swrconvert

swr_convert is a function in FFmpeg’s libswresample API that performs audio data conversion. It handles sample rate changes, sample format conversion, and channel layout or interleaving adjustments according to a configured SwrContext. The function is commonly used when an application needs to resample and reformat audio streams without relying on a specific container or file format.

The operation requires a SwrContext that has been configured with the desired input and output parameters, typically created with swr_alloc_set_opts (or swr_alloc) and initialized with swr_init. The function signature is generally swr_convert(struct SwrContext *c, uint8_t **out, int out_count, const uint8_t **in, int in_count). It converts up to in_count input samples per channel into up to out_count output samples per channel, writing into the provided output buffers. The input and output data can be either packed (interleaved) or planar, depending on the configured formats.

Usage pattern usually involves allocating and configuring the SwrContext, preparing input and output buffers, and repeatedly calling swr_convert to process blocks of samples. To flush the internal delay accumulated by resampling, a call with in_count set to 0 is used. After processing is complete, resources are released with swr_free. The function returns the actual number of output samples produced per channel, or a negative error code on failure.

Notes: swr_convert is a low-level API utility; it requires careful buffer management and awareness of potential delays in the resampling pipeline. Utilities like swr_get_delay can help estimate required output buffer sizes, and swr_get_delay may be used to determine how many samples to allocate for downstream processing.