An analog scope only presents a comb of frequencies determined by the timebase settings. A digital storage scope acquires a BW determined by the sampling rate. In typical usage a DSO displays only the same narrow range of frequencies that an analog scope displays. It is only in single shot mode that you see the full bandwidth acquired by the DSO. The rest of the time the display is dominated by the periodic component.
What a DSO displays in "normal" mode is a set of boxcars in time specified by the sweep rate, display width and trigger rate. This is a set of sinc functions in frequency. This is basic Wiener-Shannon-Nyquist mathematics. Relative to the Fourier spectrum set by the sample rate, that signal may be "sparse" within the requirements of compressive sensing. However, it may not. But overwhelmingly the odds are that it is. Donoho and Candes have presented rigorous proofs. Should you desire rigorous proofs, you must read their papers. But be warned, after spending 15 pages proving a single theorem, Donoho remarks that doubtless the reader will be relieved to know that the proofs of theorems 2 and 3 are much shorter. In fact both were 2-3 sentences.
In the figures I posted from F&R, there are 64 frequencies. However, only 5 frequencies have non-zero coefficients. There is a very sharp threshold between sparse and solvable and not sparse and not solvable. In the latter case one must revert to Wiener-Shannon-Nyquist.
If you know nothing about the signal you can only acquire it with a single shot sweep of sufficient duration to capture the entire signal. DSOs have made things much easier, but consider an analog storage scope. Do you really expect to see a 1 nS pulse on a 1 mS duration sweep? That pulse is one millionth of the sweep length. A modern low end DSO will acquire 10 million samples at 1 GS/S. But the display is typically less than 1000 samples. Zoom mode might let you find that in a one shot, but the only way you will see it in normal mode is if it is repetitive.
In practice, the signal of interest usually dominates the spectrum and is sparse. The coefficients of the transform at the other frequencies are noise and very low amplitude. This is the basis of all the lossy compression algorithms such as JPEG, MP3, etc. Compressive sensing simply merges the sampling and the compression into a single operation.
DSOs display thousands of waveforms per second. Is your eye able to discern that? No. Even if one uses a probability density display a single spike will be undetectable.
I am not in the same league with Donoho and Candes. So I cannot explain this as well as they can. Both write well and provide all the rigor you can stand. Unless you *really* care about the fine print, I suggest reading the introductions and skipping the proofs.
http://statweb.stanford.edu/~donoho/reports.htmlhttps://statweb.stanford.edu/~candes/publications.htmlThere are references on their home pages one level up about lay press discussion of the subject.
Yes, we were taught that all this is wrong. Which is why when I ran into it by accidentally doing it I *had* to know why. It cost me a few thousand hours of effort over most of 5-6 years. If after you have read their papers you still think it's wrong, take it up with them. I've cited numerous peer reviewed papers.
To repeat an earlier comment. The gist of the matter is solving Ax=y using an L1 criterion where y is a randomly sampled series and x is the positive frequency Fourier transform coefficients and then back transforming using the inverse FFT. A 5 year old probably would not understand that, but it's as simple as I know how to make it. A search on "compressive sensing" will turn up a very large number of examples in the form of graphs and images. The big breakthrough was showing the L1 has *very* different properties from L2.
If you're a computer geek and know what NP-hard is, then I strongly recommend the 2004 papers by Donoho on the equivalence of L0 and L1 solutions. To the best of my knowledge, that is a major milestone. I presume that the computational complexity crowd has been working feverishly on this, but I've not looked into the matter. So far as I know it is the first and only instance where large NP-hard problems have been solved.