 Tristan Matthews
|
be6df4d614
remove trailing whitespace
|
9 anni fa |
 Ron
|
e2d3bc3923
Tidy the speex/*.h includes some more
|
11 anni fa |
 Jean-Marc Valin
|
dc31cc4b56
removed some if()s from an ltp inner loop
|
16 anni fa |
 Jean-Marc Valin
|
f4284aea85
Added an option to implement iir_mem16() and fir_mem16() in terms of
|
16 anni fa |
|
Jean-Marc Valin
|
95fceab730
Making it more explicit that the nb filter size is constant
|
16 anni fa |
 Jean-Marc Valin
|
6008abea50
Making it possible to disable either the encoder or the decoder using macros
|
16 anni fa |
|
Jean-Marc Valin
|
785e61a593
Sorting out #include "" from #include <> and using relative paths for including
|
16 anni fa |
 jm
|
bf6e949e48
Removed some warnings
|
17 anni fa |
|
jm
|
5aefcc4833
Patch by Thom Johansen: define and use SPEEX_MEMSET, SPEEX_MEMMOVE and
|
17 anni fa |
|
jm
|
7e5c0cc752
fixed-point: making better use of memory -- saved 512 bytes of stack.
|
18 anni fa |
|
jm
|
d68a61e9f0
merged some fixed-point and floating-point code in open-loop pitch search.
|
18 anni fa |
|
jm
|
5e8289abaa
Now using only the 16-bit version of the filters.
|
18 anni fa |
|
jm
|
a702231023
Removed the generic PSHR/SHL/SHR operators and changed them to either the
|
18 anni fa |
|
jm
|
f14bb23b07
should no longer complain about unused var with the float build.
|
18 anni fa |
|
jm
|
fc6dd40040
Enhancer should no longer overflow on insanely loud signals.
|
18 anni fa |
|
jm
|
d597c8c53c
Think this should prevent overflows in the pitch search when the input is
|
18 anni fa |
|
jm
|
f4f59e83d0
copyright details, more info about the mdf learning rate
|
18 anni fa |
|
jm
|
4f4d5e9072
removed unused var
|
18 anni fa |
|
jm
|
b439ea1921
Patch by David Rowe to split the 3-tap search inner loop in a separate function
|
18 anni fa |
 tterribe
|
d5987c5bcf
Swap code/variable declaration to be pre-C99 compliant.
|
18 anni fa |
|
jm
|
3cdc6668dc
removed a float/fixed split in pitch_gain_search_3tap()
|
18 anni fa |
|
jm
|
1d0ec853c8
Better control of the pitch gain to prevent potential instabilities from
|
18 anni fa |
|
jm
|
ec8c698f45
Patch by David Rowe: sending the stack parameter to *_mem16() filters.
|
18 anni fa |
|
jm
|
6e2899889d
Minor optimisation -- should help the open loop pitch search for N=1
|
18 anni fa |
|
jm
|
1684f0b096
Cleaned up the pitch constraint a bit. Should be better and faster.
|
18 anni fa |
|
jm
|
95f146137e
oops, declarations go before the code
|
18 anni fa |
|
jm
|
1f77636011
C89 fix
|
18 anni fa |
|
jm
|
c50db2c565
Got completely rid of the 64bit type (was only used in the adaptive codebook)
|
18 anni fa |
|
jm
|
1f16e63407
excitation and target are now 16-bit (were 32) in the encoder
|
18 anni fa |
|
jm
|
034e6403d9
decoder excitation now in 16-bit precision (was 32), which saves quite a bit
|
18 anni fa |
