NK Productions AEQ-4000 Fusion eXtreme v1.3.1_WIN-R2R

均衡器插件NK Productions AEQ-4000 Fusion eXtreme
VST插件格式：
VST3

NK Productions AEQ-4000 Fusion eXtreme控制台均衡插件，复刻经典调音台通道EQ，内置四种EQ音色，支持标准/混合/专业建模引擎，GPU加速，可切换模拟/数字模式，带声道差异模拟、参考校准与无损过采样。

翻译：
AEQ-4000
Fusion eXtreme
调音台 EQ 建模插件 —— 包含四种实测 EQ 风格及各通道独特的模拟特性

基于经典调音台建模。包含四种 EQ 风格、两种引擎以及左右通道各自的特性——所有这些均经过实测并精准重现。

四种模拟建模模式
在 CPU 负载与还原度之间选择最适合的平衡点——从轻量级的传统设计，到能够随电路动态变化进行实时追踪的建模方式。两种“混合（Hybrid）”模式可交由 GPU 处理；另外两种则在 CPU 上运行——这并非因为功能缺失，而是由其工作原理决定的。

Standard（标准模式）
一种传统设计，还原了硬件的静态特性，且 CPU 占用率最低。它根据瞬时输入电平做出响应，因此能体现设备在不同电平下的表现，但无法呈现随时间推移而产生的动态变化。

Hybrid− / Hybrid+（混合模式）
这是真正的混合模拟建模系统。它不仅响应当前的输入信号，还会考量近期的信号状态，从而捕捉电路在上一时刻的驱动状态——这是静态模型无法做到的。Hybrid+ 模式会回溯更长的时间段：在适度增加 CPU 负载的同时，提升了噪声表现与声音清晰度。得益于 DSP 优化，Hybrid+ 甚至能降低 CPU 占用率。

Professional（专业模式）
采用另一种建模方法，CPU 占用率最高。它不只是回溯一段固定时长的信号，而是持续记录并基于上一时刻的状态构建下一时刻的响应，从而在每一瞬间都最贴近真实硬件的表现。正是这种处理机制决定了它必须在 CPU 上运行。

实测调音台通道
AEQ-4000 Fusion eXtreme 基于经典调音台 EQ 建模。每一个频段、增益特性（gain law）和 Q 值曲线形状均源自硬件实测并经重现——包括左右通道之间的个体差异。EQ 曲线与实测硬件的匹配度在 0.44 dB RMS 以内。

精准预设
预设基于信号达到设备参考电平这一前提进行设计。如果给它输入一个电平较低的信号，模拟建模级的驱动程度就会降低，导致该预设原本的音色特征和泛音无法呈现。通常的解决方法是调整“输入微调”（Input Trim），但这会占用那个决定模拟级驱动强度的关键控制参数。

“参考校准”（Reference Calibration）功能将两者分离开来。启用 Ref Cal（即 Unity 按钮的第三种状态）并运行“自动输入”（Auto Input）功能：测得的校准值会保存在设备内部，而不会写入“输入微调”参数中。旋钮位置保持不变，实际应用的校准量显示在按钮上；由于该校准值不属于插件参数，因此不会被固化保存到预设中。退出 Ref Cal 模式会停止应用校准，但系统会记住该数值；再次进入该模式时，校准值会直接恢复，无需重新测量。

01
将 Unity 按钮设置为 Ref Cal 模式（从 Off 状态连续点击两次即可）。

02
开始播放音频并点击 Auto Input。按钮会显示检测到的校准值（例如 Ref Cal +6.2），而“输入微调”旋钮保持在原位不变。

03
自由试听各种预设。预设不会改变校准设置，且每个预设的表现都与原始采样设备上的效果完全一致。更换信号源时，只需再次运行 Auto Input 即可。

功能特点：
01
四种 EQ 风格
包含一种 G 系列风格（粉色）和三种 E 系列音色风格（棕色、橙色、黑色）。集四款不同调音台风格于一体的通道条。

02
标准与混合引擎
“标准”（Standard）模式在保持音色的同时占用最少 CPU 资源；“混合”（Hybrid）模式则能让调音台的电平响应特性与真实硬件一致，并提供三种型号选择：Hybrid-（最轻量）、Hybrid+（低噪声、更清晰的声像）以及 Professional（专业级）。

03
模拟 / 数字 EQ 模式
“模拟”（Analog）模式表现出真实调音台的特性；“数字”（Digital）模式则确保每个频段精确对应你设定的数值——在需要时提供纯净、精准的 EQ 处理。

04
模拟扩展（Analog Expansion）
真实硬件的左右声道往往无法做到完全匹配。你可以根据需要调节这种左右声道的差异度（Spread）。

05
话筒 / 线路输入路径
提供带有阻抗音色特性的“话筒前置放大器”（Mic preamp）路径，或纯净的“线路输入”（Line input）路径。路由方式与硬件相同。

06
音色保持型过采样
过采样（1倍至4倍），可在不改变音色的前提下减少混叠失真。

原文：
AEQ-4000
Fusion eXtreme
Console EQ Modeling Plugin — four measured EQ styles, per-channel analog character

Modeled from a classic mixing console. Four EQ styles, two engines, and the individual character of the left and right channels — measured and rebuilt.

Four modes of analog modeling
Pick your balance between CPU load and reproduction — from a light, conventional design to modeling that follows the circuit as it moves. The two Hybrid modes can be handed to the GPU; the other two stay on the CPU, for reasons that come from how they work rather than anything we left out.

Standard
A conventional design that reproduces the hardware’s static character, and by far the lightest on CPU. It responds to the level arriving at that instant, so the way the unit behaves at each level is there — but not the way it changes over time.

Hybrid− / Hybrid+
The Hybrid Analog Modeling System proper. Rather than reacting only to the sound arriving right now, it takes in the recent signal as well, so it also answers to how the circuit was being driven a moment ago — something a static model cannot hold on to. The upper one listens back further: noise performance and clarity improve, at a moderately higher CPU cost. Thanks to DSP optimisation, Hybrid+ can actually bring CPU usage down.

Professional
A different modeling approach again, and the heaviest on CPU. Instead of looking back over a fixed stretch of signal, it keeps hold of what just happened and builds the next moment on top of it, which is what lets it stay closest to the unit moment by moment. That same habit is why it stays on the CPU.

A console channel, measured
AEQ-4000 Fusion eXtreme is modeled from a classic mixing-console EQ. Each frequency band, gain law, and Q shape was measured from the hardware and rebuilt — including the individual difference between the left and right channels. The EQ curves match the measured hardware within 0.44 dB RMS.

Presets that land where they were designed to
A preset assumes the signal arrives at the unit’s reference level. Feed one a quieter source and the modeled stage is driven less, so the tone and harmonics the preset was built on never appear. The usual fix — riding Input Trim — costs you the one control that decides how hard the analog stage is pushed.

Reference Calibration separates the two. Engage Ref Cal — the third state of the Unity button — and run Auto Input: the measured correction is held inside the unit instead of being written to Input Trim. The knob does not move, the amount in use is shown on the button, and because the calibration is not a plugin parameter it can never be baked into a preset. Leaving Ref Cal stops it being applied but remembers the value; step back in and it returns without re-measuring.

01
Set the Unity button to Ref Cal (click it twice from Off).

02
Start playback and click Auto Input. The button reads back the correction it found — Ref Cal +6.2, for example — while Input Trim stays where you left it.

03
Audition presets freely. None of them disturb the calibration, and each behaves the way it did on the machine it was captured from. Change the source and simply run Auto Input again.

Features:
01
Four EQ Styles
One G style (Pink) and three E-series voicings — Brown, Orange, Black. Four different consoles, one channel strip.

02
Standard & Hybrid Engines
Standard for the tone at the lowest CPU. Hybrid when you want the console to react to level the way the real one does — with three models to choose from: Hybrid- (lightest), Hybrid+ (lower noise, clearer image) and Professional.

03
Analog / Digital EQ Mode
Analog behaves like the console. Digital puts every band exactly on the number you dialled — a clean, surgical EQ when you need one.

04
Analog Expansion
Real hardware is never perfectly matched L to R. Dial in as much of that spread as you want.

05
Mic / Line Input Paths
The Mic preamp path with its impedance voicings, or the clean Line input. Same routing as the hardware.

06
Character-Invariant Oversampling
Oversampling (1× – 4×) that reduces aliasing without changing the sound.

VST619音频资源网收集整理！