Communication plays an important role in society. This can be seen in the amount of information that is spread, such as information in the form of sound, images, and videos. But communicating is not always easy. Due to the large number of sounds emitted from different sources, other sounds will be disturbed. The interference caused by noise can distort the information signal, leading a sine wave to integrate a minor noise signal. As a result, the receiver cannot differentiate the actual information signal and the added noise. This study proposed a noise-reduction system by analyzing the effect of the number of filter orders on the noise canceler system. The information signal will be processed using the STM32F4 noise canceller system, which will then be filtered using an adaptive filter with a Finite Impulse Response (FIR) structure and the Least Mean Square (LMS) algorithm. The test results show that the best SNR value is obtained at Order 40 of 5.3671 dB at a sound duration of 14 seconds, while the best PSNR value is obtained at Order 40 of 21.3557 at a sound duration of 9 seconds, and the higher the filter order value, the smaller the MSE value.