Electronic pianos and inharmonicity

[Clark]

Hi there,

The question is interesting, though. If the looping point is selected to
minimize clicking based on the lower partials, the periodicity of the upper
partials is compromised.

Manually adjusting samples say in Sound Sculptor or CoolEdit, this is not very
clear - one sees a waveform primarily as a periodic graph with interference (a
sort of fuzz); if one goes through the trouble of a Fast Fourier Transform and
select groups of close partials to loop, then compare the looping points the
looping points may not correspond.

This could very well be one of the biggest points for criticism for one of the
instruments in question: lower partials can be smooth but higher partials are
not continuous, rather forced to varying degrees into (in)harmonic
relationships with the fundamental. Indeed, inharmonicity may misregister in a
real-time measuring device like RCT or SAT.


Additive synthesis used to be measured in terms of $1,000,000 per second, since
it took a number of workstations working in parallel to synthesize an
approximate of a natural sound from component sine-waves. This technology is
now available for free through CSound which will run on most consumer
computers; sampling, on the other hand, was available (albeit at a lesser rate
and depth) to Sinclair owners. Frequency modulation was even simpler. Now both
bit depth and sampling rates are increasing, with 30-bit and 96 kHz machines
available - adjustments may be improved, but if the trend of looping samples
continues the periodicity errors will continue as well.

Now, speakers and amplifiers will add their own inharmonicity and influence to
the final sound of any electronic source, and many of the theoretical problems
may be smoothed over but it is simply a fact of the industry that old
preconceptions will be maintained until another standard is universally
adopted.

Clark