[A6] Some waveform observations II.

[Robert Skerjanc]

Hi all,

Actually these asymetrical waveforms sound more interesting than 
mathematical correct ones. But to enable "correct" FM I found following 
trick:

To the modulator's sine, add the rectangular waveform with a pulse width 
of 100% and an amount of about 2.0 . This compensates the DC offset 
caused by the asymetrical waveform, where the mean value is not equal 
zero. Now modulate linearily by an EG, and bingo, there is a 
Chowing-like sound in tune. Next step is to modulate VCFs with the same 
method, to have 4OP-FM ...
 
Polyphonic mode is possible, but the rectangle amount is a very sensible 
value to get all voices in acceptable tune.
According to that I would say the detune FM artifacts between voices are 
due to small differences in waveform assymetry. They cause different DC 
offsets and different tunes. The reason are not some VCAs which can't be 
calibrated in the tune process.

Robert


David Evans schrieb:

>On Sun, Nov 21, 2004 at 11:18:07AM +0100, Robert Skerjanc wrote:
>  
>
>>1. the basic waveforms of the VCOs look very strange, when you see them 
>>on an oscilloscope.
>>    
>>
>
>  Many synths have crazy-looking waveforms due to design of the VCOs and
>because of other parts of the signal path.  It's part of the source of
>fun.
>
>  
>
>>2. The VCOs have linear and expon. type of FM. Normally the linear type 
>>is intended for the modulation to stay in tune (negtive slopes detune 
>>the carrier in same amount as positive slopes), like the DX7. However 
>>linear FM on A6 stays not in tune. Possibly the reason are the weird 
>>waveforms mentioned above?
>>
>>    
>>
>
>  Certainly if they're asymmetrical about the zero axis then that could
>happen.  I'm afraid I'm not sufficiently awake to say any more.  ;-)
>
>  
>