Audio
Use this section when people can join the meeting but still sound hard to understand. These pages cover noise reduction, voice separation, device selection, and the limits that come from mobile hardware or Bluetooth routing.
After you finish this section, you should know whether the problem is local audio capture, the wrong device, or a broader call-quality issue.
Wireframes
Section titled “Wireframes”┌──────────────────────────────────────────────────────────────────────┐
│ Hyper app.hyper.video/settings/audio │
├──────────────────────────────────────────────────────────────────────┤
│ │
│ Audio settings │
│ │
│ Input: Studio USB Mic Output: Desk speakers │
│ Noise cancellation: High Voice isolation: On │
│ Spatial audio: Off Echo protection: Auto │
│ │
│ Live test │
│ ┌──────────────────────────────────────────────────────────────┐ │
│ │ Mic level [||||||||||----] │ │
│ │ Speaker test [ Play sample ] │ │
│ │ Network audio path Stable │ │
│ └──────────────────────────────────────────────────────────────┘ │
│ │
│ [ Save audio profile ] [ Reset ] │
│ │
└──────────────────────────────────────────────────────────────────────┘
Microphone selection, suppression, isolation, and listening tests stay close to the live call so users can adjust without leaving the room.
When to use this section
Section titled “When to use this section”- Participants often join from shared workspaces, coffee shops, travel, or home environments with unstable noise conditions.
- Teams need stronger speech intelligibility without adding separate audio tools.
- You want audio controls that work alongside call quality adaptation rather than separately.
Before you start
Section titled “Before you start”- Choose the correct microphone and output device before testing advanced audio effects.
- Use headphones for the clearest result when evaluating spatial audio or echo-sensitive environments.
- Document any org-wide defaults if you plan to enforce audio behavior for support or customer teams.
What this section covers
Noise cancellation
AI-powered noise suppression removes up to 40dB of background noise — keyboard clatter, traffic, HVAC, and more. Your voice comes through clean while distractions stay out of the call.
Voice isolation
Our AI isolates your voice from ambient sound, so you're heard clearly even in busy offices or noisy cafes. No more "can you repeat that?" — your words land the first time.
Audio controls
Fine-tune your audio with per-device settings. Toggle noise cancellation, adjust input sensitivity, and choose which microphone or speaker to use — all without leaving the call.
Key workflows and controls
Voice isolation
- AI separates your voice from background noise in real time
- Works with any microphone — built-in, headset, or external
- Preserves natural tone and expression
- No robotic artifacts or over-processing
Noise cancellation
- Up to 40dB of background noise suppression
- AI-powered — learns to distinguish voice from noise
- Handles keyboard typing, traffic, HVAC, and more
- Runs efficiently on-device for low latency
Per-device controls
- Toggle noise cancellation and voice isolation independently
- Adjust input sensitivity per microphone
- Switch between devices mid-call
- Output routing for speakers and headphones
Spatial audio
- Participants sound like they're in different positions
- Easier to follow who's speaking in larger meetings
- More natural conversation flow
- Works with stereo headphones
How the workflow fits together
Capture
Your microphone picks up your voice along with ambient sound. Hyper captures the raw audio stream and prepares it for processing — no cloud round-trip, everything happens locally for minimal latency.
Process
On-device AI isolates your voice and suppresses noise. Our models run in real time, adding less than 10ms of latency. Your voice comes through clean while keyboard clatter, fans, and street noise are filtered out.
Deliver
Your processed audio streams to other participants with sub-75ms latency. They hear you clearly — no "can you repeat that?" or "you're breaking up." Spatial audio helps them follow who's speaking in larger calls.
Common starting points
Open offices
- Voice isolation cuts through ambient chatter
- 40dB noise suppression handles keyboard noise
- Colleagues hear you clearly from your desk
- No need for a dedicated meeting room
Coffee shops
- AI filters out espresso machines and conversations
- Your voice stays clear in busy environments
- Works on laptop mics — no headset required
- Professional sound from anywhere
Home offices
- Suppress HVAC, pets, and household noise
- Per-device controls for your setup
- Spatial audio for natural team calls
- Studio-quality sound from a home desk
Questions that come up often
What types of noise does it handle?
Hyper's AI noise cancellation handles common background sounds: keyboard typing, HVAC systems, traffic, construction, coffee shop chatter, and household noise like fans or pets. We achieve up to 40dB of suppression while preserving your natural voice.
How accurate is voice isolation?
Our voice isolation model is trained to separate speech from non-speech with high accuracy. It preserves your natural tone, pauses, and expression while filtering out background noise. In testing, participants report significantly clearer audio compared to raw microphone input.
Which devices are supported?
Audio features work on all Hyper platforms: Mac, Windows, Linux, iOS, and Android. We support built-in mics, USB headsets, Bluetooth headphones, and external microphones. Per-device controls let you choose and tune each input.
Does this affect bandwidth?
No. All audio processing happens on your device before the stream is sent. We send the same compressed audio stream — just cleaner. Bandwidth usage is unchanged, and latency impact is under 10ms.
Can I disable these features?
Yes. You can turn off noise cancellation, voice isolation, or spatial audio at any time — per call or in your account settings. Some users prefer raw audio for music or specific use cases; you're always in control.
Related tasks
Limitations and rollout notes
Section titled “ Limitations and rollout notes ”- Device performance and platform support can affect which audio enhancements are available.
- Bluetooth hardware and mobile operating systems can impose their own routing or codec constraints.
Troubleshooting
Section titled “Troubleshooting”Audio still sounds poor after enabling enhancements
Section titled “Audio still sounds poor after enabling enhancements”- Verify the correct microphone is selected before tuning suppression or isolation levels.
- Switch from speaker output to headphones if echo or room bleed is part of the problem.
- Review the call-quality diagnostics when the problem may be network-related rather than local audio capture.