Talk:Spectral interferometry

Nice (School project)
Well written and interesting article! --Alessia Arrigoni (talk) 13:23, 18 June 2021 (UTC)

Nice!--FedericoVis (talk) 13:52, 18 June 2021 (UTC)

Hi, it is a good article! --Beatrice Lotesoriere (talk) 14:54, 18 June 2021 (UTC)


 * All three accounts above were set up on the same day that this article was created. All three accounts above have made no edits to Wikipedia, other than to say that this article is nice, and to make some cosmetic changes to this article, on the very same day that the article was created. (At first, I thought this was sock puppet spam, but no, I guess its not.) From what I can tell, all three, together with the article author, participated in GLAM/PoliMi/2021 which was a school project that resulted in the creation of this article. See next comment, which is a critique from that school project. 67.198.37.16 (talk) 22:20, 31 May 2024 (UTC)

Observations and suggestions for improvements
The following observations and suggestions for improvements were collected, following expert review of the article within the Science, Tecnology, Society and Wikipedia course at the Politecnico di Milano, in June 2021.

1. Spectral interferometry enables to retrieve the spectral characteristics of the light transient. In this sense, the definition "a linear technique used to measure optical pulses" may be too ambiguous, since it is not clear yet which domain we are talking about (space, frequency, time?). The fact that SI works in the spectral domain (and not, e.g., in the time domain) should be stated since the beginning; for the same reason, I would explicitly specify that the fields are written in the frequency domain on purpose.

2. Spectral interferometry works beyond the concept of pulses: in practice, it can be applied to any pair of fields (reference and test field), provided that they are mutually coherent (or phase locked); for example, reference and test fields can be obtained by an interferometer (eg, a Michelson), and interferometry can be used to measure the differential dispersion introduced in one of the two arms (as quickly mentioned in the application section). In this case, the two fields could also be derived from an incoherent source, for which the concept of pulse is inappropriate. I would hence clarify that the characterization of pulses is just one of the numerous applications of spectral interferometry.

3. The article introduces the concept of spectral fringes too soon, the reader may not be familiar with that. However the concept can be easily understood after the second equation of the article ("S_{SI}=...") in which it is clear that the spectrum gets modulated by the cosine term. Hence I would postpone the concept of spectral fringes after the equation. I also suggest to add a figure with a fringed spectrum. Even better would be an animated figure, in which the interferogram is plot as a function of the delay tau between the fields.

4. Concerning again equation "S_{SI}=...": it should be clarified that S is the intensity, and it is obtained from the previous equation by taking the square of the field. In addition, the mutual delay is should be tau, and not T.

5. Section "Comparison with the Time Domain": This section is very obscure. I can only guess what the student was meaning, but only because I have in mind the experimental implementation of SI. In addition I would not mention TIVI: there is no reason to mention an inconclusive technique.

I add one important concept, which is missing in the article, but which could be discussed in this section: extracting the spectral phase by taking the arccosine from the spectral fringes "S_{SI}=..." is in practice very challenging for various reasons: (i) arccosine is ambiguous by 2pi (the 2pi factor is mentioned in this section, but the reason is not discussed); (ii) static interferometry fails for spectral phases with both slow and rapid changes in frequency. Both issues can be solved by combining spectral interferometry with temporal scan. This can be directly deduced from equation "S_{SI}=..." by changing tau. This concept is very important and very powerful, and should be mentioned in the section discussing the time domain.

6. Section "Applications" I would distinguish between measuring dispersion (which does not require pulses nor coherent fields) and SPIDER, which is working only for laser pulses. I feel that the student has missed the key point of SPIDER: this is actually based on the interference of a pulse with itself, slightly shifted in frequency. There are various techniques to obtain this effect (not discussed in the article); but the schematic reported in the figure is none of them and is not correct.

7. Section "Experimental difficulties" The most relevant difficulty is maintaining phase coherence between reference and test field, which requires that both fields arise from the same source, and which invokes phase-stable setups. This could lead to a discussion on the stability of the setups and the involved interferometers, which is missing.

8. Section "Spectral shearing interferometry" This section is not explained properly. The expansion of phi is mathematically not correct

9. Section "Frequency-Resolved Optical Gating" In its general implementation, FROG is not related to spectral interferometry. This is also conveyed by the equation reported in the section: no interference is occurring there. I think there is one interferometric version of FROG, but it is not the main application, and for sure it is a variant which would require a long discussion. My suggestion would be to skip this section.

Ettmajor (talk) 10:09, 11 July 2021 (UTC)