circuitpython/shared-module/audiocore/__init__.c

Ignoring revisions in .git-blame-ignore-revs. Click here to bypass and see the normal blame view.

133 lines
5.0 KiB
C
Raw Normal View History

/*
* This file is part of the MicroPython project, http://micropython.org/
*
* The MIT License (MIT)
*
* Copyright (c) 2018 Scott Shawcroft for Adafruit Industries
*
* Permission is hereby granted, free of charge, to any person obtaining a copy
* of this software and associated documentation files (the "Software"), to deal
* in the Software without restriction, including without limitation the rights
* to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
* copies of the Software, and to permit persons to whom the Software is
* furnished to do so, subject to the following conditions:
*
* The above copyright notice and this permission notice shall be included in
* all copies or substantial portions of the Software.
*
* THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
* IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
* FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
* AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
* LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
* OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN
* THE SOFTWARE.
*/
#include "shared-module/audioio/__init__.h"
#include "py/obj.h"
#include "shared-bindings/audiocore/RawSample.h"
#include "shared-bindings/audiocore/WaveFile.h"
#include "shared-module/audiocore/RawSample.h"
#include "shared-module/audiocore/WaveFile.h"
#include "shared-bindings/audiomixer/Mixer.h"
#include "shared-module/audiomixer/Mixer.h"
uint32_t audiosample_sample_rate(mp_obj_t sample_obj) {
const audiosample_p_t *proto = mp_proto_get_or_throw(MP_QSTR_protocol_audiosample, sample_obj);
return proto->sample_rate(MP_OBJ_TO_PTR(sample_obj));
}
uint8_t audiosample_bits_per_sample(mp_obj_t sample_obj) {
const audiosample_p_t *proto = mp_proto_get_or_throw(MP_QSTR_protocol_audiosample, sample_obj);
return proto->bits_per_sample(MP_OBJ_TO_PTR(sample_obj));
}
uint8_t audiosample_channel_count(mp_obj_t sample_obj) {
const audiosample_p_t *proto = mp_proto_get_or_throw(MP_QSTR_protocol_audiosample, sample_obj);
return proto->channel_count(MP_OBJ_TO_PTR(sample_obj));
}
void audiosample_reset_buffer(mp_obj_t sample_obj, bool single_channel_output, uint8_t audio_channel) {
const audiosample_p_t *proto = mp_proto_get_or_throw(MP_QSTR_protocol_audiosample, sample_obj);
proto->reset_buffer(MP_OBJ_TO_PTR(sample_obj), single_channel_output, audio_channel);
}
audioio_get_buffer_result_t audiosample_get_buffer(mp_obj_t sample_obj,
bool single_channel_output,
2021-03-15 09:57:36 -04:00
uint8_t channel,
uint8_t **buffer, uint32_t *buffer_length) {
const audiosample_p_t *proto = mp_proto_get_or_throw(MP_QSTR_protocol_audiosample, sample_obj);
return proto->get_buffer(MP_OBJ_TO_PTR(sample_obj), single_channel_output, channel, buffer, buffer_length);
}
void audiosample_get_buffer_structure(mp_obj_t sample_obj, bool single_channel_output,
2021-03-15 09:57:36 -04:00
bool *single_buffer, bool *samples_signed,
uint32_t *max_buffer_length, uint8_t *spacing) {
const audiosample_p_t *proto = mp_proto_get_or_throw(MP_QSTR_protocol_audiosample, sample_obj);
proto->get_buffer_structure(MP_OBJ_TO_PTR(sample_obj), single_channel_output, single_buffer,
samples_signed, max_buffer_length, spacing);
}
2020-10-30 11:05:16 -04:00
void audiosample_convert_u8m_s16s(int16_t *buffer_out, const uint8_t *buffer_in, size_t nframes) {
2021-03-15 09:57:36 -04:00
for (; nframes--;) {
2020-10-30 11:05:16 -04:00
int16_t sample = (*buffer_in++ - 0x80) << 8;
*buffer_out++ = sample;
*buffer_out++ = sample;
}
}
void audiosample_convert_u8s_s16s(int16_t *buffer_out, const uint8_t *buffer_in, size_t nframes) {
2021-03-15 09:57:36 -04:00
size_t nsamples = 2 * nframes;
for (; nsamples--;) {
2020-10-30 11:05:16 -04:00
int16_t sample = (*buffer_in++ - 0x80) << 8;
*buffer_out++ = sample;
}
}
void audiosample_convert_s8m_s16s(int16_t *buffer_out, const int8_t *buffer_in, size_t nframes) {
2021-03-15 09:57:36 -04:00
for (; nframes--;) {
2020-10-30 11:05:16 -04:00
int16_t sample = (*buffer_in++) << 8;
*buffer_out++ = sample;
*buffer_out++ = sample;
}
}
void audiosample_convert_s8s_s16s(int16_t *buffer_out, const int8_t *buffer_in, size_t nframes) {
2021-03-15 09:57:36 -04:00
size_t nsamples = 2 * nframes;
for (; nsamples--;) {
2020-10-30 11:05:16 -04:00
int16_t sample = (*buffer_in++) << 8;
*buffer_out++ = sample;
}
}
void audiosample_convert_u16m_s16s(int16_t *buffer_out, const uint16_t *buffer_in, size_t nframes) {
2021-03-15 09:57:36 -04:00
for (; nframes--;) {
2020-10-30 11:05:16 -04:00
int16_t sample = *buffer_in++ - 0x8000;
*buffer_out++ = sample;
*buffer_out++ = sample;
}
}
void audiosample_convert_u16s_s16s(int16_t *buffer_out, const uint16_t *buffer_in, size_t nframes) {
2021-03-15 09:57:36 -04:00
size_t nsamples = 2 * nframes;
for (; nsamples--;) {
2020-10-30 11:05:16 -04:00
int16_t sample = *buffer_in++ - 0x8000;
*buffer_out++ = sample;
}
}
void audiosample_convert_s16m_s16s(int16_t *buffer_out, const int16_t *buffer_in, size_t nframes) {
2021-03-15 09:57:36 -04:00
for (; nframes--;) {
2020-10-30 11:05:16 -04:00
int16_t sample = *buffer_in++;
*buffer_out++ = sample;
*buffer_out++ = sample;
}
}