Everytime I use the audiowrite and audioread function, it adds some noise to my original signal. Does this happen just to me?

Frequency response plot of the signal before audio writing:

Frequency response plot of the signal after audio writing:

PS : The frequencies peaks of the signal are 20, 20.5 and 21KHz respectively.

[Walter Roberson]

What is the range of values and class() of the data that you are writing out ? What is the range of values and class() of what gets read back in ?

[Visweshwar Srinivasan]

Thanks for the reply, the class of the variables used is double, both before writing and after reading. The range of this signal is 0 and 2.1369, which means the modulus(maximum) of this signal is greater than 1, so is this because the signal is getting clipped due to this fact? Also, I have plotted frequencies from 0 to fs, where fs is the sampling frequency (44.1KHz). The frequency peaks are 20, 20.5 and 21 KHz, and the other part is due to complex conjugate symmetry of the signal.

[Visweshwar Srinivasan]

Thanks a lot, sir! The problem got solved now.

Answer 1

[Walter Roberson]

The data is getting clipped on output because audiowrite() clips double values that are outside the range -1 to +1