May be a simple or silly question, but;  what precision can I expect from 
simple i-rate divides ?  To the right of the decimal point.  As I can't seem 
to get below 0.001    

I can't find any mention of precision or resolution, in the manual.
And feel I should really be knowing what to expect.  

I am using the -f csound package, which I understand to be 32-bit. 
Not the -d  double precision. 

With the 32 package, can I get below 0.001  ?  

Could it be that further precision is there, but using the  print ivar  
statement is inappropriate, and  printf  would help?

Any help/clarification much appreciated.