3-4 Method of the Audio Data Development Process


3-4-1 Audio Data Development Process


To develop audio for your games, use the following processes in combination with sound development tools like N64 Sound Tools for the PC.



  • How to Make Wave Table Audio Data (SGI)
    1. Sample the voice source data by using any of the many available third-party tools to make the AIFF format file.
    2. Make a code book (a predictive coefficient table) used for the ADPCM encode by using the N64 tabledesign audio development tool.
    3. Compress the AIFF format file to the AIFC (ADPCM) format by using the N64 vadpcm_enc audio development tool.
    4. Make the source file (.inst file) used by the N64 ic audio development tool. The ic source file has defined objects that compose the bank by using a language similar to the C language.
    5. Make the bank files (.tbl, .ctl, and .sym files) by using the N64 ic audio development tool. The bank files are the informational data files used for audio synthesis by the audio library.



  • How to Make MIDI Sequence Audio Data (SGI)
    •
    1. Make the MIDI file by using any of the many available third-party tools.
    2. If the MIDI file is Type1, convert it to Type0 by using the N64 midicvt audio development tool.
    3. When you generate the compressed MIDI data generally used for a game program, convert the Type 0 MIDI file to the compact MIDI file by using the N64 midicomp audio development tool.
    4. Make the sequence bank file (.sbk file) by using the N64 audio development tools, sbc.
    5. Pass the bank files (.tbl, .ctl, and .sym files) to the programmer.

    See each respective tool manual for details on the N64 audio development tools.



    3-4-2 Setting the Sampling Rate

  • Establish the standard output rate for your game between 22.05kHz and 44.1kHz.
    • With the N64, if the rate is too high, the processing time in the RSP takes longer. If you establish a low rate, the processing time in the RSP is shortened, and you can increase the number of voices.

  • With an output rate greater than 32kHz, the human ear notices little difference.
    • A rate less than 22.05kHz produces sampling noise that significantly reduces sound quality.

  • Develop the audio data using a sampling rate that is as close as possible to the N64 reproduction output rate to ensure good sound quality.
    • If you cannot decide on the exact output rate in advance, develop the audio using a higher sampling rate first, and then convert it to a lower rate if necessary. This will minimize any decline in sound quality.