File i16u64_to_f32_core.h¶

FileList > i16u64_to_f32 > i16u64_to_f32_core.h

Q15-in-uint64 to float converter. More...

#include "clib_common.h"
#include "jm_perf.h"

Classes¶

Type	Name
struct	i16u64_to_f32_state_t I16U64ToF32 state.

Public Functions¶

Type	Name
i16u64_to_f32_state_t *	i16u64_to_f32_create (float scale) Create a i16u64_to_f32 instance.
void	i16u64_to_f32_destroy (i16u64_to_f32_state_t * state) Destroy a i16u64_to_f32 instance and release all memory.
void	i16u64_to_f32_reset (i16u64_to_f32_state_t * state) Reset i16u64_to_f32 to its post-create state.
JM_FORCEINLINE JM_HOT float	i16u64_to_f32_step (const i16u64_to_f32_state_t * state, uint64_t x) Process one input sample.
void	i16u64_to_f32_steps (i16u64_to_f32_state_t * state, const uint64_t * input, float * output, size_t n) Process a block of Q15-in-uint64 samples to float32.

Detailed Description¶

Extracts the lower 16 bits of a uint64, re-interprets them as a signed int16 (two's complement), then multiplies by 1/scale to produce a normalised float. This is the exact inverse of F32ToI16U64: a value written by that converter can be recovered here with the same scale.

uint64 0x0000000000008000 → int16 -32768 → float -1.0 uint64 0x0000000000007FFF → int16 32767 → float ~+1.0 uint64 0x0000000000000000 → int16 0 → float 0.0

All 48 upper bits are masked off; values carrying NCO phase-accumulator headroom in those bits are handled transparently.

Lifecycle: create -> (step / steps / reset)* -> destroy

>>> from doppler.cvt import I16U64ToF32
>>> import numpy as np
>>> obj = I16U64ToF32(scale=32768.0)
>>> float(obj.step(0x8000))
-1.0
>>> float(obj.step(0x0000))
0.0
>>> x = np.array([0, 0x8000, 0x7FFF], dtype=np.uint64)
>>> [round(v, 6) for v in obj.steps(x).tolist()]
[0.0, -1.0, 0.999969]

Public Functions Documentation¶

function i16u64_to_f32_create¶

Create a i16u64_to_f32 instance.

i16u64_to_f32_state_t * i16u64_to_f32_create (
    float scale
)

Pre-computes iscale = 1.0f / scale so the hot step path is a single multiply after the lower-16-bit extraction.

Parameters:

scale Denominator scale; 1/scale is applied after sign-extension (default: 32768.0f). Use 32768.0 to match F32ToI16U64 at its default scale.

Returns:

Heap-allocated state, or NULL on allocation failure.

Note:

Caller must call i16u64_to_f32_destroy() when done.

function i16u64_to_f32_destroy¶

Destroy a i16u64_to_f32 instance and release all memory.

void i16u64_to_f32_destroy (
    i16u64_to_f32_state_t * state
)

Parameters:

state May be NULL.

function i16u64_to_f32_reset¶

Reset i16u64_to_f32 to its post-create state.

void i16u64_to_f32_reset (
    i16u64_to_f32_state_t * state
)

No mutable state exists beyond the immutable iscale; reset is a no-op provided for lifecycle symmetry.

Parameters:

state Must be non-NULL.

function i16u64_to_f32_step¶

Process one input sample.

JM_FORCEINLINE  JM_HOT float i16u64_to_f32_step (
    const i16u64_to_f32_state_t * state,
    uint64_t x
)

Masks the lower 16 bits, sign-extends to int16, then multiplies by iscale. Upper 48 bits are ignored.

Parameters:

state Must be non-NULL.
x uint64 carrying a Q15 sample in its lower 16 bits.

Returns:

Scaled float32 output.

function i16u64_to_f32_steps¶

Process a block of Q15-in-uint64 samples to float32.

void i16u64_to_f32_steps (
    i16u64_to_f32_state_t * state,
    const uint64_t * input,
    float * output,
    size_t n
)

Applies step() to every element. Accepts an optional pre-allocated output array; allocates a fresh one when output is NULL.

Parameters:

state Must be non-NULL.
input Input uint64 array (Q15 packed in lower 16 bits); must contain at least n elements.
output Output float32 array; must contain at least n elements.
n Number of samples to process.

>>> from doppler.cvt import I16U64ToF32
>>> import numpy as np
>>> I16U64ToF32().steps(np.array([0, 16384], dtype=np.uint64)).tolist()
[0.0, 0.5]

The documentation for this class was generated from the following file native/inc/i16u64_to_f32/i16u64_to_f32_core.h