Hi, I am developing an audio product which requires a lot of delay memory, and it would be helpful for my application to be able to store delay samples in the 16-bit float format in order to maximize the use of on chip memory. I have looked in the help contents and it describes how the data is stored and conversion functions, but neglects to mention anything about how to declare variables of the 16bit floating point type.
My program is written in C, and I would like to declare an array of 16 bit floats. How to I do this?
Also, it would be useful if I could do the pack and unpack functions as SIMD, in order to save time in my processing loop. Is there any way to do this one the Sharc?