Digital images are used to become a visual bridge of information. The information data must be precise so that the information can be adequately conveyed, but in the process, digital images sometimes experience a change in quality. One of the causes of this change is noise, where the image affected by noise is of poor quality, so misinformation can occur. This problem can be solved using filtering methods, but there are so many filtering methods. In this study, five filtering methods were used, including the Gaussian filter, mean filter, median filter, wiener filter, and conservative filter, to be compared with two types of noise, such as salt and pepper and speckle, so that the best method for noise reduction in digital images is known based on the criteria that have been set determined. The research results were determined based on the value of the measurement parameters Mean Square Error (MSE) and Peak Signal-to-Noise Ratio (PSNR). The results show that the conservative method is the best based on the parameter values of MSE 3.21 and PSNR = 37.99. However, when viewed visually, the median method is superior for reducing noise in digital images that have been carried out. The results of the research can be used as information to develop future research, especially in the field of digital image processing.