¿Cómo convertir un archivo de audio RAW de 16 bits a otra cosa en Mac?

A menos que me equivoque, puede descargar los binarios de SoX precompilados para Mac directamente desde Sourceforge . Acabo de probarlo en Lion y parece funcionar para mí.

Alternativamente, afconvert está integrado y se puede usar para convertir una gran variedad de formatos de archivos de audio.

Usage:
afconvert [option...] input_file [output_file]
    Options may appear before or after the direct arguments. If output_file
    is not specified, a name is generated programmatically and the file
    is written into the same directory as input_file.
afconvert input_file [-o output_file [option...]]...
    Output file options apply to the previous output_file. Other options
    may appear anywhere.

General options:
    { -d | --data } data_format[@sample_rate][/format_flags][#frames_per_packet]
        [-][BE|LE]{F|[U]I}{8|16|24|32|64}          (PCM)
            e.g.   BEI16   F32@44100
        or a data format appropriate to file format (see -hf)
        format_flags: hex digits, e.g. '80'
        Frames per packet can be specified for some encoders, e.g.: samr#12
        A format of "0" specifies the same format as the source file,
            with packets copied exactly.
        A format of "N" specifies the destination format should be the
            native format of the lossless encoded source file (alac, FLAC only)
    { -c | --channels } number_of_channels
        add/remove channels without regard to order
    { -l | --channellayout } layout_tag
        layout_tag: name of a constant from CoreAudioTypes.h
          (prefix "kAudioChannelLayoutTag_" may be omitted)
        if specified once, applies to output file; if twice, the first
          applies to the input file, the second to the output file
    { -b | --bitrate } total_bit_rate_bps
         e.g. 256000 will give you roughly:
             for stereo source: 128000 bits per channel
             for 5.1 source: 51000 bits per channel
                 (the .1 channel consumes few bits and can be discounted in the
                 total bit rate calculation)
    { -q | --quality } codec_quality
        codec_quality: 0-127
    { -r | --src-quality } src_quality
        src_quality (sample rate converter quality): 0-127 (default is 127)
    { --src-complexity } src_complexity
        src_complexity (sample rate converter complexity): line, norm, bats
    { -s | --strategy } strategy
        bitrate allocation strategy for encoding an audio track
        0 for CBR, 1 for ABR, 2 for VBR_constrained, 3 for VBR
    --prime-method method
        decode priming method (see AudioConverter.h)
    --prime-override samples_prime samples_remain
        can be used to override the priming information stored in the source
        file to the specified values. If -1 is specified for either, the value
        in the file is used.
    --no-filler
        don't page-align audio data in the output file
    --soundcheck-generate
        analyze audio, add SoundCheck data to the output file
    --codec-manuf codec_manuf
        specify the codec with the specified 4-character component manufacturer
        code
    --dither algorithm
        algorithm: 1-3
    --mix
        enable channel downmixing
    { -u | --userproperty } property value
        set an arbitrary AudioConverter property to a given value
        property is a four-character code; value is signed 32-bit integer.
        A maximum of 8 properties may be set.
        e.g. '-u vbrq <sound_quality>' sets the sound quality level
             (<sound_quality>: 0-127)

Input file options:
    --read-track track_index
        For input files containing multiple tracks, the index (0..n-1)
        of the track to read and convert.
    --offset number_of_frames
        the starting offset in the input file in frames. (The first frame is
        frame zero.)
    --soundcheck-read
         read SoundCheck data from source file and set it on any destination
         file(s) of appropriate filetype (.m4a, .caf).

Output file options:
    -o filename
        specify an (additional) output file.
    { -f | --file } file_format
        use -hf for a complete list of supported file/data formats

Other options:
    { -v | --verbose }
        print progress verbosely
    { -t | --tag }
        If encoding to CAF, store the source file's format and name in a user
        chunk. If decoding from CAF, use the destination format and filename
        found in a user chunk.
    { --leaks }
        run leaks at the end of the conversion
    { --profile }
        collect and print performance information

Help options:
    { -hf | --help-formats }
        print a list of supported file/data formats
    { -h | --help }
        print this help

Además, al escribir afconvert -hf se mostrarán los formatos de audio admitidos