Conversa FAQ
Q: What is Conversa?
A: Conversa is an advanced full-duplex voice processing software that leverages traditional signal processing and machine learning techniques to provide excellent speech quality in all noise types. It includes a real-time PC-based tuning tool that provides complete control of the signal path.
Q: How many microphones does Conversa support?
A: The number of microphones supported by Conversa is scalable. Depending on the application, typically 1-4 microphones will be required, but more microphones can be supported.
Q: Which types of devices is Conversa designed for?
A: All communication applications where high-quality speech is required, such as speakerphones, communication/office headsets, wearables/smartwatches, earbuds/TWS, gaming headsets, hearables, etc.
Q: Which NXP processors can Conversa run on?
A: The list is growing, but at present Conversa is supported on the RT1060, RT1170, RT500 and RT600.
Q: What are the system resource requirements for Conversa?
A: The system resource requirement depends on several factors, such as number of microphones, number of speakers, sample rate, latency requirements, etc.
The memory required varies from around 100 kB (1-2 microphones without AEC), up to > 1 MB (multiple microphones, multiple speakers, and AEC). The MHz required varies from 50 MHz up to 1GHz.
Q: Who do I contact to learn more about whether Conversa is the right software for my voice communication product?
A: Please contact Voice@nxp.com for more information.
Q: What is the process of tuning Conversa for a product?
A: The customer can complete part of the tuning using the Conversa Tuning Tool, but the final tuning is typically done in close cooperation between the customer and the NXP Voice & Audio team.
Q: How can I have my product certified for Microsoft Teams and/or Zoom?
A: Our engineers can assist in the tuning process and provide pre-certification tests for various communication standards.
