VS1033 - MP3/AAC/WMA/MIDI AUDIO CODEC - VLSI Solution

More documents

Recommendations

Info

VLSISolution yVS1033cVS1033C8. FUNCTIONAL DESCRIPTION8.3 Data Flow of VS1033SDIBitstreamFIFOMP3WAV/ADPCM/WMA / AAC /MIDI decodeSM_ADPCM=0AIADDR = 0SB_AMPLITUDE=0ST_AMPLITUDE=0UserApplicationBassenhancerTrebleenhancerEarSpeakerAIADDR != 0SB_AMPLITUDE!=0ST_AMPLITUDE!=0VolumecontrolAudioFIFOS.rate.conv.and DACLRSCI_VOL2048 stereosamplesFigure 13: Data Flow of VS1033.First, depending on the audio data, and provided ADPCM encoding mode is not set, MP3, WMA, AAC,PCM WAV, IMA ADPCM WAV, or MIDI data is received and decoded from the SDI bus.After decoding, if SCI AIADDR is non-zero, application code is executed from the address pointed toby that register. For more details, see Application Notes for VS10XX.Then data may be sent to the Bass Enhancer and Treble Control depending on the SCI BASS register.Next, headphone processing is performed, if the EarSpeaker spatial processing is active.After that the signal is fed to the volume control unit, which also copies the data to the Audio FIFO.The Audio FIFO holds the data, which is read by the Audio interrupt (Chapter 10.14.1) and fed to thesample rate converter and DACs. The size of the audio FIFO is 2048 stereo (2×16-bit) samples, or 8KiB.The sample rate converter upsamples all different sample rates to XTALI/2, or 128 times the highest usablesample rate with 18-bit precision. This removes the need for complex PLL-based clocking schemesand allows almost unlimited sample rate accuracy with one fixed input clock frequency. With a 12.288MHz clock, the DA converter operates at 128 × 48 kHz, i.e. 6.144 MHz, and creates a stereo in-phaseanalog signal. The oversampled output is low-pass filtered by an on-chip analog filter. This signal is thenforwarded to the earphone amplifier.Version 1.00, 2008-02-01 32
VLSISolution yVS1033cVS1033C8. FUNCTIONAL DESCRIPTION8.4 EarSpeaker Spatial ProcessingWhile listening to the headphones the sound has a tendency to be localized inside the head. The soundfield becomes flat and lacking the sensation of dimensions. This is unnatural, awkward and sometimeseven disturbing situation. This phenomenon is often referred in literature as ‘lateralization’, meaning’in-the-head’ localization. Long-term listening to lateralized sound may lead to listening fatigue.All real-life sound sources are external, leaving traces to the acoustic wavefront that arrives to the eardrums. From these traces, the auditory system in the brain is able to judge the distance and angle of eachsound source. In loudspeaker listening the sound is external and these traces are available. In headphonelistening these traces are missing or ambiguous.The EarSpeaker processing makes listening via headphones more like listening the same music fromreal loudspeakers or live music. Once the EarSpeaker processing is activated, the instruments are movedfrom inside to the outside of the head, making it easier to separate the different instruments (see figure14). The listening experience becomes more natural and pleasant, and the stereo image is sharper as theinstruments are widely on front of the listener instead of being inside the head.Figure 14: EarSpeaker externalized sound sources vs. normal inside-the-head soundNote that EarSpeaker differs from any common spatial processing effects, such as echo, reverb, or bassboost. EarSpeaker simulates accurately human auditory model and real listening environment acoustics.Thus is does not change the tonal character of the music by introducing artificial effects.EarSpeaker processing can be parameterized to a few different modes, each simulating a little differenttype of acoustical situation and suiting for different personal preference and type of recording. Seesection 8.7.1 for how to activate different modes.• Off: Best option when listening through loudspeakers or if the audio to be played contains binauralpreprocessing• minimal: Suits well for listening to normal musical scores with headphones, very subtle• normal: Suits well for listening to normal musical scores with headphones, moves sound sourcefarther than minimal• extreme: Suits well for old or ’dry’ recordings, or if the audio to be played is artificial, for examplegenerated MIDIVersion 1.00, 2008-02-01 33
Page 2 and 3: VLSISolution yVS1033cVS1033CCONTENT
Page 9 and 10: VLSISolution yVS1033cVS1033C1. LICE
Page 11 and 12: VLSISolution yVS1033cVS1033C4. CHAR
Page 13 and 14: 1.10 REFVLSISolution yVS1033cVS1033
Page 15 and 16: VLSISolution yVS1033cVS1033C5. PACK
Page 17 and 18: VLSISolution yVS1033cVS1033C7. SPI
Page 25 and 26: VLSISolution yVS1033cVS1033C8. FUNC
Page 31: VLSISolution yVS1033cVS1033C8. FUNC
Page 45 and 46: VLSISolution yVS1033cVS1033C9. OPER
Page 59 and 60: VLSISolution yVS1033cVS1033C10. VS1
Page 75 and 76: VLSISolution yVS1033cVS1033C11. DOC

VS1033 - MP3/AAC/WMA/MIDI AUDIO CODEC - VLSI Solution

Create successful ePaper yourself

Delete template?

Save as template?