Maintaining good posture when using a laptop or a smartphone can prevent computer vision syndrome and text neck syndrome. However, it is difficult to remain aware of posture during an activity. Thus, wearable systems with posture feedback can help maintain good posture during daily activities. In this paper, we propose a posture feedback system that uses a commercial wearable speaker, which has been used for music and video conferencing when working from home. To judge a user's posture as good or poor, we focus on estimating the distance between the user's eyes and the screen when using a laptop and the neck tilt when using a smartphone. To estimate the distance, we use an active sensing method with ultrasound sent from a wearable speaker to a microphone on a laptop or a smartphone. The sound pressure of ultrasound changes depending on the distance between the wearable speaker and the microphone. In addition, an active sensing method can be used to estimate neck tilt because the sound pressure changes depending on the angle between the wearable speaker and the microphone. When the system judges the user's posture as poor, it will provide auditory feedback by applying digital audio effects to audible sounds (i.e., audio being listened to). Audio signal processing is implemented as a web application so users can use our system easily and immediately. We conducted three experiments to verify the feasibility of our proposed method. The sound pressure changed depending on the distance and angle between a wearable speaker and a microphone, and the system could judge posture as good or poor at almost 100 % under the experimental conditions.