enter image description here

This picture is from thorlabs' website.

Continues light can be modulated by an optical chopper. According to the fourier transform theroy, I think this kind of modulation will bring new frequency compoments to the continues light. However, there seems no energy exchange happen between the new and initial frequency light, I cannot write a suitable equation to describe the energy and momentum conservation process. Therefore, can modulation of continuous light by an optical chopper generate new frequency photon?

In non-linear-optics, Self-Phase Modulation(SPM) and other non-linear effects seems another kind of modulation through light-matter interactions. It is clear to me that new frequency light has been generated by SPM since the spectrum is wider after SPM in my experiment. And I can take the new frequency light as another light source by a suitable optical band-pass filter.

Are there any differences between these two ways of modulation? It seems to me that optical choppers cannot generate photons at new frequencies because the principle of chopping is just periodically blocking the light source.

  • 2
    $\begingroup$ Trust your Fourier transform. But it would be hard to detect unless with a carefully designed experiment since the difference in frequencies is about 1 part in 10 to the 12th. $\endgroup$
    – UVphoton
    Commented Oct 4, 2022 at 14:58
  • $\begingroup$ Yes. I can understand the modulation on math. Is the concept of photon the same with a single frequency sine wave? $\endgroup$
    – blocks
    Commented Oct 5, 2022 at 0:50
  • $\begingroup$ A single frequency sine wave is eternal. But no real signal can be eternal, so it can't be a single frequency, and Fourier tells you what the frequency components are. $\endgroup$
    – PM 2Ring
    Commented Apr 7 at 10:10

2 Answers 2


I think a chopper does alter monochromatic light.

A perfect light blocker is, in principle, generating an exact out-of-phase replica of the incoming light and adding it to the light, forming a summed electromagnetic field of zero amplitude.

That kind of light blocker is not nonlinear, isn't generating any second frequency, so should not add any Fourier components to the field that would have a frequency other than that of the incoming light.

The only hitch, is that the real blocker also has a moving edge, and in addition to diffraction at that edge, some part of the motion might impart a Doppler shift; so, yes, a light chopper of the ordinary sort (slotted disk) may somewhat alter the color.

More generally, a modulated light source is not a pure sinewave, but something akin to a doublet, which are spaced by the 'beat frequency' which is the modulation (actually, half the frequency of the amplitude modulation). I say 'akin to' a doublet, because such a doublet would not only give a modulated output, but would toggle the phase of the outgoing light.

So, both the modulation (with a splitting of the frequency) and the chopping are introducing a wave with opposite phase; the purity of a monochromatic light beam is going to become somewhat less in each kind of modulation, as a result of the introduction of the modulator's opacity; we cannot ignore the action of a light blocker as though it were an inert object, incapable of being an agent of frequency change.

  • $\begingroup$ Is the process of chopper modulation "linear"? I think it is a process of loading information on light. The information itself has some new frequency compoments, which migrate to the optical spectra. $\endgroup$
    – blocks
    Commented Oct 5, 2022 at 1:00
  • $\begingroup$ Complete blockage is linear; complete passage is linear. As for advancing-edge-of-blockage, and tricks like rotating polarizers, the math of making the wave wax and wane is strongly suggesting a spectrum effect. Any amplitude modulation makes sidelobes. $\endgroup$
    – Whit3rd
    Commented Oct 5, 2022 at 6:17

I think that harmonics arise with their differences. Because every square wave signal is, in Fourier terms, the sum of all odd harmonics. And since the transition from signal to non-signal is not a pure step signal, but takes an infinitesimal amount of time, some frequencies are distorted. I also have the operating curve of a transistor in mind, which is divided into 3 operating areas A, B, and C, of which only the linear B part is used for operation in order to avoid distortions (v-fold harmonics & co). You may also want to consider these 3 areas in your experiment. In addition, as my predecessor had already mentioned, your system also vibrates mechanically. As a result, the geometry (extensions in all directions x, y, z) and thus also the resulting waves, which preferably find space in your system, change.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.