Timbre in harmonium plays a major role in the quality of the instrument. The European free reeds were developed early on with a timbral conception that was oriented towards the human voice. The timbral kinship of harmonium and human voice is nowhere as obvious as in Indian music. If early harmoniums were often regarded as having a certain coarseness and hoarseness, all musicians agree without exception that the modern harmonium comes very close to the vocal sound.
Those singers using the harmonium also lauded its ability to adapt its timbre through the basic choice of reeds and the combination of stops. The combination of stops does not refer to the changing of stops during performance, for within the performance of a single piece a change of couplers is extremely rare; what is rather meant is that the sound is adapted to fit the range of the voice, which is done stereotypically: the three sets of reeds a harmonium usually has are called kharaj ("bass"), nar ("masculine") and madi ("feminine"); which refers primarily to the pitch and therefore to the rank of the respective set. Accordingly, a sonorous male voice is accompanied with kharaj, while a softer voice is accompanied with the nar-stop and, if necessary, with a coupled kharaj- or madi-stop. A female voice is usually accompanied with the combination nar- and madi-stops. Just because every singer has his own Sa (i.e. his own voice determines the modal system`s tonic, usually between C and E-flat`) the harmonium is tuned to the according pitch. With a sarangi - or any other stringed instrument - a scordatura would always mean a change in timbre, even if slight.
In Indian music the ideal singing voice has a natural vocal sound; by adapting the instrument to the current register of the voice this natural quality can be enhanced, because the singer does not have to go to vocal extremes to adapt to his accompaniment. This affects the vocal timbre, because the singer does not need to press his voice, for instance, to sing extremely high notes.
Because every singer usually owns a harmonium, or sings regularly with an accompanist who owns a harmonium adapted to his voice, a high level of blending is possible between instrumental and vocal sound. If the sarangi is said to come closest to imitating the human voice, then this rather vague statement can only refer to an area other than timbre; still, when asked, some of the musicians stated that the timbre is meant, without giving any details or reasons.
The harmonium`s superior ability to create a homogeneously blended sound together with the human voice, therefore, is preferred to the sarangi`s superior ability to imitate the more subtle vocal affects. Musically these are two entirely independent areas; accordingly the choice does not mean that one of these areas is rejected. A soloist has to make a choice according to his personal preferences, and this decision is complicated by many other considerations concerning the instrument (e.g. intonation).
