Go to the source code of this file.
|
| class | float16 |
| | Half-precision IEEE-754 binary16 float compressed into 2 bytes (1 sign + 5 exponent + 10 mantissa bits, no infinity/NaN encoding), used to halve the memory footprint of large float buffers (e.g. sample or FFT data) at the cost of precision: range +-131008.0, minimum positive normal ~6.1035156E-5, ~3.3 significant decimal digits. Unlike float32, this is a genuinely lossy compression – converting a value in and back out will generally not return the exact original float. More...
|
| |
|
| namespace | audio_tools |
| | Generic Implementation of sound input and output for desktop environments using portaudio.
|
| |
| namespace | std |
| |
|
| float | fabs (float16 arg) |
| |
| float | floor (float16 arg) |
| |
| float | operator* (float one, float16 two) |
| |
| float | operator* (float16 one, float two) |
| |
| float | operator* (float16 one, float16 two) |
| |
| float | operator+ (float one, float16 two) |
| |
| float | operator+ (float16 one, float two) |
| |
| float | operator+ (float16 one, float16 two) |
| |
| float | operator- (float one, float16 two) |
| |
| float | operator- (float16 one, float two) |
| |
| float | operator- (float16 one, float16 two) |
| |
| float | operator/ (float one, float16 two) |
| |
| float | operator/ (float16 one, float two) |
| |
| float | operator/ (float16 one, float16 two) |
| |
