uSound

This is a header-only implementation of a system-friendly audio XBIOS setup. It doesn't work out of the box on the STE/TT (unless equipped with EmuTOS) however it should work with most of the XBIOS emulators available.

It consists of an enum, struct and two functions:

typedef enum {
    AudioFormatSigned8,
    AudioFormatSigned16LSB,
    AudioFormatSigned16MSB,
    AudioFormatUnsigned8,
    AudioFormatUnsigned16LSB,
    AudioFormatUnsigned16MSB,

    AudioFormatCount
} AudioFormat;

typedef struct {
    uint16_t    frequency;  /* in samples per second */
    uint8_t     channels;   /* 1: mono, 2: stereo */
    AudioFormat format;     /* see AudioFormat */
    uint16_t    samples;    /* number of samples to process (2^N) */
    uint16_t    size;       /* buffer size (calculated ) */
} AudioSpec;

int AtariSoundSetupInitXbios(const AudioSpec* desired, AudioSpec* obtained);
int AtariSoundSetupDeinitXbios();

If you worked with SDL-1.2's SDL_OpenAudio this should feel familiar. The biggest difference here is that the obtained parameter is mandatory, i.e. built-in conversion is not available.

AtariSoundSetupInitXbios / AtariSoundSetupDeinitXbios return 1 (true) on success and 0 (false) on failure. The return value of 1 also implies availability of the sound XBIOS API. AtariSoundSetupInitXbios may change frequency, channels and format parameters so always check them before usage!