Skip to content

mix_comparison.md

Latest commit

 

History

History
171 lines (121 loc) · 4.11 KB

mix_comparison.md

File metadata and controls

171 lines (121 loc) · 4.11 KB

Tutorial: Compare Two Mixes

This guide explains how to use the Tonn API to compare two different versions of a mix or reference tracks, allowing you to assess progress, consistency, or alignment with a sonic benchmark.

What is Mix Comparison?

Mix Comparison allows you to:

  • Compare two different versions of your mix to track improvements
  • Benchmark your mix against a professionally released track
  • A/B test different mastering approaches

Why Compare Mixes?

Whether you're refining your sound or preparing a release, comparison helps ensure:

  • Improved sound quality: Know if your changes are actually better
  • Reference-based mixing: Match the clarity and energy of tracks you admire
  • Consistency across versions: Spot regressions or overprocessing early

Running the Comparison

cd python/examples
python 02_mix_comparison.py my_mix.wav reference.wav POP

Command Line Options

python 02_mix_comparison.py <file_a> <file_b> <musical_style> [--is-master]
Argument Description
file_a First audio file (your mix)
file_b Second audio file (reference or alternate version)
musical_style Genre (ROCK, POP, ELECTRONIC, etc.)
--is-master Flag if comparing mastered tracks

Understanding the Output

The script provides color-coded, side-by-side comparisons:

=== Production Metrics Comparison ===

integrated_loudness_lufs:
  File A: -13.3
  File B: -10.9
  → Difference of 2.4 exceeds threshold

mono_compatible:
  File A: False
  File B: True
  → Values differ

bit_depth:
  File A: 16
  File B: 24
  → Difference of 8 exceeds threshold

What Gets Compared

Technical Metrics

  • Bit depth
  • Sample rate
  • Clipping status
  • Loudness (LUFS, peak)
  • DRC evaluations
  • Mono compatibility
  • Stereo field and phase issues

Tonal Profile

  • Bass frequency content
  • Low-mid energy
  • High-mid energy
  • High frequency content

Interpreting Results

Loudness Differences

Difference Interpretation
< 1 LUFS Very similar, likely unnoticeable
1-3 LUFS Noticeable difference
> 3 LUFS Significant difference

Clipping

If one file shows clipping and the other doesn't, consider why:

  • Overprocessing during mastering?
  • Source file issues?

Stereo/Mono Compatibility

If your mix isn't mono compatible but your reference is:

  • Check for phase issues between stereo channels
  • Review any stereo widening plugins

Use Cases

1. Tracking Mix Progress

Compare v1 and v2 of your mix:

python 02_mix_comparison.py mix_v1.wav mix_v2.wav ROCK

2. Matching a Reference

Compare your mix to a commercial release:

python 02_mix_comparison.py my_mix.wav reference_track.wav POP

3. A/B Testing Mastering

Compare two different mastering approaches:

python 02_mix_comparison.py master_warm.wav master_bright.wav ELECTRONIC --is-master

Programmatic Usage

from shared import TonnClient

client = TonnClient()

# Upload both files
url_a = client.upload_local_file("mix_a.wav")
url_b = client.upload_local_file("mix_b.wav")

# Analyze both
results = {}
for label, url in [("A", url_a), ("B", url_b)]:
    response = client.post("/mixanalysis", {
        "mixDiagnosisData": {
            "audioFileLocation": url,
            "musicalStyle": "POP",
            "isMaster": True
        }
    })
    results[label] = response.get("mixDiagnosisResults", {})

# Compare specific metrics
loudness_a = results["A"].get("payload", {}).get("integrated_loudness_lufs")
loudness_b = results["B"].get("payload", {}).get("integrated_loudness_lufs")

print(f"Loudness difference: {abs(loudness_a - loudness_b):.1f} LUFS")

Tips for Better Comparisons

  1. Use the same musical style for both tracks
  2. Compare like with like (both mixes or both masters)
  3. Consider the source quality - different sample rates affect results
  4. Don't over-optimize - trust your ears alongside the data

Support

For questions or feedback: