The Challenge
When I became involved with LONO Audio, the technology existed as a complex chain of commercial plug-ins inside a professional Digital Audio Workstation (DAW). It could produce impressive results, but it required extensive human intervention and depended on third-party products.
I developed a proprietary C++ implementation that transformed the process into an integrated, substantially automated digital signal-processing system.
This was not a matter of reproducing the plug-in settings. I first had to determine what the original process was accomplishing perceptually. I could then implement those concepts directly, improve them, and invent new algorithms for effects that commercial audio tools could not produce.
Revisiting Recordings Mastered for Vinyl
Many older recordings were engineered and mastered for vinyl, a physical medium with constraints that do not apply to digital playback.
Powerful low-frequency signals require wide groove excursions. Excessive bass can reduce the amount of music that fits on a record and can make the groove difficult for a stylus to track. Out-of-phase stereo bass is especially troublesome because it produces vertical stylus movement. Mastering engineers therefore commonly reduced the lowest frequencies or narrowed them toward mono.
High frequencies presented a different problem. The standard RIAA process reduced bass and boosted treble while cutting the record, then reversed that equalization during playback. But excessive sibilance and other sharp high-frequency energy could still produce rapid groove modulations that were difficult to track without distortion. Engineers often de-essed or otherwise restrained problematic material. High-frequency reproduction also deteriorated toward the inside of a record as the linear speed of the groove beneath the stylus decreased.
Vinyl also imposed tradeoffs among dynamics, playing time, bass response, stereo width, and overall level. These were sensible engineering decisions for the intended medium, but the compromises remained embedded in the stereo master when the same recording was later distributed digitally.
LONO could revisit a finished recording without being bound by the physical limitations of vinyl. The objective was not merely to add bass and treble. It was to improve detail, imaging, and perceived power while preserving the recording’s tonal character and musical dynamics.
Building the Processing System
I implemented the LONO processing chain in C++, using the JUCE audio-development framework for standard processing components. These included dynamic multiband equalization, compression, saturation, de-essing, limiting, and related mastering functions.
Those tools were only the foundation.
Some of the soundstage processing was proprietary, and all of the transient-detail processing required original software. A patent application was filed covering aspects of the technology.
The system addressed three interrelated DSP problems.
Reshaping a Finished Stereo Soundstage
During mixing, an engineer can position individual tracks across the stereo field. Once those tracks have been combined into a two-channel master, their independent controls are gone.
The stereo signal still contains information about the instruments, however. Different instruments occupy different frequency ranges and exhibit different harmonic and timbral characteristics. I developed processing that used those characteristics to affect elements of the recording differently.
The software did not literally recover the original tracks or assign each instrument a new coordinate. It altered the listener’s perception of separation and placement. Instruments that had seemed crowded together could become more distinct, producing a clearer and wider apparent soundstage.
Increasing Detail Without Adding Harshness
A musical sound consists of more than its sustained note. Its transient (the short-lived attack at the beginning of the sound) helps the listener perceive how the note was produced.
A guitar pick striking a string, a drumstick contacting a cymbal, or a piano hammer reaching a string may last only an instant. Conventional EQ can emphasize the frequencies associated with those events, but it also emphasizes sustained energy at the same frequencies. That can make a recording brittle, harsh, or fatiguing.
LONO’s proprietary transient processing worked selectively across the frequency spectrum. It could enhance momentary details without simply raising the corresponding frequencies throughout the recording.
This distinction was important. A de-esser might reduce sustained harshness in a particular region while the transient system separately enhanced brief percussive information in that same region. The audible result could contain more detail without sounding brighter merely for the sake of brightness.
Increasing Perceived Power Without “Bricking” the Music
A common way to make music sound louder is to compress or limit it aggressively. This reduces the distance between quieter passages and louder peaks. Taken too far, the waveform becomes a nearly solid block, often called “bricked” or “sausaged” audio.
That approach sacrifices transient detail and musical dynamics.
LONO treated perceived loudness as distinct from simple amplitude. Its processing could make musical gestures more apparent while retaining the brief peaks that give those gestures definition.
The middle waveform may look more active, but it has not been flattened like the third. Its additional peaks represent transient events that remain distinguishable. The over-limited waveform may be consistently loud, but many of the short-lived details have been erased.
Converting a Technique into Software
The original DAW implementation depended on a collection of commercial tools operated by a person. My implementation replaced that workflow with a purpose-built processing system.
I wrote the processing software, integrated the conventional and proprietary stages, and automated the sequence so that it could operate repeatably across different recordings. I also created analytical comparisons between LONO, unprocessed originals, and other mastering technologies.
Those comparisons were essential because louder does not necessarily mean better. Evaluation required level-aware listening and visual analysis of waveforms, transient behavior, tonal balance, stereo imaging, and dynamic range.
The result was not a digital copy of the original plug-in chain. Some concepts that had only been approximated in the DAW became fully realizable through proprietary software.
Validation
Conductor Gerard Schwarz described hearing greater clarity and beauty in orchestral and chamber recordings processed with LONO. What would he know? Former Director of Seattle Symphony Orchestra, 4 Emmy Awards and 14 Grammy nominations.
Producer Chris Young (Izzy Gold Records) specifically praised its transient response and the natural resonance it achieved without relying on compression.
At Universal Music Group, Pat Kraus (SVP Archives and Recording Studios) brought the technology to Kevin Reeves (VP Mastering) for scrutiny. After hours of analysis in the finest mastering studio, his wonderfully understated verdict was “It doesn’t suck.” We immortalized the testimonial on T-Shirts, and he sent us a photo of his.
