Even if you have perfect microphone technique—sitting four fingers away, angled 30 degrees off-axis to avoid plosives—you still pick up background noise.
You have an air conditioner. Your PC fans ramp up when you game. You might live near a busy street. This is normal.
To fix it, you need software noise suppression. Pick the wrong tool, and you destroy your stream's audio. The wrong filter chops off laughs, mutes hype reactions, and tanks your game's framerate.
Here is how to choose the correct noise suppression for your setup.
TL;DR / Quick Rules of Thumb:
- OBS Speex: Best for 90% of gamers. Zero FPS drops.
- OBS RNNoise: Avoid for gaming. Mutes loud screams and hype reactions.
- Nvidia Broadcast: GPU heavy. Tanks your frames. Skip it.
- Discord Krisp: Turn it off. Use Standard suppression so your friends hear your reactions.
- Order of Operations: Noise suppression is always Filter #1.
The FPS Killer: Nvidia Broadcast
Nvidia Broadcast (and the Nvidia Noise Removal tool inside OBS) is heavily marketed. Most streamers should not use it.
This filter uses the Tensor and CUDA cores on your Nvidia GPU to process audio. Those are the same cores required to render frames and encode your stream via NVENC. Turn on Nvidia Broadcast, and it competes directly with your game for hardware resources. It drops your FPS. The NVIDIA Effects SDK behaves the same way directly inside OBS.
It also prioritizes noise removal over preserving natural vocal frequencies. You end up sounding compressed and robotic.
When to use it: Only if you stream from a construction site. If you have a loud AC unit or high-RPM PC fans, skip it.
The Reliable Workhorse: OBS Speex
Add a Noise Suppression filter in OBS, and you see the option Speex (Low CPU usage, low quality).
Ignore the "low quality" label. For 80% to 90% of streamers, this is the correct setting.
Speex is a traditional algorithm. It listens for repeating background sounds—like a humming PC fan or droning air conditioner—and removes them. It runs on your CPU and uses almost zero resources. It will not impact game performance. Use this as your default.
The Scream Killer: OBS RNNoise
The other default option is RNNoise (Good quality, more CPU usage). This is an AI-based neural network.
RNNoise kills repeating noises and tones down room reverb. However, it lacks sophistication regarding volume. If your voice suddenly gets loud, RNNoise treats that loudness as an anomaly. It immediately clamps down on the audio.
If you use RNNoise and play horror games, or if you jump up and yell when clutching a round, the filter mutes your voice for a split second right at the peak of the hype moment. It ruins the clip entirely.
When to use it: Avoid RNNoise for gaming. Use it for calm talking-head videos, storytimes, or business meetings where you will not project your voice.
The VST Upgrades: SpeexDSP and RNN 1.1
If you want better performance than the default OBS filters, use third-party VST plugins.
SpeexDSP: The upgraded version of standard Speex. It handles fan and AC noise with higher sophistication without eating CPU. It is a VST3 plugin. It works natively in Elgato Wavelink. To use it in OBS, install an OBS VST3 support plugin first.
RNN 1.1: A massive improvement over the original RNNoise. It handles transient peaks better. The same warning applies: it suits calm speaking rather than high-energy streaming. Available as a VST2, you drop it directly into OBS or Wavelink without extra steps.
The Best on the Market: Elgato Voice Focus
The absolute best noise suppression available right now is Elgato Voice Focus.
Voice Focus is an AI-trained model. It accurately separates human speech from background noise without chopping off loud moments or degrading voice quality. It beats the proprietary options from Shure, Rode, and Nvidia.
If you own an Elgato audio interface—like the Wave XLR, Wave:3, Wave Neo, or Stream Deck+ XLR Dock—you get Voice Focus for free inside the Wavelink software. If you do not use Elgato hardware, buy the plugin outright for $50. It runs as a VST3.
The Discord Trap: Turn Off Krisp
If you stream while sitting in a Discord call, check your voice settings immediately.
Discord uses an AI noise suppression tool called Krisp. It acts like the old RNNoise. If you laugh, clap your hands, or yell, Krisp assumes it is background noise and mutes you. Your friends hear half of what you say.
Go to Discord Voice & Video settings. Scroll down to Noise Suppression. Change it from Krisp to Standard. Standard functions like Speex. It removes fan noise but lets your personality through to your friends.
The Golden Rule: Order of Operations
Noise suppression must always be the first step in your audio processing chain.
Clean the audio before you do anything else. If you place an EQ or a compressor before noise suppression, you amplify the sound of your AC unit.
Remove the noise first. Then EQ the voice. Elgato Wavelink forces Voice Focus to be the first unmovable block in your chain for this reason. If you build your chain in OBS, make sure Noise Suppression sits at the top of the filter list.
Stop Guessing. Let Me Build Your Audio Chain.
Good audio is a process. First, get the physical mic technique right. Second, strip away room noise with the correct suppression tool. Finally, apply EQ, compression, and limiting to make your voice cut through the game mix.
You can spend weeks testing VSTs and tweaking sliders. Or you can get it done today.
GPU-heavy
Book a 1-on-1 audio setup session with me. You give me remote access via AnyDesk. We hop on a Discord call. I install, route, and perfectly calibrate your entire audio chain for your specific hardware and voice.
I handle the config. You stream.
