Vocoder
Definition
A vocoder converts acoustic features into audible sound or re-synthesizes analyzed speech.
A system that converts acoustic features or latent representations into an audio signal or analyzes and resynthesizes speech.
Simply explained
It builds an audible signal from a technical description of sound, such as the voice of a text-to-speech function.
Why it matters in practice
The term helps to select, develop, evaluate AI systems in a well-founded manner and realistically assess their technical risks.
Example from business
A TTS system generates audio from mel spectrograms and measures intelligibility and latency.
Limits and pitfalls
- A vocoder is not the entire TTS system; artifacts, speaker similarity, computational requirements, and misuse risks remain.
Origin & history
Vocoders originated in telecommunications; neural methods like WaveNet increased synthesis quality from 2016 onward.
Use in mid-sized companies
- Possible example: A TTS system generates audio from mel spectrograms and measures intelligibility and latency.
Frequently asked questions
What does vocoder mean?
A system that converts acoustic features or latent representations into an audio signal or analyzes and resynthesizes speech. What should you watch out for? A vocoder is not the entire TTS system; artifacts, speaker similarity, computational requirements, and misuse risks remain.
Sources
- WaveNet · van den Oord et al., WaveNet
Source checked on 10.10.2026. Editorial use cases are hypothetical; no evidence of real customer results.
Editorially reviewed on: