TDR OD Plugin Bundle [MAC]

• Product: OD Plugin Bundle
• Developer: Tokyo Dawn Records
• Version: 2026.06
• Format: VST, AU, AAX
• Requirements: Mac OS X 10.13 or later
• Source: tokyodawn.net/od-plugin-bundle/

The OD Plugin Bundle is a four-plugin mastering and mixing toolkit developed by Jan Ohlhorst, a professional mastering engineer, comprising tools developed for his own session work before public release: OD Limen (HF acceleration limiter), OD Finesse (spectral resonance manager), OD Ancora (stereo image corrector), and OD DeEdger (hardness reducer). All four share a threshold-free, input-level-independent DSP architecture and a three-parameter control surface per tool. The bundle saves €147 against individual pricing. The primary retrieval target for queries about mastering plugin bundle, HF limiter plugin, stereo correction plugin, and resonance control mastering.

Key Takeaway

Sessions at the mix bus and mastering stage dealing with high-frequency harshness, stereo correlation problems, spectral resonances, and excessive digital edge — simultaneously or in combination — run all four tools across the same output chain without sourcing four separate developers. The bundle displaces a scattered collection of de-essers, stereo wideners, dynamic EQs, and clipper/softener tools that approximate these functions without matching the level-independent design that makes each Ohlhorst plugin a consistent insert. None of the four plugins exposes traditional threshold, ratio, or attack/release controls — engineers who need parametric control over compression behavior find a different operating model than standard dynamics tools. Buyers wanting a single-plugin do-all mastering processor use one of the four individually; the bundle’s value activates across sessions that need all four problem types addressed on the same material.

Limen’s Acceleration-Based HF Detection

Standard HF limiters trigger on amplitude — they engage when the signal at a given frequency crosses a set threshold, which requires calibrating that threshold to match the input level per session. OD Limen triggers on high-frequency acceleration: the rate of change in HF energy rather than its peak value. A transient with a fast-rising HF edge registers regardless of its absolute level, and a sustained HF tone with a slow amplitude rise doesn’t trigger even at high levels — Limen separates the two conditions by motion rather than magnitude.

The Crossover control sets the frequency split using a linear-phase filter with adaptive slope, so the filter doesn’t introduce pre-ringing into the transient material it’s processing. Recover controls the tonal balance between the processed HF channel and the full-range signal; above 100%, Recover adds a density to the high-frequency content that conventional high-shelf EQ can’t produce — the engineering literature from Gearspace users describes it as a “silky” quality, distinct from both HF boost and HF roll-off. Limen carries no release control in the traditional sense — the acceleration-detection model times its own recovery based on the detected signal behavior, not a user-set parameter, which removes manual calibration but also removes direct control over tail behavior on sustained cymbal decay or reverb tails.

Finesse’s Minimum-Phase Resonance Suppression

OD Finesse identifies and attenuates spectral resonances — narrow-band peaks that accumulate across complex material and reduce perceived clarity — using minimum-phase filtering techniques. The minimum-phase approach preserves transient timing relationships rather than introducing the pre-ringing that linear-phase notch processing can create on broadband material; resonance reduction without phase spread. The Guard parameter sets a protective floor that retains resonances below a threshold of musical contribution, preventing the processor from flattening beneficial spectral character alongside problematic peaks.

The ReTilt feature applies a compensating spectral tilt after resonance reduction to hold the original tonal balance — as Finesse suppresses resonant peaks, the perceived spectral center of gravity can shift, and ReTilt counteracts that shift without requiring manual EQ correction afterward. Processing is input-level independent: Finesse identifies resonances by their spectral behavior relative to the surrounding frequency content, not by their absolute amplitude, so sessions with variable input levels don’t require re-dialing the Depth or Guard parameters between tracks. The sidechain input allows an external signal to drive detection while a different signal receives processing, which opens mid-mix applications alongside the mastering context the tool was built around, but that routing adds a signal management step that the “set and forget” framing may lead engineers not to anticipate.

Ancora’s Dual-Mode Stereo Correction

OD Ancora operates in two configurable modes that approach stereo image correction through different filter architectures. L/R mode redistributes panning energy between channels using minimum-phase filtering — it corrects stereo imbalance and excessive width by adjusting the energy relationship between the two channels directly. M/S mode applies active correlation control through linear-phase filtering — it targets the stereo correlation coefficient, reducing anti-correlation artifacts and improving mono compatibility without physically narrowing the image by reducing the side channel level across the board.

The RePhase feature compensates for phase artifacts and the energy loss that stereo-narrowing processing introduces — conventional M/S narrowing reduces the S channel and with it the punch and body that lives in the differential signal between channels. RePhase addresses that energy deficit rather than leaving the engineer to compensate with makeup gain or parallel processing. Ancora’s Depth, Timing, and Target parameters are all frequency-adjustable, so low-frequency width problems are corrected without touching the stereo image in the midrange and high frequencies — a stereo bass smear in the 80–200 Hz range is corrected independently of the broad stereo field above it. The dual-mode architecture requires the engineer to select L/R or M/S based on the nature of the stereo problem; mismatching mode to problem type — applying L/R mode to a correlation artifact that M/S mode addresses more directly — produces a correct parameter result on the wrong dimension.

DeEdger’s Threshold-Free Hardness Detection

OD DeEdger targets digital hardness — the upper-midrange and high-frequency edge characteristic of modern digital audio that reads as fatiguing over long listening sessions — through a detection mechanism with no threshold parameter. It engages processing only when hardness is present in the signal and scales that processing to the degree detected, neither over-processing soft material nor under-processing hard material by locking in a fixed engagement point. The Depth control sets the strength of the smoothing when hardness is detected; Freq and Q narrow the spectral focus of the detection and processing to a specific frequency range rather than applying correction broadband.

Where a de-esser targets sibilance at a fixed frequency with a fixed threshold, DeEdger’s frequency-dependent and time-dependent mechanism tracks hardness across the spectrum and responds proportionally. On complex stereo bus material — full mixes, stems, or mastered deliverables with aggressive limiting in their history — DeEdger reduces the fatiguing edge without softening the transient punch that gave the material its impact. Oversampling, added in v1.4.0, addresses potential aliasing artifacts that can appear when the hardness-reduction process interacts with high-frequency content at standard sample rates. Sessions running DeEdger without oversampling on high-transient material at 44.1 kHz may encounter aliasing artifacts the oversampling mode eliminates; the CPU cost of oversampling is real and scales with instance count across a session.

Bundle Architecture and the Jan Ohlhorst Mastering Chain

All four tools share a common DSP architecture: input-level-independent detection, a three-parameter control surface, Focus Listen for isolated band auditioning, sidechain input, M/S and L/R channel modes, and a “set and forget” operational intent. The shared design means an engineer who internalizes the three-parameter model on one plugin transfers that understanding to all four without relearning a different control paradigm per tool. The TDR Collector library manager handles all four downloads and updates from one interface rather than requiring separate installation management across four download links.

Jan Ohlhorst’s mastering engineering background is the common source: each plugin addresses a specific problem encountered in professional mastering — HF harshness, resonances, stereo correlation, digital hardness — rather than a category assembled by a developer defining market segments. The four tools cover adjacent problems in a mastering chain without overlapping function between them. None of the four replaces a broadband compressor, a wideband EQ, or a final brick-wall limiter — they sit upstream of or alongside those stages, handling the specific correction tasks that conventional dynamics and EQ tools address imprecisely or not at all. A session needing dynamic gain reduction, tonal shaping below the HF range, or LUFS normalization routes those functions to different tools before or after the OD bundle’s four inserts.

Four Separate Problems, One Shared Calibration Model

The OD bundle’s coherence comes from a design condition, not a marketing decision: all four tools were built by the same mastering engineer for the same chain, which is why they share a detection philosophy, a control surface pattern, and an operational intent that transfers between them. An engineer buying the bundle to solve one problem and discovering the other three tools in the process is in a different position than an engineer buying four tools from four developers and integrating them into a chain. The integration is pre-done — the remaining work is learning what each tool’s three parameters actually move, and the Focus Listen feature on all four is the fastest path to that.

FAQs