can someone please explain why these 4096 tables have table lengths of 4032 ?
#define ARMBITREVINDEXTABLE_4096_TABLE_LENGTH ((uint16_t)4032)
#define ARMBITREVINDEXTABLE_FIXED_4096_TABLE_LENGTH ((uint16_t)4032)
I would like to know the formula that generates the 4032 so that I can produce other table sizes
For the 512 and 1024 tables it is 32 smaller
For the 2048 and 4096 tables it is 64 smaller
I suspect if you dig into it there are some short-cut methods, or it uses other tables, or has two tap points within the table providing the overlap required.
Thank you for your response.
I have seen elsewhere that table size is square root of radix times square root of fft size. But none of these numbers tally.
It would be good to be able to dervive this number
View all questions in Keil forum