SoundFont2 Structure
This section explains a bit about the general SoundFont2 structure. It's not necessarily required to know this information in order to use the library, but it may make it easier to understand how the library and SoundFonts work.
Table of Contents
General Structure
A SoundFont has one or more presets, which can normally be selected with the MIDI Program Change command. The presets have one or more preset zones, that have one instrument. The instruments have instrument zones and these have one sample each.
Multiple preset zones can refer to the same instrument and multiple instrument zones can refer to the same samples.
RIFF Chunks
A SoundFont2 file is a RIFF (Resource Interchange File Format) file, which contains multiple chunks. Those chunks are identified by a FourCC (four-character code). If an identifier is not for characters long, it will be padded by spaces. For example, "foo" would be read as "foo ".
Every SoundFont2 file should have at least three top-level sub-chunks: an INFO chunk containing metadata, a sdta chunk containing Wave Audio (WAV) samples and a pdta header containing the preset, instrument and sample headers.
All the raw (unparsed) data is available in the SoundFont2 class, as metaData, sampleData and presetData.
INFO Chunk
The INFO chunk contains at least the following sub-chunks.
ifil- The SoundFont specification version.Available as
metaData.version.isng- The sound engine for which the SoundFont was optimized.Available as
metaData.soundEngine.INAM- The name of the SoundFont.Available as
metaData.name.
The other sub-chunks are optional.
irom- A sound data ROM to which any ROM samples refer.Available as
metaData.rom.iver- A sound data ROM revision to which any ROM samples refer.Available as
metaData.romVersion.ICRD- The creation date of the SoundFont, conventionally in the 'Month Day, Year' format.Available as
metaData.creationDate.IENG- The author or authors of the SoundFont.Available as
metaData.author.IPDR- The product for which the SoundFont is intended.Available as
metaData.product.ICOP- Copyright assertion string associated with the SoundFont.Available as
metaData.copyright.ICMT- Any comments associated with the SoundFont.Available as
metaData.comments.ISFT- The tool used to create the SoundFont.Available as
metaData.createdBy.
sdta Chunk
The sample chunk has one or two sub-chunks.
smpl- The 16-bit WAV sample data.Available as
sampleData.sm24- The 8-bit WAV sample data, in addition to the 16-bit data.Currently not available in the API.
The sdta and sm24 sub-chunks can be combined to get 24-bit WAV sample data, but the library currently does not support this yet.
pdta Chunk
The pdta sub-chunk contains data for presets, instruments and samples in the SoundFont. There are four preset and four instrument chunks, which form the instruments and presets. There is just one sample header chunk, that forms the samples together with the data from the sdta chunk.
Preset Data
The pdta has four sub-chunks that form the preset data.
phdr- The preset headers.Available as
presetData.presetHeaders.pbag- The preset zone indices.Available as
presetData.presetZones.pmod- The preset modulators.Available as
presetData.presetModulatorspgen- The preset generators.Available as
presetData.presetGenerators
Instrument Data
The pdta has four sub-chunks that form the instrument data.
inst- The instrument headers.Available as
presetData.instrumentHeaders.ibag- The instrument zone indices.Available as
presetData.instrumentZones.imod- The instrument modulators.Available as
presetData.instrumentModulatorsigen- The instrument generators.Available as
presetData.instrumentGenerators
Sample Data
The pdta has one sub-chunk with the sample header data.
shdr- The sample headers.Available as
presetData.sampleHeaders
Presets
TODO
Instruments
TODO