Electrocardiogram (ECG) was recorded electrical signals produced by the body from the heart using electrodes placed on the surface of the body. Electrocardiograph is an electrical activity that are important in diagnosing cardiac disorders such as myocardial infarction, conduction defects, and arrhythmia. But in fact, the ECG signal is often contaminated by a noise. Therefore it is necessary an ECG signal processing is capable of addressing the issue. Where to ECG feature extraction and detection of QRS complex needs a technique to overcome the baseline wandering and minimize disturbance in the ECG signal. In this paper noise reduction techniques that will be used is based on wavelet Daubechies 04 (db 04) with decomposition level 3. The algorithm was developed using Advanced Signal Processing Tool (ASPT) LabVIEW programming. In this study, of the five types tested wavelet (Haar, Daubechies 02, Daubechies 03, Daubechies Daubechies 04 and 06), the type of wavelet that delivers the best RMSE values are Daubechies 04. Because it can do a enhancement value is good for AWGN 0.1, 0.2, 0.3, and 0.4 with the percentage
value of 2.69%, 4.97% ,7.97%, and 11.31 %.