Use the gradient function to produce results that are the same sizes as your data, not diff.
If ‘EMG’ is your (30000x4) signal matrix, the derivative (‘dEMG’) is:
[~,dEMG] = gradient(EMG);
Note that taking the derivative of a a ‘real world’ signal will significantly amplify the noise. It would be best for you to use a lowpass filter on your EMG signal matrix first, to eliminate high-frequency noise. If you have R2018a or later, use the lowpass function (or bandpass if you also want to eliminate d-c offset and low-frequency baseline drift). Otherwise, efficient filters are not difficult to design and implement in MATLAB.