[sdiy] Digital noise sources

[Scott Gravenhorst]

Tom Wiltshire <tom at electricdruid.net> wrote:
>
>On 13 Jun 2008, at 19:00, Scott Gravenhorst wrote:
>
>> xyzzy at sysabend.org wrote:
>>> Something else I've been pondering.
>>>
>>> I noticed in Tom Wiltshire's noise generator PIC that he preloads  
>>> the shift
>>> registers with some random data.  I'm curious about something  
>>> though...  In
>>> a "normal" logic based noise source, are we depending on the power- 
>>> on state
>>> of the shift register to be random, or are we just loading it from  
>>> 0 with
>>> the pattern generated from the XOR's?
>>
>> I've done it by seeding with some value not equal to all ones and  
>> also with a seed of
>> zero.  I use LFSRs for audio noise, not pseudorandom sequences at  
>> low rates of speed.
>> For my use, it doesn't seem to matter, it sounds like noise.  There  
>> is only one "evil"
>> value in a max length LFSR, so avoid that and you're good - every  
>> other value will come
>> up eventually anyway, so no matter what your seed is, it's always  
>> just one number in a
>> sequence of many.
>
>One related idea I've seen recently is a feature on the noise source  
>on the Nord Lead synth.
>This makes deliberate use of the fact that LFSRs aren't random at  
>all, by allowing you to reset the noise source back to some start  
>value. By feeding an oscillator to this reset input, you have a  
>'synced' noise source. By changing the start value (and hence the  
>part of the sequence produced) you can change the tone of the result.
>It's an interesting idea, and for the 20 minutes it'd take to program  
>up, I might give it a whirl.
>

That's interesting, if you do it, please post something.  

It sounds like it's using an LFSR to produce a randomly shaped, but repeatable
waveform.  

I've thought of trying something similar in an FPGA, a wave table approach, but use
an LFSR.  The system could reseed the LFSR periodicallywith a new seed every N
reseeds, the new seed would be the next value in the series when seeded with the old
seed.  This would cause the pattern and the waveform to change by one sample every N
full "table" cycles.  I would think that there would be a timbre change, but I have
no idea what it would sound like.


-- ScottG

-----------------------------------------------------------------

-- Scott Gravenhorst
-- GateManPoly - FPGA Based Polyphonic MIDI LA/FM Synthesizer
-- GateMan-III - FPGA Based Monophonic MIDI LA/FM Synthesizer
-- PolyDaWG/8 - FPGA Based 8 Voice Polyphonic MIDI Synthesizer
-- FatMan: home1.gte.net/res0658s/fatman/
-- NonFatMan: home1.gte.net/res0658s/electronics/
-- When the going gets tough, the tough use the command line.