will be close to the original but not quite exact. The more "other" frequencies you include, the better the summation will approach your original signal.