Intel® Integrated Performance Primitives
Community support and discussions relating to developing high-performance vision, signal, security, and storage applications.
Announcements
This community is designed for sharing of public information. Please do not share Intel or third-party confidential information here.

IPP library non-matching results

postaquestion
Beginner
163 Views

Hello,

The DCT and IDCT functions provided by your IPP library do not produce matching results.

See, in the case like:

[X0] -> DCT -> [Y1] -> IDCT -> [X1] -> DCT -> [Y2]

In result, [Y1] != [Y2].

Any suggestions? Thank you.

0 Kudos
1 Reply
Vladimir_Dudnik
Employee
163 Views

The suggestion is to carefully read IPP documentation and properly use IPP functions.

If you use 16-bit DCT functions you need to provide data in some particular range, 9-bit if I remember it correctly. You also mayexpect rounding errors in integer arithmetic used in 16-bit functions. You may use floating point IPP DCT functions if the accuracy of 16-bit integer functions is not enough for your application.

Vladimir

Reply