rlm@0: #LyX 1.6.4 created this file. For more info see http://www.lyx.org/
rlm@0: \lyxformat 345
rlm@0: \begin_document
rlm@0: \begin_header
rlm@0: \textclass article
rlm@0: \use_default_options true
rlm@0: \language english
rlm@0: \inputencoding auto
rlm@0: \font_roman default
rlm@0: \font_sans default
rlm@0: \font_typewriter default
rlm@0: \font_default_family default
rlm@0: \font_sc false
rlm@0: \font_osf false
rlm@0: \font_sf_scale 100
rlm@0: \font_tt_scale 100
rlm@0: 
rlm@0: \graphics default
rlm@0: \paperfontsize default
rlm@0: \use_hyperref false
rlm@0: \papersize default
rlm@0: \use_geometry false
rlm@0: \use_amsmath 1
rlm@0: \use_esint 1
rlm@0: \cite_engine basic
rlm@0: \use_bibtopic false
rlm@0: \paperorientation portrait
rlm@0: \secnumdepth 3
rlm@0: \tocdepth 3
rlm@0: \paragraph_separation indent
rlm@0: \defskip medskip
rlm@0: \quotes_language english
rlm@0: \papercolumns 1
rlm@0: \papersides 1
rlm@0: \paperpagestyle default
rlm@0: \tracking_changes false
rlm@0: \output_changes false
rlm@0: \author "" 
rlm@0: \end_header
rlm@0: 
rlm@0: \begin_body
rlm@0: 
rlm@0: \begin_layout Title
rlm@0: Multi-Voice Audio Playback
rlm@0: \end_layout
rlm@0: 
rlm@0: \begin_layout Author
rlm@0: Laurel Pardue, Robert McIntyre
rlm@0: \end_layout
rlm@0: 
rlm@0: \begin_layout Section*
rlm@0: Micro-Architecture
rlm@0: \end_layout
rlm@0: 
rlm@0: \begin_layout Subsection*
rlm@0: A quick summary of our system
rlm@0: \end_layout
rlm@0: 
rlm@0: \begin_layout Standard
rlm@0: The audio data (
rlm@0: \begin_inset Quotes eld
rlm@0: \end_inset
rlm@0: 
rlm@0: samples
rlm@0: \begin_inset Quotes erd
rlm@0: \end_inset
rlm@0: 
rlm@0: ) start in the memory, but are soon pulled into action by the DMA (direct
rlm@0:  memory access).
rlm@0:  The DMA sends the samples to a chain of 0 or more 
rlm@0: \emph on
rlm@0: soft-cores
rlm@0: \emph default
rlm@0: , where they are transformed according to the soft-cores' algorithms.
rlm@0:  After running the gauntlet of soft-cores, the samples flow first to a buffering
rlm@0:  FIFO, and finally to a 
rlm@0: \emph on
rlm@0: mixer
rlm@0: \emph default
rlm@0: , which sends the samples off to be played by speakers of stored in a file.
rlm@0: \end_layout
rlm@0: 
rlm@0: \begin_layout Subsection*
rlm@0: A more detailed look a our system
rlm@0: \end_layout
rlm@0: 
rlm@0: \begin_layout Standard
rlm@0: Our audio pipeline starts its life with whatever music we want to play already
rlm@0:  in memory.
rlm@0:  The music has multiple tracks, or 
rlm@0: \begin_inset Quotes eld
rlm@0: \end_inset
rlm@0: 
rlm@0: voices,
rlm@0: \begin_inset Quotes erd
rlm@0: \end_inset
rlm@0: 
rlm@0:  which we arrange on the memory in contiguous sequences.
rlm@0:  Therefore, each voice starts at its own start-address, and continues on
rlm@0:  for as long as it needs.
rlm@0:  We cap each voice with an end-of-file marker.
rlm@0:  We make every voice the same length to simplify timing issues, even though
rlm@0:  that might mean that there is a lot of wasted space in the form of dead
rlm@0:  areas with no sound.
rlm@0: \end_layout
rlm@0: 
rlm@0: \begin_layout Standard
rlm@0: When the processing starts, the DMA reads in each voice and sends the audio
rlm@0:  data (
rlm@0: \begin_inset Quotes eld
rlm@0: \end_inset
rlm@0: 
rlm@0: samples
rlm@0: \begin_inset Quotes erd
rlm@0: \end_inset
rlm@0: 
rlm@0: ) off to a chain of soft-cores.
rlm@0:  The DMA knows to which chain the samples should be forwarded by looking
rlm@0:  up the samples' voice in a hash table read at startup.
rlm@0:  Once the samples enter the soft-core chain, the first soft-core in line
rlm@0:  takes the samples and performs some sort of audio operation, producing
rlm@0:  modified audio samples.
rlm@0:  These samples feed into the next soft-core which performs a different algorithm.
rlm@0:  As condition for functional correctness, we require the algorithm that
rlm@0:  each soft-core contains to be able to produce audio samples at the desired
rlm@0:  44khz rate, but we allow an arbitrary setup time for the soft-core to get
rlm@0:  ready.
rlm@0:  That way, a pipelined processor can take its time and fill up its pipeline
rlm@0:  completely with no worries.
rlm@0:  Once the soft-core starts producing samples, it must produce them at a
rlm@0:  rate of at least 44khz forever.
rlm@0:  If every soft-core in the chain obeys this timing condition, then the whole
rlm@0:  chain will as well, so the entire chain of soft-cores can be treated as
rlm@0:  a single big soft-core for timing purposes.
rlm@0:  Our soft-cores are just Lab 5 processors preloaded with specific audio
rlm@0:  processing algorithms.
rlm@0: \end_layout
rlm@0: 
rlm@0: \begin_layout Standard
rlm@0: Eventually, the samples make their way to the end of the chain of soft-cores.
rlm@0:  Every chain of soft-cores ends in a FIFO.
rlm@0:  For this project, we will have 12 separate voices.
rlm@0:  Therefore, there will be 12 soft-core chains, each with a FIFO at its end,
rlm@0:  one for each voice.
rlm@0:  Each of the 12 FIFOs leads to the mixer, which only begins operation once
rlm@0:  all 12 FIFOs are full.
rlm@0:  Because of the timing constraints we impose on our soft-cores, once every
rlm@0:  FIFO has at least one sample, all 12 will forever more be able to supply
rlm@0:  samples at 44khz.
rlm@0:  The mixer therefore waits until all 12 FIFOs have elements and then reads
rlm@0:  samples from all FIFOS at a rate of 44khz.
rlm@0:  The mixer combines the samples through simple addition and scaling, creating
rlm@0:  a single stream of sample data, which is sent back from the FPGA to the
rlm@0:  host computer where we play the stream and/or save it to a file.
rlm@0:  The mixer ensures that it outputs samples at a rate of 44khz by running
rlm@0:  everything through its own internal buffer FIFO clocked at 44khz.
rlm@0: \end_layout
rlm@0: 
rlm@0: \end_body
rlm@0: \end_document