The firmware that makes sound smarter.
We develop audio signal-processing algorithms and embedded firmware that turn hardware into a complete system: from digital filters to the final acoustic tuning.
Dedicated DSPs and ARM Cortex-M cores
The wide availability of DSPs with high computing power at low cost has opened new horizons for applications that require complex real-time audio processing. Alongside the classic Analog Devices and Texas Instruments DSPs, we use the newer ARM Cortex-M4/M7 cores (with integrated FPU, single- and multi-core) for real-time digital audio processing, with very interesting results.
From algorithm to finished product
Real-time DSP processing
FIR and IIR filters, active crossovers (Butterworth, Linkwitz-Riley), limiters, compressors and parametric EQ, within strict latency constraints.
Embedded firmware
Bare-metal and RTOS development for microcontrollers and dedicated DSPs, audio peripheral and control-interface management.
Digital audio protocols
Integration of I2S, TDM, SPDIF and control buses (I2C, SPI, UART) between system components.
Fixed/floating-point optimisation
Porting and optimisation of algorithms on architectures with resource, memory and power constraints.
Acoustic tuning
Fine-tuning of DSP parameters based on the mechanical and acoustic characteristics of the final system.
Update & diagnostics
Bootloaders, bus-based or OTA firmware updates, diagnostic and telemetry tooling.
Digital audio processing
- FIR and IIR digital filtering algorithms, equalization and parametric tone controls
- Digital audio effects: delay, echo, chorus, flanger, reverb
- Active crossover filters: Butterworth and Linkwitz-Riley
- FFT applications, fast convolution, frequency-domain filtering
- Digital compression, limiting and gain-control techniques
- Sound spatialization algorithms
- Sound intensity measurement: RMS value, Leq, A- and C-weighting curves, third-octave analysis
- Fractional adaptive resampling for multi-rate systems
- Synthesis algorithms (additive, subtractive, FM, physical) for sound generation and modelling
DSP that tunes itself
Beyond classic DSP processing, we develop firmware capable of adapting autonomously to the real conditions the system operates in, without manual intervention.
Auto-tuning & automatic calibration
The system measures acoustic conditions and automatically adjusts DSP parameters, with no manual calibration needed.
Real-time adaptation
Parameters that change dynamically based on variable operating conditions: signal level, ambient noise, temperature, load.
Machine-learning-based classification
ML models to recognise and classify the audio signal's content, beyond fixed-rule DSP processing.
Long-distance transmission of digital audio signals
The Bluetooth standard allows audio to be streamed from mobile devices to processing and amplification systems, typically over fairly short distances. WiFi connectivity, on the other hand, allows audio streams to propagate over wireless networks and existing infrastructure, even over greater distances. Other proprietary technologies (usually on the 2.4GHz ISM band) allow wireless audio transmission with the desired range, topology (point/multipoint, unicast/multicast or broadcast), high quality and minimal latency.
There are also several standard protocols for transmitting digital audio signals over wired, TCP/IP-based communication networks. We design communication boards with DSP processors or Embedded Linux platforms suited to implementing audio streaming protocols over wired networks (Audio over LAN, AES67) or wireless ones (Bluetooth, BLE, WiFi, 2.4GHz, UWB).
Firmware and hardware, designed together
The success of an audio processing firmware is strongly influenced by the characteristics of the hardware it runs on. That's why our firmware team works closely with hardware design from the earliest architectural choices.
Hardware Design →Need an audio algorithm or dedicated firmware?
Get in touch.