Mastering

In-Depth Explanation

Mastering is the final step in audio production before a song is distributed. It takes a finished stereo mix and applies a last layer of processing to ensure the track sounds consistent across all playback systems (club speakers, earbuds, car stereos) and meets the loudness standards of streaming platforms. Mastering uses EQ, compression, stereo enhancement, and limiting to polish the mix to commercial quality.

How Mastering Works

If mixing is the process of balancing individual instruments (drums, bass, vocals) so they blend together, mastering is the process of taking that single stereo file and applying final processing to the whole thing.

A professional mastering engineer works in an acoustically treated room with highly accurate speakers. They use a chain of audio processors (analog hardware or digital plugins) to shape the stereo file. A typical mastering chain includes:

1. Corrective EQ: Removing harsh resonant frequencies or muddy low-end buildup that the mixing engineer missed.
2. Multiband compression: Compressing specific frequency ranges independently. For example, tightening bass frequencies without affecting vocal dynamics.
3. Stereo enhancement: Widening high frequencies to make the mix feel more expansive while keeping low frequencies centered for club playback.
4. Harmonic saturation: Adding subtle harmonic distortion for analog warmth and presence.
5. Limiting: The final step. A limiter prevents audio from exceeding a specific volume threshold (usually -1.0 dB True Peak), allowing the engineer to raise overall volume without digital clipping.

Mastering for Different Formats

A single master file does not serve every distribution format. Professional engineers create format-specific masters:

• Streaming (Spotify, Apple Music): Mastered to specific LUFS targets. The industry sweet spot in 2026 is approximately -14 LUFS integrated with -1.0 dBTP true peak. However, many genre-specific masters still hit -7 to -9 LUFS for sonic density, accepting that platforms will turn them down.
• Vinyl: A specialized master that controls stereo width in bass frequencies and excessive high-end sibilance, both of which can cause a record player needle to jump out of the groove.
• Club/DJ: Often mastered louder with less dynamic range than streaming masters, ensuring the track hits hard on festival sound systems.
• Social media (TikTok, Reels): Many engineers now print a separate "social master" at -8 LUFS with a hard ceiling at -0.1 dBTP, since TikTok and Reels apply their own aggressive limiting.
• Hi-res/Sales (Bandcamp, Qobuz): A dynamic, high-resolution master at 24-bit/96kHz with minimal limiting, preserving full dynamic range for audiophile listeners.

Real-World Example

An independent producer finishes mixing a hip-hop single. The mix sits at -18 LUFS with peaks hitting -3 dBFS. It sounds quiet and limp compared to commercial releases.

They send the mix to a mastering engineer, who applies corrective EQ to tame a 250 Hz mud buildup, uses multiband compression to tighten the low end, adds stereo width to the hi-hats, and sets a limiter ceiling at -1.0 dBTP. The engineer raises the integrated loudness to -9 LUFS. The track now sounds dense, glued, and competitive with major-label hip-hop releases. When uploaded to Spotify, the platform turns it down by 5 dB to meet its -14 LUFS normalization target, but the dense, limited character of the master remains intact.

Total cost for a single-song master from a mid-tier engineer in 2026: approximately $50 to $150. A top-tier engineer charges $200 to $500 per song. AI mastering services (LANDR, eMastered) charge $5 to $15 per track or a monthly subscription.

AI vs. Human Mastering in 2026

AI mastering has improved significantly. Services like LANDR, eMastered, and CloudBounce now use neural networks trained on millions of commercial tracks. They analyze your mix, compare its frequency curve to genre-specific references, and automatically apply EQ and limiting.

Where AI mastering works well: Electronic music, lofi, hip-hop instrumentals, drum and bass, and future bass. These genres have well-defined loudness and tonal conventions, and the training datasets are enormous. For demo tracks, SoundCloud releases, or artists on a tight budget, AI mastering delivers a genuine improvement over an unmastered mix.

Where AI mastering falls short: Acoustic guitar records, jazz, classical, and complex vocal arrangements. AI tools tend to aggressively brighten and compress mixes, which works for pop but destroys the intentional mood of a dark, intimate recording. A human engineer hears emotional intent and respects the artist's vision while fixing technical flaws.

The hybrid approach: In 2026, many producers use AI mastering as a starting point, then tweak the results manually. Tools like iZotope Ozone's Master Assistant provide a starting chain that you can adjust by ear. This gives you the speed of AI with the taste of a human.

Even with AI or professional mastering, the garbage-in-garbage-out rule applies. If the vocals are too quiet or the bass is out of tune in the mix, mastering will only make those problems louder.

Read our guide on mastering for streaming platforms for platform-specific workflows, our guide on mixing vs. mastering to understand where mastering fits in the chain, and our guide on stem mastering to see when stem-based mastering is worth the extra cost.

Related Terms

• DAW (Digital Audio Workstation)
• LUFS
• EQ (Equalization)
• Compression
• Lossless