# Definition of boundaries within the electromagnetic spectrum

As far as I know, there are no universally accepted boundaries within the EM spectrum. This means that different authors will provide a different frequency or wavelength to separate the different portions of the spectrum: radio waves, IR, visible light and so on. For example, the radio waves article from Wikipedia in English states that this portion corresponds to $$30 \textrm{ Hz} but the article in Spanish states $$10 \textrm{ kHz}.

Does the portion of the spectrum that a given signal belongs to, depend on anything else than the frequency content of the signal? I am wondering this because it looks like it would be easy to set universal boundaries if it was only a matter of agreement in a couple of numbers.

• It is a matter of "naming" frequencies, why would you want strict boundaries? Even for visible light, the "visible" depends on biologic response of individuals. Jul 18, 2021 at 4:19
• @annav I don't want strict boundaries. I was just wondering why they do not exist. As you point out with the visible example, these boundaries are not defined by the frequencies themselves but by the physical characteristics of the waves of each portion of the spectrum.
– Javi
Jul 18, 2021 at 20:21

The boundaries are all "fuzzy" or approximate because the characteristics of the waves continuously and smoothly change as their wavelength changes. Note that there is no practical detriment related to the fuzziness of those boundaries.

The "visible" light band has well-defined boundaries for humans because of the biochemistry of our retina. In the case of radio waves, for example, the practical or engineering upper frequency cutoff happens where we run out of technology for producing radio waves with shorter wavelengths, and the lower frequency cutoff occurs where the antenna size necessary to radiate well becomes unmanageably and impractically large.

• I recognize there is a continuous change of the characteristics in the waves as wavelength changes. I believe you are implying that the definition of each portion of the spectrum relies more on these characteristics than the specific wavelengths composing the signal, am I right?
– Javi
Jul 18, 2021 at 20:09

When I was a practising electromagnetics test engineer a few decades ago, the various design specifications and test standards differed. Audio was typically treated as 50 Hz to anywhere in the range 80-200 kHz, with RF up to around 1 GHz and microwave above that.

These definitions were geared around the physical effects of concern and the test equipment employed. Different equipment standards led to different definitions of the changeover points.

The radio communications bands - LF, MF, HF, SHF, UHF and the like tend to be fairly stable, as they are agreed by international communications standards bodies and enshrined in many national laws and practices.

I would also think that things which depend on natural phenomena, like the visible spectrum or the threshold of ionising radiation (somewhere in the ultraviolet range) must be pretty well tied down by the scientists.

• From your answer, I understand that your opinion is that any given EM signal is classified to a different portion of the spectrum depending on its "physical effects of concern and test equipment employed", but not necessarily according to the exact frequencies composing the signal. Do you agree?
– Javi
Jul 18, 2021 at 20:13
• @Javi Broadly, yes. Obviously the definitions of the various bands are not too different. One standard might classify say a 120 Hz susceptibility test as audio and require only bulk current injection, while another would classify it as RF and require only E- and H-field antenna irradiation. Some might require both in different tests and be inconsistent as to how they describe it. That is not opinion, it is (or used to be in my day) enshrined in the various standards. We didn't worry about it, we just did the tests and troubleshot the resonance! Jul 18, 2021 at 21:27