Mixes with seemingly healthy stereo correlation can hide problematic low-end phase issues that only appear on mono playback. Sub stereo width targets the sub-bass region specifically, where phase problems cause the most damage.
What sub-stereo-width reveals (and why it matters)
Sub stereo width measures how much stereo difference exists in the sub-bass frequency range below 100Hz, expressed as a width score from 0 to 1. A score of 0 means the sub region is fully mono. A score of 1 means the left and right channels are fully uncorrelated. Most commercial mixes keep the sub region tightly mono to preserve bass energy on mono systems and prevent phase cancellation (Source: inputs/articles/sub-stereo-width/brief.md#Core message).
This metric matters because it reveals problems that full-mix correlation might miss. A mix can show healthy overall correlation whilst hiding significant stereo spread concentrated in the low end. When bass energy is out of phase between channels, mono playback systems lose that energy entirely through phase cancellation.
Sub frequencies are omni-directional. Human ears cannot localize bass below roughly 80 to 100Hz. Stereo information in this range adds no perceived width but creates problems on mono playback, reduces bass energy on summed systems, and causes issues with vinyl cutting (Source: inputs/articles/sub-stereo-width/brief.md#Core message).
How sub-stereo-width works: technical methodology
The measurement process isolates the sub-bass frequency range and calculates correlation specifically for that region. Separate STFTs are computed for the left and right channels. Frequency bins below 100Hz are isolated from each channel. These bins are then flattened into vectors representing the sub-bass content of each channel (Source: inputs/articles/sub-stereo-width/brief.md#Page structure sections).
Pearson correlation is calculated between the left and right sub-bass vectors. This correlation value describes how similar or different the two channels are in the sub region. The final width score is derived using the formula: width = 1.0 − sub_correlation (Source: inputs/articles/sub-stereo-width/brief.md#Key accuracy requirements).
When the left and right channels are identical in the sub region, correlation equals 1.0 and width equals 0.0. This represents perfect mono. When the channels have low or negative correlation in the sub region, width approaches 1.0. This represents significant stereo spread that will cause problems.
The 100Hz cutoff was chosen because it aligns with the psychoacoustic threshold below which humans cannot localize sound sources. Industry practice typically keeps everything below 100Hz mono, and often extends this to 200Hz for bass-heavy genres.
Interpreting sub-stereo-width values and outputs
Understanding the width scale allows you to make informed mixing decisions. Values between 0.0 and 0.05 indicate excellent mono compatibility. The sub region is effectively mono with negligible stereo difference. This is the target range for most commercial mixes.
Values between 0.05 and 0.15 are acceptable. There is minimal stereo spread in the sub region. Most playback systems will handle this without significant problems, though tighter control is preferred (Source: inputs/articles/sub-stereo-width/brief.md#Key accuracy requirements).
Values between 0.15 and 0.25 are problematic. Noticeable stereo difference exists in the sub region, creating phase risk on mono playback. Bass energy will be compromised on summed systems. Vinyl cutting may encounter difficulties. This range requires corrective action.
Values above 0.25 are critical. Significant sub stereo spread will cause audible problems. Mono playback will lose substantial bass energy. Club systems that sum bass to mono will suffer. Vinyl cutting will be difficult or impossible without additional processing.
The actionable threshold is typically set between 0.15 and 0.2. Above this point, sub stereo width represents a measurable problem regardless of how healthy the full-mix correlation appears (Source: inputs/articles/sub-stereo-width/brief.md#Key accuracy requirements).
| Width range | Status | Description | Action needed |
|---|---|---|---|
| 0.0–0.05 | Excellent | Sub is effectively mono | None |
| 0.05–0.15 | Acceptable | Minimal stereo spread | Monitor |
| 0.15–0.25 | Problematic | Noticeable stereo difference | Correct |
| Above 0.25 | Critical | Significant stereo spread | Immediate correction |
How sub-stereo-width integrates with other systems
Sub stereo width is included in the raw measurements block passed to AI coaching systems. High values are flagged as stereo and phase issues requiring correction. The metric typically triggers prescriptive fixes that specify crossover frequencies for mono summing (Source: inputs/articles/sub-stereo-width/brief.md#Page structure sections).
Common AI recommendations include mid-side processing with a high-pass filter on the side channel below 80 to 100Hz. This preserves stereo width in higher frequencies whilst ensuring sub-bass mono compatibility. Alternative recommendations may suggest applying a high-pass filter on the side channel below a specific frequency threshold determined by the severity of the issue.
The metric complements full-mix stereo correlation by providing targeted low-frequency analysis. Full-mix correlation measures the overall left-right relationship across all frequencies. Sub stereo width isolates just the low-frequency region. A mix can exhibit healthy full-mix correlation whilst hiding problematic sub stereo spread because wide energy concentrated in the low end may not significantly affect the overall correlation value. Both metrics are needed for complete stereo field assessment.
Sub stereo width also relates to tonal balance metrics, particularly the sub-band in five-band energy analysis. High sub stereo width may appear alongside sub-bass level issues, forming part of a comprehensive low-end assessment that ensures both appropriate energy levels and phase coherence.
Practical application and workflow
Use sub stereo width during the late mixing stage when finalising the low end. Check the metric after applying stereo bus processing, particularly mid-side EQ or stereo widening plugins. These tools often affect the sub region unintentionally.
Common causes of high sub stereo width include wide stereo bass synthesizers with sub-octave oscillators panned across the stereo field, stereo bus processing that applies wide mid-side EQ or stereo widening to the full spectrum, room microphone leakage capturing low-end rumble in stereo, stereo reverb returns that extend into the sub range, and excessive stereo widening from plugins affecting frequencies they should not touch (Source: inputs/articles/sub-stereo-width/brief.md#Page structure sections).
To correct high sub stereo width, apply mid-side processing with a high-pass filter on the side channel below 80 to 100Hz. This is the most common and effective fix. Ensure kick drum and bass fundamentals are mono from the source. Use linear-phase crossover EQ to sum sub frequencies to mono without introducing phase shifts elsewhere. Check that stereo widening plugins have high-frequency-only modes engaged. Remove sub frequencies from stereo effects sends by high-passing reverbs, delays, and widening plugins (Source: inputs/articles/sub-stereo-width/brief.md#Page structure sections).
Sub stereo width is critical for mono compatibility verification. Phone speakers, laptop speakers, and many club systems sum bass to mono. Bluetooth speakers often use mono bass processing to maximise loudness from small drivers. Vinyl cutting lathes require mono bass to prevent the stylus from jumping the groove. Testing on these systems will immediately reveal problems that the metric predicts.
Summary and key takeaways
Sub stereo width measures stereo difference in the sub-bass region below 100Hz, expressed as a width score from 0 (fully mono) to 1 (fully uncorrelated). Most commercial mixes keep the sub region tightly mono to preserve bass energy on mono systems and prevent phase cancellation. High values above 0.15 to 0.2 need immediate attention regardless of full-mix correlation health (Source: inputs/articles/sub-stereo-width/brief.md#Core message).
The metric reveals problems through targeted low-frequency analysis that full-mix correlation might miss. Sub-bass frequencies are omni-directional and non-localizable, meaning stereo information adds no perceived width but creates measurable problems on summed playback systems.
Common fixes include mid-side processing with high-pass filtering on the side channel, ensuring mono bass sources, and removing sub frequencies from stereo effects. The metric integrates with AI coaching systems to provide specific crossover frequencies for corrective processing.