We propose to design and implement a real-time voice-changer. The system will capture live microphone input, process it on the FPGA with a selectable chain of DSP effects (pitch shift, formant shift, robot/vocoder, delay/reverb, ect), and output the transformed audio to headphones or speakers.
This project reuses and extends concepts and modules from the existing lab 5 with the aim to produce a usable, demonstrable voice-changer.
- Capture audio from the Pmod I2S2 (CS5343) at 48 kHz sample rate
- Implement at least three real-time effects: pitch shifting, formant shifting (timbre control), and a robot/vocoder-like effect -- really whatever we come up with.
- Provide user controls (switches/buttons or UART/USB commands) to select effect type and parameters.
The existing lab material for the Nexys A7 (Lab 5: DAC Siren) demonstrates:
- Interfacing to the Pmod I2S / Pmod I2S2 hardware.
- A
dac_ifmodule that serializes parallel audio words into I2S-format output. - Simple tone-generation modules (
tone,wail) and top-level wiring (siren).
We will reuse dac_if.vhd and the constraint approach in siren.xdc, and add an ADC receive interface (adc_if / i2s_rx), buffering, DSP blocks, and control logic.
Files and modules to examine / reuse from the repository:
dac_if.vhd— I2S/DAC transmit interface (adapt to 24-bit stereo output)tone.vhd,wail.vhd,siren.vhd— examples of timing, sample-rate control, and top-level wiringsiren.xdc— constraint file to adapt for Pmod I2S2 pin assignments