• State TI Thinks Resolved
  • Locked Locked
  • Replies 2 replies
  • Answers 1 answer
  • Subscribers 99 subscribers
  • Views 1030 views
  • Users 0 members are here
Support feedback
Options
Options
Related

This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

how to convert fract32 to float?

user6000372
Intellectual 560 points

Hi Sir,

Dsplib_c66x_3.4.0 is used in our project,we want to convert fract32 to float,but I can't find these function the lib.

Could you give me a reference code?

Many thanks.

  • 0
    TI__Guru**** 161600 points

    Could you share which device are you using?

    Best Regards,
    Yordan

  • 0
    TI__Guru** 116170 points

    Can you clarify what device are you using for development I am not sure where you have come across fract32 data type in TI software as this is not a data type we define in C66x libraries. In C66x context, we have only seen this usage with TI third party partner audio weaver software. Is this what you are using.

    From some other users our understanding is that the fract32 data type is a signed 32-bit fixed-point data type using 31 fractional bits. This uses the Q format definition is Q1.31, and it has a constant resolution from -1.0 to 0.9999999995. On the other hand, floating point data types have a much larger range but variable resolution. See the IEEE 754 standard for more information on floating point numbers.

    For handling Q format which is typically used for floating point data operations on C64x+ fixed point DSP, we provide a library called IQMATH Lib which should help with the format conversion:

    http://processors.wiki.ti.com/index.php/Software_libraries#IQMath

    Regards,

    Rahul