The sound energy density maintained in a room at the (nearly) discrete frequency ω produced by a musical instrument is proportional to the quantityA(ω)=D(ω)⋅δω⋅T〈∫ψ̇φdS〉av. HereD(ω) is the frequency density of room modes, the φ's are pressure wavefunctions belonging to the room,Tis the mean reverberation time for these modes, and ψ is the surface displacement of the second source. Integration is over the source surface, and the averaging is over room modes within the bandwidth δω. The latter depends on (1/T) via the Schroeder frequency, and on the “natural linewidth” of the source spectrum:Ais dependent on the linear sizeLof the source via some fraction of (ωL/c). Room selection and microphone and player location for meaningful spectrum measurement may be guided by general properties ofA, φ, andD. A violin recording and transposition to synthesize the sounds of the Hutchins family of new violins will be demonstrated as an application of these principles. Calibration procedures will also be described.