[sDIY](early) wavetable synthesis

[mbartkow at ET.PUT.Poznan.PL]

Paul,

> >If you reproduce the wave on the basis of some other sampled wave, you
> >are resampling a signal which already has been sampled, so its spectrum
> >is not limited. Lots of aliasing typically occur in such cases (the cha-
> >racteristical lo-fi sound of the Amiga/tracker music). Pro samplers
> >usually apply interpolation to the waveform stored in the memory before
> >the resampling process. Interpolation may be thought of as lowpass fil-
> >tering whose aim is to limit the spectrum (to have more smooth waveform.)
> >
> 
> Hmmm, My PPG VCO sounds fine, the only thing I have vagually described as a 
> filter is a uA741 as the output opamp, it filters enough of the signal for 
> me.. mine sounds fine.. :-)

Well, as far as I understand your circuit in the Modulus No. 6, you are not
skipping samples in the waveform. If there are 64 samples per wave cycle,
there is enough bandwidth for spectrally rich waveforms without aliasing.
There is virtually no aliasing if the waveforms contain up to 32 harmonic
partials (more precisely speaking, the energy of the spectrum part above  
32nd harmonic is kept below 48dB, which is the quantization floor level,  
due to 8-bit representation).

> anyway, 99% of the time, the output of most VCO's goes through a filter...

Mind you, it is not possible to remove aliasing by putting a filter after
sampling. You can only remove the aliased portion of the spectrum together with
the original part of the one.

> 
> >
> >Well, AFAIK it IS used. You must not skip samples from the wavetable, if
> >the wave is not appropriately smoothed (i.e. its bandwidth is appropriately
> >limited), otherwise heavy aliasing occurs.
> 
> yes, haveing spoken to wolfram franke and looked extensivly at the PPG  
> stuff, as far as I can tell wavetable synths play ALL the samples in a wave 
> (in the case of the wave 256 samples per cycle) at a higher frequency,  
> simply skipping bits is just not on, and only make for a far far more 
> complex circuit to do this than just a plain high speed osc, or even plain 
> phase-accumulator oscilator.

A simple phase-acc oscilator may be used for skipping samples. The number
of the samples skipped depends on the increment of the accumulator.
In fact, this phase-acc might be used for dual purpose of DCO and sample  
counter. I have never seen the PPG schematics, but I can imagine a 24-bit
phase-acc driven from 1.6MHz crystal osc. This gives you 0.1Hz pitch reso-
lution and you can use the highest 8 bits to address the samples of one cycle
of the waveform. Commercially available direct frequency synthesis circuits 
work in this way.

> >I understand it might be confusing without a deep knowledge on the
> >Fourier transform properties. Try the excellent Bristow-Johnson tutor
> >on wavetable synthesis from the music-dsp site.
> >
> 
> Whats the URL for this music-dsp site?

Unfortunatelly, it seems that the paper has been removed from their site.
The URL is http://shoko.calarts.edu/~glmrboy/musicdsp/music-dsp.html 

I have a printed version of the paper, which has been published at 
AES 101st Convention, Los Angeles Convention Center, Los Angeles, CA, USA, 
1996 November 8–11. Robert Bristow-Johnson, Wavetable Synthesis 101, A  
Fundamental Perspective.
Some time ago it was available in HTML version. Maybe you will be able
to locate it.

I recommend also:
http://www.digital-daydreams.com/library/programming/samplefx.htm


best regards,

mb


--

Maciej Bartkowiak, PhD
========================================================================
Institute of Electronics and Telecommunication     fax: (+48 61) 8782572
Poznan University of Technology          phone: (+48 61) 8791016 int.171
Piotrowo 3A                             email: mbartkow at et.put.poznan.pl
60-965 Poznan POLAND               http://www.et.put.poznan.pl/~mbartkow
========================================================================