I'm confused about the unit of x-axis.
The below signal has a duration of 3 second and sampling frequency 44100. So should i write Time(s) for the unit of x-axis.
However, in the graph, it shows the number of samples i.e 44100 * 3 = 132300.
I'm confused about the unit of x-axis.
The below signal has a duration of 3 second and sampling frequency 44100. So should i write Time(s) for the unit of x-axis.
However, in the graph, it shows the number of samples i.e 44100 * 3 = 132300.
The below signal has a duration of 3 second and sampling frequency 44100. So should i write Time(s) for the unit of x-axis.
No. In your graph, the X axis seems to be simply the sample number. I mean, it's a bit strange to ask us, considering you generated the graph, but if you benefit from having it spelled out:
In DSP, signals are represented by a sequence of samples. And if you plot these samples, there's no inherent "units" on the X axis. Your signal has been reduced to a sequence of digital numbers, and the fact that there was a sampling rate that implies some period of time between each sample is simply not there in DSP.
It is always a good practice to have proper axis labels, on both the $x$ and $y$-axis. And a legend or a title. Digital signal processing usually deals with regularly sampled data, in which the actual sampling frequency "almost" vanishes: samples are separated by a sort of unitless length $1$ interval. So, many low-level processing use a "sample index" $x$-axis. I prefer sample index to sample number.
However, when it comes to event location, comparison of frames of different length, data with different sampling rates, a label with a multiple of seconds can be preferred.