Raw audio
The uncompressed digital sound signal, like raw microphone input, before any processing or compression.
Raw audio is the actual sound data captured by a microphone. Think of it as the direct electrical signal from the microphone before any changes are made.
This raw signal is then converted into a digital format. It's a sequence of numbers representing the sound's loudness at very frequent intervals. This format is uncompressed, meaning it contains all the original sound information.
Raw audio refers to the uncompressed digital representation of sound waves. It is typically stored as a sequence of amplitude values sampled at a high frequency (e.g., 44.1 kHz or 48 kHz) and bit depth (e.g., 16-bit or 24-bit). This format preserves the full fidelity of the original sound, making it ideal for audio editing and processing but resulting in large file sizes.
Models that generate raw audio, like OpenAI's Jukebox, aim to produce speech or music directly in this uncompressed format. This contrasts with systems that generate intermediate representations (like spectrograms) and then use a separate vocoder to synthesize the final audio waveform.
Frontier AI labs are increasingly focused on generating realistic audio, including speech and music. Producing raw audio directly is a challenging goal that pushes the boundaries of generative models, moving beyond simpler text-to-speech or symbolic music generation.
- JukeboxOpenAI · OpenAI News ·
- OpenAI ChatGPT
- OpenAI's Jukebox model generates music with singing in raw audio.
- Compressed audio
- Compressed audio formats like MP3 reduce file size by removing
- Spectrogram
- A spectrogram is a visual representation of sound frequencies over
- MIDI
- MIDI is a set of instructions for musical instruments, not an actual audio
The term 'raw audio' originates from digital signal processing and computer audio engineering, referring to the unmanipulated, uncompressed digital data representing sound.
You can experience raw audio by recording your voice using a simple voice recorder app on your phone or computer. Then, play it back without applying any filters or effects to hear the unedited sound. Many audio editing software programs allow you to import and view the raw waveform of a sound file.
The Playbook: reusable how-tos, prompts, configs and tips pulled out of the news.
- PromptRole-play a student struggling with math for AI tutor simulationSimulate a student's math struggles and uncertainty for AI tutor evaluation. · 5 min
- TemplateReusable workflow templates for healthcare teamsTeams can create and share templates for common workflows like documentation or prior authorizations. · 15 min
- TemplateTemplate for building an AI tool tailored to your workA structured approach to defining and building a custom AI tool for specific business needs. · 15 min
- TemplateEvals CSV prompt list for agent skillsA CSV file template for listing prompts used in evaluating agent skills. · 20 min
- HeuristicDouble coding agent task completion with docs linksProviding coding agents with a documentation link significantly increases their task completion rate. · 2 min
Questions people ask
- What is the difference between raw audio and compressed audio?
- Raw audio is uncompressed and contains all original sound data, leading to high quality but large file sizes. Compressed audio reduces file size by removing some data, which can slightly reduce quality.
- Why is raw audio important for AI models?
- Generating raw audio directly is a more complex and advanced task for AI models, allowing for more realistic and nuanced sound generation compared to using intermediate representations.
First seen on Subvolts Apr 30, 2020. Updated Sep 15, 2026. Written by an AI model from the items above and reviewed against them weekly; report an error.