Audio Processing
Sample rate conversion, graphic/parametric equalisation and spectral display — the processing a player needs around the decoder.
Where each algorithm sits
Decoded PCM in, DAC out. Click a node to jump to its demonstration.
Stream
(VINJEY)
Rate Converter
Equalizer
Output
pre-IMDCT tap
See what each algorithm does
Move the controls and watch the response update.
VSRC: Alias Attenuation Vs CPU Resource Footprint
44.1 KHz to 48 KHz. Select Quality Preset configured at build time.
VEQ — 5-band graphic equalizer
Drag a band to boost or cut it.
VSPEC: 30-band spectrum without the FFT
Switch between the two paths and watch the pipeline lose a stage. Same 30-band output, 3–6 MHz cheaper.
Find your algorithm, then check what it takes
The Product/Platform matrix shows which algorithms are supported on which platforms, with a datasheet for each. The calculator gives measured resource footprint — MHz and memory — for a specific core and stream.
| Algorithm & Product Code | ARM Cortex-M | AArch32 Cortex-A/R | AArch64 Cortex-A | EspressIf RISC-V | Renesas RX | TI C6x DSP | Status | Datasheet |
|---|
CPU Performance Parameter Selector
Select target silicon architecture, core family, operation, and stream profile to view measured MIPS and memory footprints.
SALIENT FEATURES
Common and algorithm-specific characteristics of our audio processing algorithms.
- Performance engineered – Optimized for CPU Resource Footprint. Tradeoff between CPU resource footprint and audio quality is provided.
- Simple API with documentation and a sample application
- Reentrant – verified with multiple instances
- Bit-accurate PC implementation alongside the target code
- Trial version available on request
- Availability: Available on widely used DSPs and MCUs
- Supports both 16/24 bit interleaved/non-interleaved PCM inputs
- Audio Quality Vs CPU Resource footprint decided at build time.
- Three standard configurations, or one designed to customer's requirements
- Input Sampling Rate Supported: 8000, 11025, 12000, 16000, 22050, 24000, 32000, 44100, 48000, 88200, 96000, 176400 and 192000 Hz
- Output Sampling Rate Supported: 44100, 48000, 96000 and 192000 Hz
- One library, two modes. Fixed bands for graphical equalizer and configurable centre frequency, band width and band count for parametric equalizer.
- Supplied in 3, 5 and 10 band configurations.
- Shelving filters for low/high pass and peaking filters for band pass.
- Configurability of PEQ happens at initialization time, so no run-time overhead.
- Produces a 30-band spectral output from the inside of decoder.
- FFT based implementation available for use with lossless audio codecs like FLAC
- Useful for generating content based visualizations
Ready to evaluate our Audio Codecs?
Request an evaluation build and check our numbers on your machine.