What if your background music could support your message without stealing the show? That is exactly what the right dB choices do. In this post you will get an evidence-based, step-by-step guide to the best decibel (dB) settings and workflow in Audacity so music enhances your audio instead of drowning the voice. You’ll learn recommended target levels, when to use normalise versus gain sliders, how to automate dips with the Envelope Tool, and quick fixes that sound professional without complicated plugins.
This article is for creators who publish branded media – podcasts, explainer videos, eLearning, or promo pieces – and want reliable settings that keep dialogue intelligible and natural.
Table of Contents
Quick Summary
Objective: Give you a replicable Audacity workflow so background music never obscures voice. Use settings and techniques that are simple, repeatable, and suitable for branded media.
Tip: Aim for voice peaks around -6 dB to -12 dB and background music peaks around -18 dB to -25 dB, ensuring at least a ~12–20 dB difference depending on context. For precision, normalise music to -18 dB then fine tune with the gain slider or Envelope Tool. These ranges prioritise clarity while preserving musical presence, and align with accessibility guidance recommending a large enough speech-to-background difference. (W3C)
Why dB levels matter for background music vs voice
Levels determine intelligibility and emotional weight. If the music sits too close to the voice’s level it competes for the same frequencies and attention. Accessibility guidance notes that speech should be substantially louder than background audio to be understood by people with hearing difficulties; a 20 dB foreground advantage makes speech about four times louder than background audio. (W3C)
On the production side, maintaining headroom prevents clipping and preserves dynamic control for compression and final mastering. Audacity’s normalization and meters are the basic tools for headroom management. (manual.audacityteam.org)
Recommended target levels (practical chart)
| Element | Typical target peak | Notes |
| Primary voice (narration/dialogue) | -6 dB to -12 dB | Clear, present, leaves headroom for compression. (Music Guy Mixing) |
| Background music (general) | -18 dB to -25 dB | Start around -18 dB; push quieter to -25 dB for dense or busy music. |
| Minimum speech-to-music difference | 12–20 dB | Use higher difference for complex music or accessibility. (W3C) |
| Final peak ceiling (master/export) | -1.0 dB to -0.5 dB | Leave tiny headroom to avoid inter-sample clipping; many creators use -1 dB. (Delenzo Studio) |
| Loudness (if measuring LUFS) | Speech-focused content: -16 to -18 LUFS | Streaming and broadcast standards differ; use LUFS if platform requires it. (Mastering.com) |
Why these numbers? They balance presence and headroom. Industry practitioners often record and process with an average vocal around -18 dB RMS before compression, which gives good results when adding effects. (Music Guy Mixing)
Step-by-step Audacity workflows
Below are three practical workflows. Pick the one that matches your speed and quality needs.
Method A – Normalise method (precise, recommended)
Best when you want reproducible levels across projects.
- Import both voice and music into separate tracks.
- Process voice first:
- Clean noise (Effect > Noise Reduction) if needed. (manual.audacityteam.org)
- Normalise the voice track to a peak of -6 dB to -8 dB (or to -12 dB if you plan heavy compression). (Delenzo Studio)
- Select the music track only.
- Effect > Volume and Compression > Normalize. Set peak amplitude to -18 dB and apply. (manual.audacityteam.org)
- Play both together and check meters. If music still feels intrusive, nudge music down with the track Gain Slider in the Track Control Panel by -3 dB to -6 dB.
- Use the Envelope Tool to automate extra dips under short dialogue bursts.
Why normalise music to -18 dB? It guarantees you start with music that never exceeds a safe peak, letting voice sit on top predictably. Many podcasts and branded creators use -18 dB as the music baseline before fine adjustments. (swellai.com)
Method B – Gain Slider method (fast)
Best when you are iterating quickly and prefer listening.
- Set up playback and meters.
- Lower the music track Gain Slider until the meters show music peaks around -18 to -20 dB, eyeballing relative difference with voice.
- Adjust voice gain so speech peaks around -8 to -10 dB.
- Use the Envelope Tool for small, manual dips.
This method prioritises listening and quick deployment. It is less precise than normalise but often faster for short projects. Practical forums commonly recommend this real-time approach. (Audacity Forum)
Method C – Envelope automation + EQ (best control)
Best when your music has vocal sections or wide frequency content that clashes with speech.
- After normalising/setting gain, insert subtle EQ on the music:
- Reduce 300 Hz–3 kHz by 2–4 dB to carve space for voice frequencies.
- Use Envelope Tool to automate a 6–12 dB dip when someone speaks.
- For long-form content, use multiple small dips rather than one large one to preserve musical flow.
EQ carve plus envelope automation prevents frequency masking and retains musical energy without masking the dialogue.
When to use compression, limiting and LUFS meters
- Compression: Apply gentle compression on voice (ratio 2:1–4:1, threshold tuned so gain reduction is 2–6 dB). This reduces dynamic swings so voice remains audible over music. (Music Guy Mixing)
- Limiter: Use a soft limiter on the master track at -1 dB to prevent clipping at export. (Delenzo Studio)
- LUFS/Loudness: If publishing to platforms with loudness targets, measure Integrated LUFS and adjust target accordingly (eLearning/podcast platforms vary). Aim for platform guidance; otherwise keep speech-centred projects around -16 to -18 LUFS. (Mastering.com)
Troubleshooting – common problems and fixes
Problem: Music sounds fine alone but clashes with voice.
Fix: Apply a small EQ dip in the 500 Hz–2 kHz band on the music track and automate a 4–8 dB envelope dip during speech.
Problem: Dialogue sometimes gets buried when music swells.
Fix: Create transient envelope points around those moments and drop music by 8–12 dB just for the sentence.
Problem: After normalising, exported audio feels softer than expected.
Fix: Normalise sets peak level; perceived loudness depends on RMS/LUFS and compression. Add gentle compression or raise overall track gain, then reapply final limiter to -1 dB. (BassGorilla.com)
Problem: Audacity meters differ from other apps.
Fix: Audacity displays peak dBFS. When comparing with LUFS-aware meters, expect differences; measure loudness with a dedicated loudness plugin if platform compliance is required. (Mastering.com)
Suggested project settings for branded media
- Sample rate: 48 kHz (standard for video).
- Bit depth (export): 24-bit WAV for masters; 16-bit for final MP3 if required.
- Voice target peak: -8 dB.
- Music target peak: start -18 dB, adjust to taste.
- Final export peak: -1 dB ceiling.
- LUFS (optional): -16 to -18 LUFS for speech-led content unless platform specifies otherwise. (Mastering.com)
FAQs
Q: What exact dB should I normalise music to in Audacity?
A: Start at -18 dB. That is a practical baseline that leaves room for voice to be clearly audible. For very busy or vocal-heavy music, normalise to -20 or -22 dB and then fine tune with gain. (manual.audacityteam.org)
Q: Should I normalise voice too?
A: Yes. Normalise voice to a peak between -6 and -12 dB depending on whether you will compress heavily later. Normalise before gentle compression. (Delenzo Studio)
Q: Is normalising the same as compressing?
A: No. Normalising adjusts peaks to a target amplitude. Compression reduces dynamic range so quieter parts are closer in level to louder parts. Use both: normalise first for headroom, then compress if you need level consistency. (manual.audacityteam.org)
Q: How much should music dip when someone speaks?
A: Typically 6–12 dB during short phrases; for sustained speech or critical dialogue, consider 12–20 dB. The goal is clarity with musical continuity. (W3C)
Q: Do I need LUFS metering in Audacity?
A: Audacity does not include a LUFS meter by default. If you need platform compliance, use an external LUFS meter plugin or export and measure with a loudness tool. Aim for platform targets. (Mastering.com)
Key takeaways
- For branded media keep voice peaks around -6 to -12 dB and music around -18 to -25 dB to preserve clarity.
- Normalise music to -18 dB as a reliable baseline, then refine with gain and envelope automation. (manual.audacityteam.org)
- Use EQ to carve space and light compression on voice if needed; final limiter at -1 dB prevents clipping. Read our article on Quick EQ tips for clear speech
Albert Einstein