Low-frequency p-modes have relatively small amplitudes and are set against the background of granulation on the solar surface. In this work, the calibrated signals of both the MDI and GOLF instruments are processed in the following steps: 1) prewhitening 2) signal extraction and demodulation in given frequency bands and 3) separation of noise and signal by singular spectrum analysis. The resulting signals are compared in both time and frequency domains and mode parameters are determined by fitting autoregressive models and Lorentzian spectral profiles. Particular attention is paid to spectral features at very low frequency among the investigated data sets, which can lead to the identification of low radial order p-modes.