<p>Environmental conditions such as temperature and humidity significantly influence the acoustic response of bowed-string instruments by affecting string tension, damping, and resonance properties. This paper proposes a feature-based feedback framework that regulates environmental conditions to achieve desired timbral characteristics of string instruments. A sinusoidal-plus-noise model is employed to represent short instrument sound signals, and model parameters—including harmonic amplitudes, decay constants, damping characteristics, and noise intensity—are estimated using time–frequency analysis. Perceptually relevant audio features, such as spectral centroid, roll-off frequency, root mean square (RMS) energy, harmonic-to-noise ratio (HNR), and fundamental frequency drift, are extracted to construct a composite feature index that characterizes the sound signal. Based on experimental data collected under varying temperature and humidity conditions, a sensitivity matrix is identified to model the relationship between environmental variations and acoustic feature changes. This enables the design of an iterative feedback controller that adaptively adjusts environmental variables to drive the composite feature index toward a predefined target value. Experimental results demonstrate that the proposed framework effectively regulates environmental conditions and consistently achieves the desired acoustic feature characteristics from various initial states.</p>