Cheap scope expectation management.
I have an old 15MHz analog scope. It is honestly 15MHz. If you put in a higher-bandwidth signal, you see a 15MHz low-pass filtered version of it. For example, a 6MHz square wave will look pretty rounded, but rock steady with no strange artifacts.
This thing claims to be 50MHz bandwidth, which may well be true, but I don't think there's anything even resembling a 50MHz "brick wall" low-pass filter. So if you sample a higher-bandwidth signal you can get some gruesome artifacts. I don't have access to high-quality lab equiment, specifically no function generator and no better quality scope to compare to. What I do have is a 3.6864MHz clock oscillator module which, based on lab experience in a past life, puts out a clean signal - it had better, being a clock source. Let's put it on (first pic).
Whoa! I could handle "ringing" on the edge, but every edge looks different! 10s persistence shows the variation.
Well, that's overmagnification. In high speed mode, this thing does something like 250Ms/s but at 50ns/div that's only about one sample every other pixel. So zoom out to 100ns/div (second picture). This shows several different "versions" of the rising and falling edges, now at roughly 1:1 ratio between horizontal pixels and samples.
Right, get a hold of yourself. This is not a $20K lab scope. This is a scopemeter. What do people use them for? Debug power supply issues, audio issues, and, in the old days, standard definition TV issues (that's the world my 15MHz analog scope was born into). I do still have a single SD video source in this house - an old video camera.
Third picture: Pretty! That's the 3.58MHz colour burst on the "front porch" of a scan line. Zoomed in (4th pic) it still looks pretty good, though not particularly sinusoidal. So back to reality. This thing totally does what hobbyist scopes need to do. You could debug a switching power supply with it, for example.
Just for laughs, back to the square wave at the same horizontal resolution (250ns/div) as the last capture. You can see the artifacts, but all in all it looks square wave-ish. Note the probe was compensation adjusted at 1KHz so it may not be perfect at this zoom level.
It's not the probes by the way. I tried a high quality, Tektronix 250MHz scope probe that I happened to have, and the waveforms look essentially the same.