1. 05 Jul, 2015 1 commit
  2. 05 May, 2015 1 commit
  3. 01 Apr, 2015 1 commit
  4. 15 Dec, 2014 1 commit
  5. 07 Dec, 2014 9 commits
  6. 06 Dec, 2014 3 commits
  7. 12 Oct, 2014 1 commit
  8. 26 Sep, 2014 1 commit
  9. 17 Aug, 2014 1 commit
  10. 10 Aug, 2014 2 commits
  11. 07 Aug, 2014 4 commits
  12. 23 Jul, 2014 2 commits
  13. 03 Jul, 2014 4 commits
  14. 02 Jul, 2014 6 commits
    • Tristan Matthews's avatar
      build testresample · 0234d452
      Tristan Matthews authored
      0234d452
    • Jyri Sarha's avatar
      resample: Add NEON optimized inner_product_single for floating point · 4e6faa45
      Jyri Sarha authored
      Also adds inline asm implementations of WORD2INT(x) macro for fixed
      and floating point.
      4e6faa45
    • Jyri Sarha's avatar
      resample: Add NEON optimized inner_product_single for fixed point · 0e5d424f
      Jyri Sarha authored
      Semantics of inner_product_single have also been changed to contain
      the final right shift and saturation so it can also be implemented in
      the optimal way for the used platform. This change affects fixed point
      calculations only.
      
      I also added a new fixed point macro SATURATE32PSHR(x, shift, a). It
      does pretty much the same thing as SATURATE32(PSHR32(x, shift), a),
      but it avoids over flowing in rounding up phase in the rare occasion
      where x has already been saturated. It should also be slightly faster.
      0e5d424f
    • Jyri Sarha's avatar
      configure.ac: Add ARM NEON support · 46b1545c
      Jyri Sarha authored
      Use --enable-neon to force NEON optimization on. The auto detection
      should also work if your CFLAGS supports NEON.
      46b1545c
    • Jyri Sarha's avatar
      resample: Calculate full sinc table (e.g. no sinc interpolation) · 0280b024
      Jyri Sarha authored
      On many systems taking couple of kilo-bytes for full sinc table is not
      a problem. With a straight filter table the number of MAC operations
      needed drops to one quarter. Also the inner loop of the filter becomes
      more simple and easier to optimize. So the full sinc table version
      runs much faster in most cases. The trade off is some kilo-bytes
      bigger memory usage and more CPU needed when initializing the
      resampler.
      
      configure.ac: Add --enable-resample-full-sinc-table conf flag
      
      Use this flag to enable resampling with full sinc table.
      0280b024
    • Ron's avatar
      Make M_PI the same as what math.h defines it as · 2721dca6
      Ron authored
      Since the math.h one is what most systems will use anyway.
      2721dca6
  15. 27 Jun, 2014 2 commits
  16. 25 Jun, 2014 1 commit