Channel impulse response (CIR) datasets, which represent the characteristics of wireless channels, play a vital role in industrial applications. Unfortunately, systematic errors have been discovered within CIR datasets. These errors have undesirable effects in wireless localization, physical layer security, and related applications. In addition, existing CIR calibration methods do not consider applicability to industrial scenarios. To address this concern, two channel sounder error calibration (CSEC) methods are proposed. The first CSEC method based on phase compensation (CSEC-Phase) extracts the direct path signal to estimate errors to avoid interference from rich multipath signals for stationary measurements in complex industrial environments. Moreover, to solve the problem of the inability to estimate the CIR error in mobile measurements due to the rapid dynamic change of the channel, a second CSEC method based on carrier frequency recovery (CSEC-Frequency) is proposed that extracts the frequency offset of the system clock to compensate for systematic error utilizing the stability of clock drift. Furthermore, the L1-norm distance metric is employed to measure the differences between CIRs and evaluate the calibration performance. Our results reveal that the proposed CSEC methods are effective on diverse CIR datasets measured by different11The MATLAB code of the CSEC methods has been shared on GitHub: https://github.com/QiWang-GH/Industrial-Wireless-Channel-Sounding-Error-Calibration-IW-CSEC. equipment in different scenarios.