| 1 |
$Id: DEFINES,v 1.23 2003-02-07 18:17:41 bouvigne Exp $ |
| 2 |
|
| 3 |
USE_FAST_LOG: |
| 4 |
- use of log/log10 approximation (uses IEEE754 float format) |
| 5 |
(absolute precision of log10 is then around 1e-6) |
| 6 |
|
| 7 |
KLEMM_36: |
| 8 |
- portability fixes in the IO code |
| 9 |
|
| 10 |
NON_LINEAR_PSYMODEL |
| 11 |
- use a non linear psymodel in the GPSYCHO case |
| 12 |
|
| 13 |
USE_GOGO_SUBBAND: |
| 14 |
? |
| 15 |
|
| 16 |
NOTABLES (default): |
| 17 |
? |
| 18 |
|
| 19 |
NEWS3: |
| 20 |
? |
| 21 |
|
| 22 |
NORES_TEST (allways on): |
| 23 |
- don't let the 2nd granule use bits saved by the 1st granule |
| 24 |
- useful for testing only |
| 25 |
|
| 26 |
NEW_DRAIN (theres a define above: NEW_DRAINXX): |
| 27 |
? |
| 28 |
comment: |
| 29 |
mdb_bytes = x/8; m ?= n ?= o ?= 8* mdb_bytes; p ?= mdb_bytes |
| 30 |
( ? == [+-] ) |
| 31 |
- optimization should handle this, but it looks ugly |
| 32 |
- do we lose precision here? |
| 33 |
|
| 34 |
LAME_STD_PRINT: |
| 35 |
- more verbose output? |
| 36 |
|
| 37 |
PRECOMPUTE (always on, multiple defines): |
| 38 |
- precomputes blackman window? |
| 39 |
|
| 40 |
USE_GNUC_ASM: |
| 41 |
- speed optimization |
| 42 |
(should move into configure.in) |
| 43 |
|
| 44 |
... alot of #if 0 / #if 1 not evaluated ... |
| 45 |
|