Csound Csound-dev Csound-tekno Search About

[Csnd] UDO question..

Date2008-11-24 16:30
From"Rory Walsh"
Subject[Csnd] UDO question..
AttachmentsNone  

Date2008-11-24 16:56
From"Steven Yi"
Subject[Csnd] Re: UDO question..
AttachmentsNone  

Date2008-11-24 18:00
FromRichard Dobson
Subject[Csnd] Re: UDO question..
Note that (as indicated in the docs) the buzz algorithm by its nature 
does not support the smooth fade in/out of harmonics.  It uses what is 
called a "closed-form solution" to the otherwise expensive arithmetic of 
generating N partials very inexpensively; it cannot be worked around. 
The price of this is that N cannot be fractional, so dynamic variations 
of knh (e.g. in response to a broad  change of pitch) will result in pops.

So  the answer to your student's question is "no". It's something of a 
Holy Grail question.  Were there such a cheap way of generating 
dynamically band-limited signals (from which beloved analog waveshapes 
such as square and sawtooth can be obtained), we would all be using it, 
and analog-modelled synths would be much cheaper than they are!


Richard Dobson


Rory Walsh wrote:
> One of my students was asking if there is a version of the the buzz
> opcode that has a fade-in for each harmonic so that pops do not occur
> in the output. I didn't know if there was and couldn't find on in the
> manual so I set about doing a simple UDO for him. Here's what I have,
> the problem is explained below:
> 




Date2008-11-24 19:58
From"Rory Walsh"
Subject[Csnd] Re: Re: UDO question..
AttachmentsNone  

Date2008-11-25 00:15
From"Steven Yi"
Subject[Csnd] Re: Re: Re: UDO question..
AttachmentsNone  

Date2008-11-25 09:42
From"Rory Walsh"
Subject[Csnd] Re: Re: Re: Re: UDO question..
AttachmentsNone  

Date2008-11-25 11:26
From"Rory Walsh"
Subject[Csnd] Re: Re: Re: Re: UDO question..
AttachmentsNone  

Date2008-11-25 16:38
From"Steven Yi"
Subject[Csnd] Re: Re: Re: Re: Re: UDO question..
AttachmentsNone  

Date2008-11-25 16:58
From"Rory Walsh"
Subject[Csnd] Re: Re: Re: Re: Re: Re: UDO question..
AttachmentsNone