Talk:Signal frequency spectrum

Latest comment: 10 years ago by Dicklyon in topic What is this article supposed to be about?

Wrong example?

edit

The example shows a sine wave and then shows a spectrum with three frequency components?

Also, we should do the continuous case first, either with a continuous fourier transform or a fourier series of a simple three-component signal or something. — Omegatron 00:58, 5 December 2005 (UTC)Reply

The "color"

edit

I changed the effect is the "color" white to the color of the light is white. Just making a note of this here because I am a n00b to fourier stuff and may have missed some subtlety. The only way I could understand the quote marks, though, was as meaning "metaphorically but not actually white". Surely we're talking (for the moment) about actual light which is actually white so the quote marks just introduce confusion. 213.122.22.138 (talk) 14:42, 20 March 2010 (UTC)Reply

What is this article supposed to be about?

edit

And why does it have no useful sources? The term "frequency spectrum" is not very defined, either here or in reliable sources. We have various other articles on Frequency domain, Spectral estimation, Spectrum analyzer, Spectrogram, etc. Is there something that this article should be about, other than those? Or should we just delete it? Dicklyon (talk) 02:28, 13 October 2014 (UTC)Reply

It's been tagged for merging with Spectral density. The problem with frequency spectrum and power spectrum (and phase spectrum) is that each focuses only on either abscissa or ordinate axes. Ideally these dimensions would be discussed together. One clear demarcation point though, is that Spectral density estimation is outside the scope. More practically, the current version's first three sections (Light, Sound, Radio) can easily be dumped into Spectral density#Applications; the only section left, Spectrum analysis, could go into Spectral estimation. Fgnievinski (talk) 04:11, 13 October 2014 (UTC)Reply
Ah, yes, I see you tagged it. But you didn't start a merge discussion to give the rationale; neither did anyone else. I'd say either start the discussion, or if you think lack of opposition is enough, just go ahead and merge. Dicklyon (talk) 06:40, 13 October 2014 (UTC)Reply
OK, I'll give it a couple of days then I just WP:BEBOLD. Fgnievinski (talk) 13:59, 13 October 2014 (UTC)Reply
Different fields have different names for the same or similar concepts. In the audio world, "Frequency spectrum" is a more common term than "power spectrum" or "spectral density"- Gscholar gives
  • 40K hits for audio "frequency spectrum"
  • 26.7K hits for audio "power spectrum"
  • 30.9K hits for audio "spectral density"
And although it is harder to measure, I would expect "frequency spectrum" to be even more common in the popular press. Broadcasting also uses the term "frequency spectrum", but for a different purpose.
This article is mostly about the audio version, with unsourced light and radio sections thrown in (those two could be deleted, or merged somewhere else). It could be renamed to Frequency spectrum (acoustics) to indicate this article is mostly about the audio concept. As an article, start class is about right--it provides a nice gentle introduction to the concept and some nice examples, but has inadequate sourcing and citations. It could be expanded with better references and some more examples and technical material later in the article.
I can understand Fgnievinski wanting to merge material from this article into Spectral density; that article is completely inadequate as an introduction to the concept. A reader who encounters the term "frequency spectrum" for the first time, maybe in a music class or playing around with a graphic equalizer, and wants to know what it is all about, will hit a brick wall when confronted in the first paragraph in the lead with concepts like measure theory, cumulative distribution function and absolute continuity. I added some example images there to make it more accessible, but it was lipstick on a pig. The Spectral density article needs restructuring per WP:EXPLAINLEAD and WP:UPFRONT, with introductory material and examples/applications near the top and the heavy math sent to the latter part of the article. --Mark viking (talk) 18:33, 13 October 2014 (UTC)Reply
Most of the gScholar hits for "frequency spectrum" are portions of a longer meaningful phrase; see here. Dicklyon (talk) 02:40, 14 October 2014 (UTC)Reply
I wouldn't object with Introduction to spectral density (e.g., Introduction to quantum mechanics). But we can't have two names for the name thing. Fgnievinski (talk) 18:37, 13 October 2014 (UTC)Reply
Moving the article to Introduction to spectral density would be fine with me. --Mark viking (talk) 19:31, 13 October 2014 (UTC)Reply
Shall we outline what sections ideally this introductory article would have? I'm not sure how far back into the Fourier transform theory we'd like to explain. On the other hand, although a more practical, less abstract introduction would be desirable, we don't want just an Introduction to spectral density estimation. Again, I think that any of "spectral density", "Fourier space", or "Fourier spectrum" would convey the idea that we're not talking only about the independent variable (frequency domain). Fgnievinski (talk) 01:42, 14 October 2014 (UTC)Reply