Budgie Harev1.0.4 · USER MANUAL · 用户说明书

TWENTY-FOUR TERMS. YOUR OWN SOUND.

让声音,
有它的季节。
Give your sound
its season.

从第一次载入,到节气 EQ、天气与细节参数。理解每个入口,让空间、颗粒与自然声跟上你的演奏。

From the first insert to seasonal EQ, weather and detailed shaping. Learn each control and let space, grains and nature follow your playing.

macOS 13+ · Apple Silicon / Intel · AU / VST3 / AAX

macOS 13+ · Apple Silicon / Intel · AU / VST3 / AAX

01

60 秒上手Quick start

  1. 在有声音的音频轨道或乐器轨道效果器插槽中载入 Budgie Hare。
  2. 播放或演奏,先听当前节气;从较低的混合量开始。
  3. 用顶部箭头、节气名称菜单或左下节气旋钮切换预设。
  4. 调整生长、光泽与颗粒,再用 XY 决定空间与质感的分量。
  5. 打开细节 → 质感,比较染色 0 与 50;将 Mix 提到 70% 以上,更容易听清季节音色。
  6. 保存为「我的预设」,再保存 DAW 工程。
  1. Insert Budgie Hare in an audio-effect slot on an audio or instrument track.
  2. Play audio and audition the current term, starting with a modest Mix.
  3. Change terms with the header arrows, name menu or bottom-left term dial.
  4. Shape Growth, Shine and Grains, then balance space and texture with XY.
  5. Open Detail → Texture. Compare Colour at 0 and 50; Mix above 70% makes seasonal colour easier to hear.
  6. Save a user preset and your DAW session.
02

界面与信号流Interface and signal flow

Budgie Harev1.0‹ 01 立春 · 东风解冻 ›我的预设 A / B 复制 粘贴⚙晴立春 · 东风解冻琴弦日光鸟鸣风自然声跟随回声空间输入音量0.0 dB输出音量0.0 dB音高CC♯DD♯EFF♯GG♯AA♯B节气生长光泽颗粒混合空间 / 质感细节 ⌄
图 01 · 主界面矢量参考

矢量控件参考 · 数值为演示值 · 点击可放大

顶部依次是节气选择、我的预设、A/B、复制/粘贴、闪避与设置。画面中的大字是天气按钮;右侧是层次与输入/输出音量;下方是四个主旋钮、XY 与细节开关。

输入 → 原声与效果纹理按 Mix 混合 → 季节染色 → 与自然声层共同输出

颗粒切自输入,回声重复输入,空间承接纹理;自然声来自录音。染色作用于原声与效果纹理的合成信号,自然声录音不经过这层染色。不要把输入静音后的自然声误认为插件凭空生成了乐器。

主声音路径与自然声路径,在输出前相遇输入原声 + 纹理Mix → 染色输出录音 → 跟随 / 应答两条路径,在输出前相遇输入原声 + 纹理Mix → 染色录音跟随 / 应答输出
图 02 · 信号流向:染色不作用于自然声录音

矢量原理图 · 说明声音关系,不代表实测波形 · 点击可放大。

Budgie Harev1.0‹ 01 Start of Spring ›My presets A / B Copy Paste⚙晴CLEARStringsSunlightChirpWindNatureFollowEchoSpaceInput0.0 dBOutput0.0 dBPitchCC♯DD♯EFF♯GG♯AA♯BSolar termGrowthShineGrainsMixSpace / TextureDetail ⌄
Figure 01 · Vector interface reference

Vector control reference · illustrative settings · click to enlarge

The header contains term selection, user presets, A/B, copy/paste, Ducking and Settings. The large character in the picture changes weather. Layer and input/output controls sit on the right; four main knobs, XY and Detail sit below.

Input → source/texture Mix → seasonal Colour → output with nature layer

Grains use the input, echoes repeat it and the space receives the texture. Nature uses recordings. Colour processes the combined source and texture; nature recordings bypass this stage.

Source/texture and nature meet at the outputInputSource + textureMix → ColourOutputRecordings → Follow / ReplyTwo paths meet at the outputInputSource + textureMix → ColourRecordingsFollow / ReplyOutput
Figure 02 · Signal flow: nature bypasses Colour

Vector concept diagram · relationships, not measured audio · click to enlarge.

03

WET 发送模式与自然声退出WET send mode and nature release

Hare 1.0.4 在 Mix 旁提供 WET「发送模式」按钮。把 Hare 放在 DAW 的辅助/返回轨道上,用发送把原轨的声音送来,再由辅助轨推子决定整体效果量。

  1. 打开 WET,效果引擎固定使用 Mix 100%,不再把原轨的干声重复送入返回轨;关闭后恢复 Mix 旋钮原先保存的数值。
  2. 开启时,若闪避低于 40%,会先抬到 40%,为原声留出空间。关闭时,如果你没有手动改过闪避,会恢复开启前的数值;手动调整过则保留你的选择。
  3. 有输入时,自然声维持节气和当前自然声设置决定的音量,即使轻弹也不会因为输入变小而额外变弱。输入停止后保持 3 秒,再平滑淡出,包含自然声自身的空间尾音;下次输入到来时平滑返回。

重要:输入只决定自然声何时进入或退出,不按输入响度缩放自然声。发送推子主要改变送入的演奏及其效果;要改变自然声层本身,请用「自然声」旋钮;要同时改变整个返回效果,请用 DAW 的辅助轨推子。WET 关闭时,自然声保持原先的常驻层行为。低于约 −80 dBFS 的输入按静音处理。

WET:演奏停止后的自然声输入自然声输入停止保持 3 秒平滑淡出示意包络;录音本身仍有自然起伏,不是固定电平的测试信号。
WET 发送模式:输入停止、保持与淡出

矢量控件参考 · 数值为演示值 · 点击可放大

WET 的主要用途是辅助返回轨。用作普通插入效果时,通常先保持 WET 关闭,通过 Mix 混合原声与效果;旧工程默认不会自动开启 WET。

同时建议更新到 HUB 0.8.10:试用转为正式授权后会立即刷新本机许可证;从插件打开授权或云端预设入口会复用已有 HUB 窗口。

Hare 1.0.4 adds the WET send-mode button beside Mix. Insert Hare on an aux/return track, send your source to it and use the return fader to set the overall effect level.

  1. WET fixes the engine at Mix 100%, keeping duplicate dry audio out of the return. Switching it off restores the Mix knob’s saved value.
  2. When enabled, Ducking is raised to 40% if it was lower. Switching WET off restores the previous value unless you adjusted Ducking yourself; your manual adjustment is retained.
  3. While input is present, nature retains the level set by the term and your Nature controls, even with very soft playing. After input stops, nature holds for 3 seconds, then fades smoothly, including its own room tail. It returns smoothly with the next input.

Important:Input controls when nature enters and leaves, not its loudness. The send fader mainly changes the incoming performance and its effects. Use Nature to adjust the nature layer, or the DAW’s aux/return fader to change the entire return. With WET off, nature keeps its previous ambient-bed behavior. Input below approximately −80 dBFS is treated as silence.

WET: nature after the input stopsInputNatureInput stopsHold for 3 secondsFade smoothlySchematic envelope; the recordings retain their natural level variations.
WET send mode: stop, hold and release

Vector control reference · illustrative settings · click to enlarge

WET is primarily for aux returns. For a normal insert, leave WET off and blend source and effects with Mix. Older sessions do not automatically enable WET.

HUB 0.8.10 is recommended: upgrading a trial to a full license refreshes the local certificate immediately, and licensing/cloud links from the plug-in reuse the existing HUB window.

04

安装与 DAW 使用Installation and DAW use

当前支持 macOS 13 或更高版本、Apple Silicon 与 Intel,插件格式为 AU / VST3 / AAX。Hare 是普通音频效果器,不需要 MIDI 轨道或 MIDI 输入;音高从音频分析取得。

  • Logic Pro:在轨道的 Audio FX 插槽选择 Budgie Hare。
  • Ableton Live、Cubase、Nuendo、REAPER:在音频效果器位置载入 AU 或 VST3,具体可用格式取决于宿主。
  • 插件不产生测试音;先确认轨道输入、乐器输出和监听正常。
  • 旧版 MIDI 控制 AU 的工程实例可能无法自动替换为当前普通效果 AU。保留原工程副本,重新插入当前效果器并检查设置。

从官网或 Budgie Hub 下载 Hare 1.0.4 安装包。安装时可选择 AU、VST3、AAX;Pro Tools 使用 AAX。安装后重启 DAW 并重新扫描插件。

Requires macOS 13 or later, Apple Silicon or Intel. Formats: AU / VST3 / AAX. Hare is a regular audio effect. It needs no MIDI input; pitch information comes from audio analysis.

  • Logic Pro: insert Budgie Hare in an Audio FX slot.
  • Live, Cubase, Nuendo and REAPER: load an AU or VST3 in an audio-effect slot, as supported by the host.
  • Hare does not generate a test instrument. Check the track input, instrument and monitoring first.
  • Sessions using the older MIDI-controlled AU may need the current audio effect inserted again. Keep a session backup and check recalled settings.

Download Hare 1.0.4 from the website or Budgie Hub. Select AU, VST3 and/or AAX; Pro Tools uses AAX. Restart your DAW and rescan after installation.

05

四个旋钮、XY 与层次Main controls and layers

控制怎么用
生长围绕当前预设自己的起点塑造纹理形态。向低处更稀疏,向高处更繁茂,连动颗粒密度、长度、反向比例、回声反馈、泛音和珠落;不是第二个音量旋钮。
光泽增加纹理的闪光与泛音感,画面色彩也随之改变。
颗粒调整输入切片形成的颗粒层,画面水墨质感也随之改变。
混合 Mix改变原声与纹理的比例,同时决定季节染色的介入程度。响度经过校准和补偿,实际结果仍需用耳朵与电平表判断。
XY横向调整空间/回声,纵向调整光泽/颗粒。双击回到中间的预设基准位置。

右侧琴弦、日光、鸟鸣、风、自然声与回声用于平衡各层;「空间」控制混响时间(0.5–20 秒)。自然声总量与鸟鸣/风分层需要共同调整;鸟鸣控制在夏季也涉及虫声,并不保证每个预设都有鸟。

输入/输出音量范围为 −24 至 +12 dB,双击归零。旁通对比时留意输出电平,避免把更响误认为更好。

ControlUse
GrowthShapes texture around the preset’s own starting position. Lower is sparse, higher lush: density, length, reverse share, echo feedback, shimmer and pearls move together. It is not a second volume knob.
ShineRaises sparkle and overtone character; the picture’s colour responds too.
GrainsSets the layer of input-derived slices and the picture’s ink texture.
MixBalances source and texture and controls how strongly seasonal Colour enters. Calibration and level matching help, but listen and watch the meters.
XYHorizontal: space/echo. Vertical: shine/grains. Double-click to return to the central preset reference.

The right-hand layers balance strings, sunlight, birds/insects, wind, nature and echo. Space sets decay time (0.5–20 s). Nature and its bird/wind layers work together; not every preset contains birds.

Input/output gains span −24 to +12 dB; double-click resets to unity. Match output levels when comparing bypass.

06

每个节气的默认 EQ 与染色Seasonal EQ and Colour

染色:先由 Mix 决定介入量,再由染色推子限制强度0%10%70%100%目标染色为零逐渐介入达到推子允许的程度曲线不表示输出响度;输入频谱还会影响自适应结果Mix 决定染色介入量染色推子决定最高程度01070100%Mix不是输出响度曲线
图 05 · Mix 与季节染色的关系

矢量原理图 · 说明声音关系,不代表实测波形 · 点击可放大。

每个节气都有一套内置的默认音色轮廓。它不是可以逐段编辑的五段均衡器:切换节气时轮廓随之变化,用户通过「细节 → 质感 → 染色」控制整体程度。

  • 五个宽频段:120 Hz 低架、300 Hz、1 kHz、3 kHz,以及 9 kHz 高架;此外有少量二、三次谐波。
  • 春季趋向清亮,夏季趋向生动,秋季偏温暖,冬季更厚重。相邻节气之间连续插值,不是四套完全相同的季节 EQ。
  • 自适应会观察输入的频谱倾向。如果输入本来已经很亮或很暖,就减轻同方向的处理,避免一味叠加。
  • Mix ≤10% 时目标染色量为零;10–70% 之间逐渐介入;≥70% 时达到当前染色推子允许的程度。转动后有平滑过渡。
  • 染色推子初始值为 50%。切换节气载入新的曲线,但当前实现不会把染色推子强制重置到 50%;用户预设与工程会保存它。
想听清区别:固定输入与节气,将 Mix 设到约 80%,切换 A/B 比较染色 0 与 50,再比较春分与冬至。请用同一段音频持续播放,给自适应与响度补偿一点稳定时间。

染色、明暗、空间滤波有什么区别?

染色改变原声与纹理的整体音色,并带自适应与轻微谐波;明暗倾斜效果纹理的明暗平衡;空间低频/高频决定混响路径的频谱;自然声低切/中频/高切只整理自然录音层。想保留原声音色时,先把染色降到 0,而不是只移动明暗。

Colour: Mix engagement × Colour amount0%10%70%100%Zero targetRises graduallyFader-limited maximumThis is not output level; input balance also affects adaptationMix engages ColourColour limits the maximum01070100%MixNot an output-level curve
Figure 03 · How Mix engages seasonal Colour

Vector concept diagram · relationships, not measured audio · click to enlarge.

Every term has a built-in tonal profile. This is not a five-band EQ you edit band by band. Selecting a term loads its profile; Detail → Texture → Colour controls its overall strength.

  • Five broad bands: 120 Hz low shelf, 300 Hz, 1 kHz, 3 kHz and 9 kHz high shelf, plus a small share of second/third harmonics.
  • Spring is clearer, summer livelier, autumn warmer and winter deeper. Profiles interpolate across terms rather than repeating one EQ throughout each season.
  • Adaptation observes the input balance and softens processing where the input already leans in the same direction.
  • At Mix ≤10%, target colour is zero. It rises between 10–70%; above 70%, the Colour fader sets the available strength. Changes are smoothed.
  • The initial Colour value is 50%. Changing terms changes the profile but does not force this fader back to 50. User presets and sessions save it.
To compare: loop the same audio, set Mix near 80%, and use A/B for Colour 0 versus 50. Compare Spring Equinox and Winter Solstice. Allow the adaptive balance and level matching to settle.

Colour versus Tone and filters

Colour shapes source plus texture with adaptive EQ and subtle harmonics. Tone tilts the effect texture. Space filters shape the reverb path. Nature filters affect only recordings. Set Colour to 0 when you want to remove seasonal colouring of the source.

07

二十四节气与天气Terms and weather

切换节气会载入相应的声音设计与默认天气,也会清掉已选择的灵感配方。先保存手动调整,再换节气。天气按钮按晴 → 雨 → 雪 → 风循环,手选的天气持续到下次切换节气。

默认天气节气
雨雨水、清明、谷雨
雪小雪、大雪
风立秋
晴其余节气

雨随演奏力度加重,并加强珠落;雪使高频柔和、空隙中的空间更容易凝住;风加强阵风和声场运动。天气不仅是画面。当前实现避免天气录音与同类节气底层同时重复叠加。

春分、大暑、秋分默认采用反向空间;立冬、冬至、大寒默认冻结。冻结并非无条件永久锁住,后续演奏会改变空间状态。

A term change loads its sound design and default weather, and clears the selected Inspired recipe. Save manual edits before changing terms. The picture button cycles Clear → Rain → Snow → Wind; your weather choice lasts until the next term change.

DefaultTerms
RainRain Water, Clear and Bright, Grain Rain
SnowMinor Snow, Major Snow
WindAutumn Begins
ClearAll others

Rain grows with playing intensity and adds pearls. Snow softens the top and holds the space in gaps. Wind adds gusts and field movement. Weather changes sound as well as the picture; equivalent recording beds are not doubled.

Spring Equinox, Major Heat and Autumn Equinox use reverse space. Winter Begins, Winter Solstice and Major Cold use freeze. Freeze is responsive to subsequent playing rather than a permanent latch.

08

跟随、应答与闪避Follow, Reply and Ducking

同一个乐句:三种不同的处理对象演奏 / 输入跟随:自然声贴着演奏起落闪避:纹理为演奏让位跟随与闪避:对象不同演奏 / 输入跟随:自然声随演奏起落闪避:纹理为演奏让位
图 06 · 跟随与闪避:自然声包络与纹理避让

矢量原理图 · 说明声音关系,不代表实测波形 · 点击可放大。

控制含义
跟随自然声跟随输入包络的程度。0 是稳定底层;100 时不演奏就安静。雨的力度变化仍受演奏影响,不等于把所有天气都锁成恒定音量。
应答乐句结束后,录音事件与环境层可能回应。它有触发条件和随机性,不保证每个音之后都响,也不用于产生合成鸟叫。
闪避顶部伞形控件:演奏时压低效果纹理,为输入让位。上下拖动调节,双击归零;它与自然声跟随是两个控制。

只想听空间与颗粒:先降低自然声,并检查鸟鸣、风及天气设置。没有成功载入录音时,自然声静音,不会自动替换成合成自然声。

One phrase, three different targetsPlaying / inputFollow: nature follows inputDuck: texture yields to inputFollow and Duck: different pathsPlaying / inputFollow: nature traces playingDuck: texture yields to input
Figure 04 · Follow versus Ducking: nature and texture

Vector concept diagram · relationships, not measured audio · click to enlarge.

ControlMeaning
FollowHow strongly nature follows the input envelope. At 0 it forms a steady bed; at 100 it falls silent without playing. Rain intensity can still respond to playing.
ReplyRecorded events and ambience may answer after phrases. Trigger conditions and probability apply; it is not guaranteed after every note.
DuckingThe umbrella control lowers effect texture while you play. Drag vertically; double-click to reset. It is separate from nature Follow.

For space and grains alone, lower Nature and check Birds, Wind and weather too. Missing nature recordings stay silent; no synthetic substitutes are used.

09

细节抽屉:从入口到七组控制Detail drawer: seven control groups

点击底部「细节」展开。以下各组配有局部界面参考图和逐项解释;颗粒、碎响与琶音另有原理图和试听步骤。参考图用于认位置,正文说明实际作用;点击图可放大。

Click Detail at the bottom to expand the drawer. Each group has a clean panel reference and individual explanations. Grains, stutters and arpeggios also have conceptual diagrams and listening exercises. Click an image to enlarge.

10

回声:重复、反馈与珠落Echo: repeats, feedback and pearls

回声同步毫秒450 ms时间45%反馈6.0 kHz色彩30%珠落70%叠加+12 st音程回声同步毫秒450 ms时间45%反馈6.0 kHz色彩30%珠落70%叠加+12 st音程
图 07 · 回声控制区

矢量控件参考 · 数值为演示值 · 点击可放大

同步 / 毫秒

同步按 DAW 的拍子安排重复;毫秒按绝对时间。同步适合希望回声和鼓点对齐的段落,毫秒适合自由的空间节奏。宿主停播或未提供速度时,先确认宿主时钟,不要把同步当成自动侦测节奏。

拍值 / 时间

拍值是两次回声之间占多少拍,例如 120 BPM 下 1/4 为 500 ms,1/8 为 250 ms。数字越小的分母不一定代表更快:1/4 比 1/8 更慢。附点延长到 1.5 倍,三连音 T 为同拍值的 2/3;自由时间范围 0–2000 ms。

反馈

把一部分回声重新送回延迟,让重复继续发生。提高后尾巴更长、更容易积累,降低后更快消失。它不同于右侧回声音量:一个决定持续多久,一个决定这一层有多明显。先弹一个短音再停下,最容易听清反馈。

色彩

控制重复声的频谱倾向。向低频率调整通常更暗、更远;提高频率保留更多亮度。原声已经尖锐时,先把回声调暗,让重复融入背景,而不是把整条轨道削暗。

珠落

从输入中取不规则的短片段,形成像珠子散落一样的密集回响。它不是另加一套水滴乐器。先降低常规回声与自然声,再从小量增加珠落,听它带来的细碎动态;太多会遮住快速乐句。

叠加 / 音程

叠加决定回声中移调部分的份额,音程决定它相对输入升降多少。同度保持原音高,八度相差 12 半音,五度为 7 半音;手动四度等组合可能与和声冲突。随机会在同度、八度、五度、低八度之间变化。先固定音程,再增加叠加,方便判断变化来自哪里。

EchoSyncms450 msTime45%Feedback6.0 kHzTone30%Pearls70%Layer+12 stIntervalEchoSyncms450 msTime45%Feedback6.0 kHzTone30%Pearls70%Layer+12 stInterval
Figure 05 · Echo controls

Vector control reference · illustrative settings · click to enlarge

Sync / milliseconds

Sync uses the DAW beat; milliseconds use absolute time. Use sync for repeats aligned to drums and ms for free timing. Check the host clock if transport/tempo information is unavailable.

Division / time

A division sets the spacing between repeats: at 120 BPM, 1/4 is 500 ms and 1/8 is 250 ms. Dotted values last 1.5× as long; triplets (T) last 2/3. Free time spans 0–2000 ms.

Feedback

Returns part of each echo to the delay so it repeats again. More feedback sustains and accumulates; less dies sooner. Echo level sets audibility, not duration. Play one short note, stop and listen to the repeats.

Tone

Shapes the repeats’ spectrum. Lower frequency generally sounds darker and more distant; higher preserves brightness. Darken sharp repeats before darkening the entire source track.

Pearls

Irregular short slices of the input make scattered, dense reflections. This adds no separate water-drop instrument. Lower normal echo/nature, then raise Pearls gently and listen for fine movement; too much can obscure fast phrases.

Layer / interval

Layer sets the shifted share; Interval sets its transposition. Unison retains pitch, an octave is 12 semitones and a fifth is 7. Manual fourths can clash. Random varies unison, octave, fifth and octave below. Fix an interval first, then raise Layer.

11

空间:场景、长度与远近Space: scene, decay and distance

空间自然空灵暮色星野场景常态特效18 ms预延迟×1.4大小20%流动150 Hz低频9.0 kHz高频空间自然空灵暮色星野场景常态特效18 ms预延迟×1.4大小20%流动150 Hz低频9.0 kHz高频
图 08 · 空间控制区

矢量控件参考 · 数值为演示值 · 点击可放大

场景

场景决定空间的基础形状和早期反射。庭院像较近的墙面,寺院或石窟更有石质反射,星野可以很深远;这是听感设计,并非实测建筑的脉冲响应。换场景会套用相关空间设定,先保存当前声音再比较。

自然 / 空灵 / 暮色

自然保留较清楚的早期反射;空灵偏连续、均匀的长尾;暮色让空间缓慢展开,低频更悠长。比较时固定节气、Mix 和输入,避免同时改多个参数。

空间时间 / 大小

右侧「空间」是尾音持续时间,单位秒。细节里的「大小」改变延迟网络的尺度与反射间距,并不等于同一只时间旋钮。小房间也能有长尾,大空间也可设置短尾;分别比较才容易听清。

预延迟

输入出现后,空间隔多久才开始跟上,范围 0–250 ms。较短时原声和空间贴在一起;适度延后能让起音先出来,更清楚。过长会像单独的回声。用短促钢琴或敲击音比较比持续 Pad 更明显。

流动

让混响中的延迟线路轻微游移,产生活的宽度和流动感,不是自动左右声像旋钮。提高可能更丰富,也可能使尾音不够稳定。独奏乐器可小量使用,长 Pad 可稍多,再检查单声道兼容。

反向 / 冻结

反向让尾音以倒放片段涌起,听起来像从远处向音头吸入;冻结在演奏停下时延长或保持空间能量,下一次演奏会改变它。两者都不等于把整段原声倒放或永久暂停音频。短音后留出空白,才能听见区别。

低频 / 高频

低频是空间低切:提高会减少尾音里的低频堆积。高频控制进入尾音的亮度:降低通常更柔暗、更远。它们不等于季节染色 EQ。先听低音是否被尾巴糊住,再逐步提高空间低切。

SpaceNaturalEtherealDuskStarry skySceneNormalFX18 msPre-delay×1.4Size20%Motion150 HzLow cut9.0 kHzHigh cutSpaceNaturalEtherealDuskStarry skySceneNormalFX18 msPre-delay×1.4Size20%Motion150 HzLow cut9.0 kHzHigh cut
Figure 06 · Space controls

Vector control reference · illustrative settings · click to enlarge

Scene

Scene sets the base space and early-reflection character: nearby courtyard walls, stone-like temple/grotto reflections, or a deep starry space. These are designed impressions, not measured building impulse responses. Save edits before changing scenes.

Natural / Ethereal / Dusk

Natural has more distinct early reflections; Ethereal favours an even continuous tail; Dusk blooms slowly with longer low-frequency decay. Keep term, Mix and input fixed when comparing.

Decay / size

The right-hand Space control sets decay in seconds. Detail’s Size changes the delay-network scale and reflection spacing. They are different controls: a small space may have a long tail and a large one a short tail.

Pre-delay

The gap before the space follows the input, 0–250 ms. Short values blend the attack and space; a little delay lets the attack speak. Too long can read as a separate echo. Try short piano or percussive notes.

Motion

Gently varies reverb delay lines for a living, spacious tail; it is not an automatic pan control. More can feel richer but less stable. Use modestly on exposed instruments, more on pads, and check mono.

Reverse / freeze

Reverse makes tail slices swell backwards. Freeze holds or extends space in gaps, with subsequent playing changing the state. Neither reverses the whole dry track nor permanently pauses audio. Leave silence after a short note to hear them.

Low / high filters

Low is the space low-cut: raise it to reduce low-frequency build-up. High sets tail brightness; lower generally feels softer and farther. These filters are separate from seasonal Colour. Raise low-cut gradually if bass becomes muddy.

12

泛音与琴弦:音程与共鸣Shimmer and strings: intervals and resonance

泛音 · 琴弦移调尾音 / 溪语共鸣20%泛音+12 st音程0%溪语
图 09 · 泛音 · 琴弦控制区

矢量控件参考 · 数值为演示值 · 点击可放大

泛音

这里是混响尾音中的移调层,不是均衡器的高频提升。上八度会让尾巴向高处浮起,低八度会增加深度。先用一个长音,保持原声不动,小量增加泛音,再选择音程。

音程

定义移调层与输入的音高关系,例如八度、五加八、低八度。它不是选择整首歌调号的开关。即使音程本身常见,在复杂和弦上仍可能出现不需要的音;用整段乐句判断,不要只试单音。

琴弦 / 溪语 / 光晕

琴弦是右侧的共鸣层。溪语增加成对轻微失谐的共鸣,彼此产生细小拍动,像细流;这里的失谐是两根共鸣之间的差异,不是把整条原声拉跑调。光晕在质感组中增加和弦周围的共鸣感。若声音太厚,先分开降低溪语、琴弦与光晕,定位是哪一层。

Shimmer · StringsShimmer overtones / Brook resonance20%Shimmer+12 stVoicing0%Brook
Figure 07 · Shimmer · Strings controls

Vector control reference · illustrative settings · click to enlarge

Shimmer

Here Shimmer is a pitch-shifted layer in the tail, not a high-frequency EQ boost. An octave up lifts it; an octave down deepens it. Hold a note and increase it gently before choosing a voicing.

Voicing

Defines pitch relationships such as octave, fifth plus octave and octave below. It does not select the song key. Even familiar intervals can introduce unwanted notes over complex chords; test full phrases.

Strings / Brook / Halo

Strings is the right-hand resonance layer. Brook adds gently detuned resonance pairs that beat softly; this is not global detuning of the dry source. Halo in Texture adds resonant glow. Reduce Brook, Strings and Halo separately to locate excessive thickness.

13

颗粒:从小切片到纹理云Grains: from slices to a cloud

颗粒芽脉籽霜10 /s密度40%碎响50 ms间隔重复速度
图 10 · 颗粒控制区

矢量控件参考 · 数值为演示值 · 点击可放大

一段输入,不断取小片段,再让它们重叠加上柔和起落的包络密度低:颗粒之间有空隙密度高:更多片段交叠颗粒:切片、包络与交叠01 从输入截取短片段02 柔和起落,减小切片边缘03 密度改变重叠程度密度低:能分辨单独事件密度高:形成连续的纹理云
图 11 · 颗粒是什么:切片、包络与密度

矢量原理图 · 说明声音关系,不代表实测波形 · 点击可放大。

碎响:同一小片段被连续重读短间隔:更快的细碎重复AAAAAAA长间隔:更慢、更能分辨的重复AAA碎响:同一切片反复重读短间隔 → 重复更快AAAAAAA长间隔 → 重复更慢AAAA 始终是同一小段输入
图 12 · 碎响和间隔:重复比例与重复速度分开调

矢量原理图 · 说明声音关系,不代表实测波形 · 点击可放大。

什么是颗粒?

把最近输入的声音截成一个很短的片段,加上淡入淡出,再重复、移调、反向或分散到左右,就是一个颗粒。许多颗粒叠在一起成为纹理云。不是把钢琴换成另一套采样器;你弹得不同,片段素材也跟着变。

芽 / 萌芽

较短的片段,按输入相关的音程序列走动,像细小枝芽向外展开。适合想保留灵敏度、又增加一点分解运动的弹拨或键盘。先降低混合,再逐渐提高颗粒,避免细节盖过主旋律。

脉 / 叶脉

片段更长,带五度等关系,接合更平滑,像一条条拉长的纹理线。适合拉长短音或让和弦更连贯;如果起音失去清晰度,可减少颗粒或用闪避保留原声。

籽 / 种子

短小、以同度为主的片段,更多保留输入本来的音高轮廓,容易听成细碎落点。它不意味着必定强烈卡顿;卡顿比例仍由碎响控制。想要小颗粒却不要重复感时,把碎响降下来。

霜 / 霜晶

更长、更柔的片段,含较多反向成分,容易形成缓缓凝结的背景。适合慢速铺底和空隙中的延伸。快速密集旋律上可能模糊节奏,可降低生长或颗粒,而不是一味提高亮度。

密度

表示颗粒云每秒触发多少片段,界面基准范围为 1–40 次/秒。低时能分辨零散事件,高时更多片段重叠成连续云。生长和主颗粒旋钮还会影响最终密度,所以显示数值不等于输出里永远准确可数的颗粒数。它不是音量,也不是琶音速度。

碎响

决定有多少片段会被连续重读成轻微的卡顿。低时多数片段只走一次,较高时更容易听见同一小片段反复。先固定密度,只改碎响,就能听清“事件更多”和“同一事件重复”是两回事。

间隔

控制碎响片段的重复时间尺度,范围 15–400 ms,并有细小随机变化。短时重复更快、更细,长时更慢、更能听清片段本身。它不控制普通颗粒之间的空隙;碎响很低时,改变它可能不明显。先稍加碎响,再比较 40 与 200 ms。

两分钟分辨颗粒参数
  1. 循环一段有明显起音的音频,固定节气与 Mix,适当降低回声和自然声。
  2. 只改密度:听稀疏点状到连续云的变化。
  3. 固定密度,加一点碎响:听同一片段开始反复。
  4. 保持碎响,比较短/长间隔:听重复速度,而不是音量。
  5. 最后在芽、脉、籽、霜间比较;切换质感会改变一组底层设计,不只是一个滤波器。
GrainsSproutLeafSeedFrost10 /sDensity40%Glitch50 msIntervalrepeat speed
Figure 08 · Grains controls

Vector control reference · illustrative settings · click to enlarge

One input → short slices → overlapping grainsApply a soft envelopeLow density: gaps between eventsHigh density: more overlapGrains: slices and overlap01 Slice the input02 Soften the grain edges03 Density changes overlapLow: distinct eventsHigh: an overlapping cloud
Figure 09 · Grains: slicing, envelope and density

Vector concept diagram · relationships, not measured audio · click to enlarge.

Glitch: read the same short slice againShort interval: faster repeatsAAAAAAALong interval: slower repeatsAAAGlitch: repeat the same sliceShort interval → fasterAAAAAAALong interval → slowerAAAA is the same input slice
Figure 10 · Glitch and interval: amount versus repeat speed

Vector concept diagram · relationships, not measured audio · click to enlarge.

What is a grain?

A grain is a short piece of recent input with a fade-in/out, optionally repeated, shifted, reversed or spread across stereo. Many overlapping grains form a cloud. It uses your playing rather than replacing it with another sample instrument.

Sprout

Shorter slices step through an interval run, like small shoots branching outward. Useful for responsive plucked or keyboard textures. Start with lower Mix, then raise Grains gradually.

Leaf

Longer, smoother slices with fifth-related colour make extended texture lines. Good for connecting chords or lengthening short notes. Reduce Grains or use Ducking if attacks lose clarity.

Seed

Short, mostly unison slices retain more of the input pitch contour and make small scattered events. Seed does not automatically mean heavy stuttering; lower Glitch for small grains without obvious repeats.

Frost

Longer, softer slices with more reverse content make a slowly forming background. Good for beds and gap tails. On fast lines it may blur rhythm; reduce Growth or Grains rather than simply adding treble.

Density

The grain cloud’s base trigger rate, displayed as 1–40 events/s. Low feels scattered; high overlaps into a cloud. Growth and the main Grains knob also affect effective density, so the display is not an exact count at the output. It is neither level nor arpeggio speed.

Glitch

Sets the tendency for slices to stutter. Low mostly plays a slice once; higher makes repeated reads more apparent. Hold Density fixed while changing Glitch to distinguish more events from repeated events.

Interval

Sets the stutter slice’s time scale, 15–400 ms with small random variation. Short gives fast fine repeats; long gives slower recognisable ones. It is not the spacing between ordinary grains. With little Glitch, changes can be subtle; raise Glitch slightly before comparing 40 and 200 ms.

A two-minute grain comparison
  1. Loop audio with clear attacks. Fix term/Mix and lower Echo/Nature.
  2. Change only Density: scattered events to cloud.
  3. Fix Density and add Glitch: hear the same slice repeat.
  4. Keep Glitch and compare short/long Interval: listen for repeat speed.
  5. Compare the four textures last; each changes a group of underlying design choices.
14

琶音:节奏、音程与原声粒Arpeggio: rhythm, intervals and source grains

琶音同步自由1/8速度45%律动预留40%原声粒
图 13 · 琶音控制区

矢量控件参考 · 数值为演示值 · 点击可放大

琶音:同一输入的片段,依次站在不同音程上仅为顺序示意,具体顺序由当前节气设计决定+0 st+7 st+12 st+7 st+0 st-12 st同步:跟随 DAW 拍值 | 自由:每秒步数速度=时钟快慢;原声粒=这一层音量;律动当前未接入发声路径琶音:依次展开输入片段顺序仅为示意,随节气设计变化1+0 st2+7 st3+12 st4+7 st5+0 st6-12 st同步:跟随拍值;自由:每秒步数律动在 v1.0 中为预留参数
图 14 · 琶音、速度、律动与原声粒

矢量原理图 · 说明声音关系,不代表实测波形 · 点击可放大。

什么是琶音?

把音依次展开,而不是同时堆在一起,叫琶音。Hare 用输入的小切片,按当前节气设计的音程序列依次播放。它与颗粒云可以同时存在:云是分散交叠的纹理,琶音是更有步进顺序的片段。没有输入时,它不会像独立音源那样演奏一串新音符。

同步 / 自由

同步时以宿主速度为基准,速度推子选择拍值;自由时按每秒步数运行,不随工程 BPM 一起改变。想跟鼓组贴合先选同步;想要漂浮、与节拍错开的运动可试自由。

速度

决定步进时钟走得多快。同步模式可选 1/2 到 1/32 等拍值;自由模式约 0.25–16 步/秒。120 BPM 下,1/8 每 250 ms 一步,1/16 每 125 ms 一步。速度提高不等于一定更响,也不等于颗粒云密度提高。

律动

这个术语通常指事件与稳定拍点之间的关系,但当前 Hare 的实现需要特别区分:界面会改变并保存律动数值,实际原声琶音直接按速度时钟触发,没有读取这个数值;旧粒子路径已停用。因此当前不应期待它改变琶音的松紧或 Swing。若听不出变化,不必反复放大输入;请使用速度、同步与原声粒来控制当前可用的琶音。

原声粒

控制琶音这一层的输出分量。“原声”指片段来自你的输入,不是干声直通音量。把它降到 0 可减少或关闭这一路琶音;仍可能听见颗粒云和回声。想单独辨认琶音,可降低其他纹理层,再从 0 慢慢推高原声粒。

两分钟辨认琶音
  1. 用持续和弦作输入,减小其他层,逐渐提高原声粒。
  2. 同步到宿主,先比较 1/8 与 1/16:前者更慢。
  3. 固定速度比较原声粒的低量与高量;律动当前未接入这一路发声,不作为试听步骤。
  4. 把原声粒拉低,确认消失的是琶音层;颗粒云、回声可能仍在。
ArpeggioSyncFree1/8Speed45%Pulsereserved40%Source grains
Figure 11 · Arpeggio controls

Vector control reference · illustrative settings · click to enlarge

Arpeggio: source slices step through intervalsIllustrative sequence; actual patterns depend on the term+0 st+7 st+12 st+7 st+0 st-12 stSync: DAW divisions | Free: steps per secondSpeed = clock; Source grains = level; Pulse currently inactiveArpeggio: source slices in orderIllustrative, term-dependent order1+0 st2+7 st3+12 st4+7 st5+0 st6-12 stSync: divisions; Free: steps/secondPulse is reserved in v1.0
Figure 12 · Arpeggio, speed, pulse and source-grain level

Vector concept diagram · relationships, not measured audio · click to enlarge.

What is an arpeggio?

An arpeggio unfolds notes in sequence rather than stacking them simultaneously. Hare uses input slices following a term-designed interval pattern. It can coexist with the grain cloud: the cloud scatters/overlaps, while the arp has a more stepped sequence. It is not a standalone instrument.

Sync / free

Sync follows host tempo and Speed selects divisions. Free runs in steps per second independently of BPM. Use Sync to lock to drums and Free for floating patterns.

Speed

Sets the step clock. Sync divisions range from 1/2 to 1/32 with dotted/triplet choices; Free spans about 0.25–16 steps/s. At 120 BPM, 1/8 is 250 ms per step and 1/16 is 125 ms. Faster is not necessarily louder and does not increase cloud Density.

Pulse

Pulse usually describes how events relate to a regular beat, but the current Hare implementation differs: the UI stores its value while the active source arpeggio triggers directly from the speed clock without reading it. The former particle path is disabled. Do not expect this control to change arp tightness or Swing; use Speed, Sync and Source grains for the active arp.

Source grains

Sets the arpeggio layer’s level. “Source” means slices originate from your input, not dry-through gain. At 0 this arp path is reduced/removed, but cloud grains and echo may remain. Lower other texture layers and raise this control from 0 to identify the arp.

A two-minute arpeggio comparison
  1. Use sustained chords, reduce other layers and raise Source grains.
  2. Sync to the host and compare 1/8 versus 1/16: 1/8 is slower.
  3. Fix Speed and compare Source grains levels. Pulse is not wired into this active path.
  4. Lower Source grains to identify the arp path; cloud/echo may remain.
15

质感:光晕、明暗、染色与总量Texture: Halo, Tone, Colour and Amount

质感分层塑形20%光晕0明暗50%染色50%总量
图 15 · 质感控制区

矢量控件参考 · 数值为演示值 · 点击可放大

光晕

增加输入和弦周围的共鸣感,像音的边缘更发亮、更有延展,不是简单复制原声。持续和弦上更易听清。和琴弦、溪语同时很高时会显得拥挤,先逐层比较。

明暗

倾斜效果纹理的低高频平衡。向暗处减少纹理的亮感,向亮处强调上方细节。它不会替代空间滤波,也不会单独撤销季节 EQ 对合成信号的影响。觉得尾巴尖锐,先辨认尖锐来自回声、空间还是颗粒,再动对应控制。

染色

控制当前节气默认 EQ 与少量谐波的程度。设为 0 可作无季节染色的对照;提高后还需有足够 Mix 才能听清。自适应不是把你的声音修成标准频谱,而是在输入已经偏向该季节时收敛同方向处理。具体五段、Mix 曲线和测试方法见前面的染色章节。

总量

调整多层纹理的整体表现力,是声音设计宏控制,不等同于输出音量。想只改变听音响度用输出音量;想让纹理更克制或更夸张用总量,再复核 Mix。

TexturePer-layer shaping20%Halo0Tone50%Colour50%Amount
Figure 13 · Texture controls

Vector control reference · illustrative settings · click to enlarge

Halo

Adds resonant glow around input chords rather than a simple dry copy. Sustained chords reveal it well. High Halo, Strings and Brook together can crowd the sound; compare them individually.

Tone

Tilts the effect texture’s low/high balance. Dark softens its brightness; bright emphasises detail. It neither replaces space filtering nor independently cancels seasonal Colour. Locate sharpness in echo, space or grains before adjusting.

Colour

Sets seasonal EQ and subtle harmonics. Use 0 as an uncoloured reference; raising it needs sufficient Mix to become apparent. Adaptation does not force a standard spectrum: it reduces same-direction processing when the input already leans that way. See the Colour chapter for bands and Mix engagement.

Amount

A design macro for the texture’s overall expression, not output gain. Use Output for listening level; use Amount for restrained versus exaggerated texture, then reassess Mix.

16

自然声:频谱、距离与应答Nature: spectrum, distance and Reply

自然声分层塑形150 Hz低切6 dB中频50%应答4.5 kHz高切
图 16 · 自然声控制区

矢量控件参考 · 数值为演示值 · 点击可放大

低切

削弱自然录音低频,减少风声或环境底层的轰鸣。提高后更轻、更不占低音空间;过高会失去风雪的厚度。先在低音乐器进入时判断是否需要,而不是预先把自然层全部削薄。

中频

在自然声中频挖出一块位置,为乐器让路,范围 0–18 dB。这是衰减量,不是选择中频中心频率。提高会更避开乐器主体,但过多可能使录音空洞;先试小量,再与关闭比较。

应答

乐句结束后,环境层与录音事件可能回应,像演奏停下后自然浮上来。它有乐句检测、延迟和概率;不是每个音都必定触发,也不合成新的鸟声。弹一个完整短句后留一两秒,再比较低量与高量,比连续弹满更能听清。

高切

削弱自然层的高频,让录音更柔、更像远景。过低会使雨声或虫声失去辨识度。它只影响自然层,不会让原声乐器整体变暗。

距离不是单一旋钮

自然声远近由音量、频谱、空间与跟随共同决定:降低音量、高切更低通常更远;把跟随调高会贴着演奏起落。想让环境退后,先小量减自然声,再调高切,最后决定是否需要更强跟随。

NaturePer-layer shaping150 HzLow cut6 dBMid dip50%Reply4.5 kHzHigh cut
Figure 14 · Nature controls

Vector control reference · illustrative settings · click to enlarge

Low cut

Removes low-frequency rumble from nature recordings. Raise it for lighter ambience with less bass masking; too high loses body. Judge with the bass instrument playing.

Mid dip

Cuts a midrange region in nature to make room for the instrument, 0–18 dB. This controls cut depth, not centre frequency. More clears room but can sound hollow; compare small amounts with 0.

Reply

Recorded events/ambience may answer after phrases. Phrase detection, delay and probability apply; it is not a guaranteed response to every note and synthesises no new bird call. Play a short phrase, leave a gap and compare low/high Reply.

High cut

Softens nature’s high frequencies for a more distant impression. Too low can obscure rain/insect identity. It does not darken the dry instrument.

Distance is a combination

Distance combines level, spectrum, space and Follow. Lower level and a lower high-cut generally recede; more Follow ties the envelope to playing. Reduce Nature slightly, adjust high-cut, then decide whether stronger Follow helps.

17

预设、A/B、复制与混合锁Presets, A/B, copy and Mix lock

「我的预设」输入名称后保存,列表可载入或删除。打开预设文件夹可备份或分享 ~/Library/Audio/Presets/Budgie Hare 中的 .budgiehare 文件。

A/B 是两个可比较的声音状态;复制把当前完整声音设置送到剪贴板,粘贴载入兼容的 Hare 设置文本。它不是音频导出功能。

混合旋钮旁的小锁开启时,换节气、载入预设、A/B 与粘贴保持当前 Mix。需要完整还原预设混合量时,先解锁再载入。锁状态随工程保存,而不是用户预设内容。

语言和缩放属于界面设置,不用声音预设交换。保存 DAW 工程保留插件的当前参数;仅保存用户预设不能替代保存音轨和路由。

Type a name in Presets and save; use the list to load or delete. Open the preset folder to back up/share .budgiehare files in ~/Library/Audio/Presets/Budgie Hare.

A/B compares two sound states. Copy puts the current sound settings on the clipboard; Paste reads compatible Hare text. Neither exports audio.

The padlock beside Mix preserves Mix across term changes, presets, A/B and paste. Unlock before loading when you want the preset’s saved Mix. The lock is stored with the session rather than the user preset.

Language and scale are UI settings, separate from sound presets. Save your DAW session to retain the track, routing and current plugin state.

18

灵感配方与恢复默认Inspired recipes and reset

  • 空灵 · 远山:清透的长空间。
  • 溪语 · 山涧:成对共鸣与流动重音。
  • 暮语 · 晚钟:更湿、更暗、缓慢展开的空间。

配方叠加在当前节气上,主要改动空间、回声和琴弦等参数;再次点击同一配方取消。切换节气会取消配方。想回到节气设计,可在齿轮菜单选择「恢复当前节气的出厂设置」。若混合锁开启,恢复时仍需留意 Mix。

设置提供中文/English、80–150% 的界面缩放。打开细节时界面会扩展;宿主窗口较小时先降低缩放。

  • Distant Hills: a clear, long space.
  • Mountain Brook: paired resonance and moving accents.
  • Evening Bell: wetter, darker and slowly blooming.

Recipes overlay the current term, changing space, echo, strings and related settings. Click the active recipe again to cancel. A term change clears it. Use Settings → Restore this term’s factory settings to return to its design; check Mix if its lock is on.

Settings offers Chinese/English and 80–150% UI scaling. Detail expands the window; reduce scale in a small host window.

19

三个实战工作流Three practical workflows

清晰的钢琴或拨弦

从春季节气开始,适度混合,降低自然声;用生长增加颗粒形态,用闪避保留起音。若原声已经明亮,先减少染色,再调整光泽。

厚重的冬季铺底

选择冬至或大寒,缓慢演奏并留出空隙。逐渐提高 Mix,比较染色 0/50;自然声与跟随决定风雪是持续远景还是随音起落。降低低频过多时,检查输入本身与后续总线处理。

通过 MB Rack 只处理高频

把 Hare 放入 MB Rack 的高频处理链,低频保持独立。先降低自然声,确认跨频段的空间不会掩盖原声;这需要单独使用 MB Rack,不是 Hare 内置分频功能。

Clear piano or plucked parts

Start with spring, a moderate Mix and less Nature. Raise Growth for texture and use Ducking to protect the attack. With an already bright source, reduce Colour before raising Shine.

A deep winter bed

Choose Winter Solstice or Major Cold, play slowly and leave gaps. Raise Mix gradually and compare Colour 0/50. Nature and Follow decide whether wind/snow stays behind the music or follows notes.

High-band ambience with MB Rack

Insert Hare in MB Rack’s high-band chain and keep the lows separate. Start with less Nature. This requires MB Rack; Hare has no built-in multiband crossover.

20

常见问题与排查Troubleshooting

为什么换节气 EQ 不明显?

确认染色不是 0,Mix 高于 10%;可先设 Mix 80%、染色 50%,循环同一段音频。输入已偏向该季节的频谱时,自适应会减轻处理。EQ 目标值不是实时固定增益。

为什么没有 MIDI/调性按钮?

当前插件已改为普通音频效果器,自动分析音频。旧手册或旧截图中的 MIDI 设置不适用于当前界面。

没演奏也有风雨声?

检查自然声、跟随和天气。跟随较低时环境层可以持续;想安静,减少自然声并提高跟随。雨仍会随演奏力度改变。

怎么回到原声?

将 Mix 降到 0,并检查输入/输出增益。做严格对比时也可使用宿主旁通。

切预设后天气或参数变了?

这是当前设计:预设载入自己的参数及默认天气。先存用户预设,再切换;需要保留 Mix 时开启混合锁。

旧工程找不到 AU/插件不出现?

确认 macOS、格式与宿主扫描。旧 MIDI 控制 AU 与当前普通 AU 类型不同,可能要重新插入。先备份工程,不要覆盖唯一副本。

需要反馈问题时提供什么?

提供 macOS、DAW、插件格式、设置中的版本号、复现步骤及截图。可以用复制功能附上 Hare 设置文本。联系 hello@budgieaudio.com。

Why is the EQ change subtle?

Check Colour is above 0 and Mix above 10%. Try Mix 80%, Colour 50%, and loop identical audio. Adaptation softens the change when the input already leans that way. Target EQ values are not fixed live gains.

Where are MIDI and key controls?

The current plugin is a regular audio effect using audio analysis. Older MIDI-control instructions do not apply.

Why is nature audible without playing?

Check Nature, Follow and weather. Low Follow allows an ambient bed; reduce Nature and raise Follow for quieter gaps. Rain intensity also responds to playing.

How do I return to the source?

Set Mix to 0 and check input/output gains. Use host bypass for a strict comparison.

Why did weather/settings change with the term?

Terms load their own settings and default weather. Save a user preset before switching. Use Mix lock to preserve Mix.

Missing AU in an older session?

Check OS, format and host scan. The former MIDI-controlled AU and current audio-effect AU have different types; reinsertion may be needed. Back up the session first.

What should a support report include?

Include macOS, DAW, format, version shown in Settings, steps and screenshots. Copy can supply the Hare settings text. Contact hello@budgieaudio.com.

21

术语:从声音现象理解参数Vocabulary: hear what the terms mean

这些词贯穿整本手册。先理解声音对象,再操作对应的旋钮;同一个词出现在不同组里,可能控制不同路径。

干声、湿声与 Mix

干声是输入原本的声音;湿声是处理产生的效果。Mix 改变二者比例。在 Hare 中提高 Mix 还会引入更多季节染色,所以比较预设时固定 Mix,才能区分“更湿”和“音色本身改变”。把 Mix 降到 0 可检查原声与输入输出增益。

Hz、kHz 与明暗

Hz 表示每秒振动次数,kHz 是千 Hz。通常较低频率对应厚度和低音,较高频率对应清晰度、齿音或空气感,但实际听感取决于原声。滤波器上写 9 kHz,不意味着只处理一个极窄的 9 kHz 音;这里的架式滤波会影响一整片高频。

dB 与音量比较

dB 是相对比例的对数表示。EQ 的 +3 dB 表示该频段目标提升,不是整段音乐必然更响 3 dB;多个宽频段也会叠加。Hare 染色后还有响度补偿。做 A/B 时用相同乐句并留意输出电平,不要把更响当作细节更多。

ms、秒与节奏

1000 ms 等于 1 秒。20 ms 的预延迟是很短的起音间隔,500 ms 的回声更容易听成单独一次重复。空间用秒表示长尾,碎响间隔用毫秒描述短片段重读;单位一样也不代表控制同一件事。

低切 / 高通 与 高切 / 低通

低切也叫高通:削弱截止频率以下的低频,让上方通过。高切也叫低通:削弱高于截止频率的部分。滤波不是砖墙式切断,过渡是渐变的。空间太浑时提高空间低切;自然声太近、太亮时降低自然声高切,注意操作对象不同。

架式 EQ 与峰形 EQ

架式 EQ 像把低频或高频整片抬高或压低;峰形 EQ 则围绕一个中心区域形成宽的凸起或凹陷。季节染色的 120 Hz、9 kHz 分别是低架与高架,中间三段是宽峰形。因此表格中的数值是滤波器目标,不应直接当作五个频点的实际测量结果。

包络与起音

包络描述声音随时间的强弱形状:起音、保持与消退。给颗粒加柔和起落能减少生硬边缘;跟随让自然声的强弱贴着演奏走。它不等于压缩器,也不意味着“跟随”会识别你想要的节奏。用一长一短两种音比较,比只看推子数值更容易理解。

八度、五度与半音

半音是常见十二平均律中相邻音之间的一步;八度相差 12 半音,五度常指 7 半音。移调颗粒是把同一片段用不同速度或比例重放,不是自动作曲。五度虽然常见,在复杂和声里也可能不合适。先用简单和弦试听,再检查整首曲子的转调段。

泛音、谐波与泛音推子

谐波是与基音呈整数倍关系的频率成分。季节染色添加的是少量二、三次谐波;界面上的“泛音”推子控制的却是尾音中的移调层。这两个地方都可能增加丰富度,但机制和听感不同:要减少整体染色用染色,要减少上浮尾音用泛音。

失谐与拍动

两个很接近但不完全相同的频率同时响,会产生缓慢的强弱起伏,叫拍动。溪语中的成对共鸣和颗粒的细微失谐会带来宽度与流动,不等于把整条输入变调。若独奏声部觉得晃动太多,分别降低溪语、光泽或相关纹理层来定位。

自适应与响度补偿

自适应先观察输入的频谱倾向,再减轻已经顺着季节方向的处理;不会为每种乐器固定套用同样的增益。响度补偿尝试让处理前后的整体强弱接近,它不保证每个瞬态、电平表或每段音乐都完全一致。比较时让同一段音频持续一会儿,等待分析稳定。

混响与延迟

延迟通常能辨认出一次次重复;混响是大量反射重叠成的空间尾音。Hare 会让它们与颗粒互相融合,但调节入口仍然不同。想减少“一下一下”的重复先查回声和珠落;想缩短持续的空间背景先查空间时间、泛音与冻结。

These terms recur throughout the manual. Identify the sound path before turning a control; similar names in different groups can affect different paths.

Dry, wet and Mix

Dry is the original input; wet is the processed effect. Mix blends them. In Hare, raising Mix also engages seasonal Colour. Fix Mix when comparing presets to separate a wetter blend from a tonal change. Use Mix 0 to check the source and gains.

Hz, kHz and brightness

Hz means cycles per second; kHz means thousands of Hz. Lower regions often carry body/bass and higher ones detail or air, depending on the source. A 9 kHz shelf affects a broad high-frequency region, not one narrow tone.

dB and level comparisons

dB describes a logarithmic ratio. An EQ target of +3 dB does not mean the whole track gets 3 dB louder; broad bands overlap, and Colour also level-matches. Compare the same phrase and check output level rather than assuming louder means more detailed.

ms, seconds and rhythm

1000 ms equals one second. A 20 ms pre-delay is a short onset separation; a 500 ms echo is more easily heard as a separate repeat. Space decay uses seconds and stutter interval uses milliseconds, but timing controls can affect different processes.

Low-cut / high-pass and high-cut / low-pass

Low-cut (high-pass) reduces lows while passing higher frequencies. High-cut (low-pass) reduces highs. These are gradual filters, not brick walls. Raise the space low-cut for muddy tails; lower the nature high-cut for overly bright ambience. Choose the correct path.

Shelving and bell EQ

A shelf raises/lowers a broad low- or high-frequency region. A bell forms a hump/dip around a centre. Seasonal Colour uses shelves at 120 Hz/9 kHz and broad bells in between. Table entries are filter targets, not measurements of response at five points.

Envelope and attack

An envelope is the sound’s level shape over time: onset, sustain and decay. Soft grain envelopes avoid hard edges; Follow makes nature trace playing dynamics. An envelope is not itself a compressor or an intention-aware rhythm detector. Compare short and held notes.

Octaves, fifths and semitones

A semitone is one step in common 12-tone equal temperament; an octave spans 12 and a fifth usually 7. Shifted grains replay source slices at pitch ratios rather than composing new parts. Even fifths can be unsuitable over complex harmony. Check modulations, not only one chord.

Harmonics versus the Shimmer control

Harmonics occur at integer multiples of a fundamental. Colour adds small second/third-harmonic shares, while the UI’s Shimmer fader controls a pitch-shifted tail layer. Reduce Colour for overall colouring and Shimmer for rising tails; they are different processes.

Detuning and beating

Two nearby frequencies produce slow level fluctuations called beating. Brook’s resonance pairs and slight grain detuning add width/motion rather than globally retuning the source. If an exposed part feels too unstable, reduce Brook, Shine or texture layers separately.

Adaptation and level matching

Adaptation observes input balance and softens processing already aligned with the seasonal direction. It is not one fixed gain for every instrument. Level matching aims for similar overall loudness, not identical transients/meters. Loop the same audio and allow analysis to settle.

Reverb and delay

Delay often reveals distinct repeats; reverb blends many reflections into a spatial tail. Hare connects these layers, but their controls differ. Reduce Echo/Pearls for repeated hits and inspect Space decay, Shimmer or Freeze for a sustained background.

22

附录:24 套 EQ 目标曲线Appendix: 24 EQ target profiles

单位 dB。此表是完整染色、适应前的五段滤波器目标值,包含源码的 1.4 倍轮廓系数。不是默认 50% 染色时的实际频响,也不是五个频率处实测的输出增益;宽频段会彼此叠加。实际处理还受 Mix、染色推子、自适应输入与响度补偿影响。表中未列出少量谐波。

节气 / Term120 Hz300 Hz1 kHz3 kHz9 kHz
立春 / Spring Begins-0.70-2.10+0.00+0.70+2.80
雨水 / Rain Water-0.93-2.10-0.23+0.70+3.03
惊蛰 / Insects Awaken-1.17-2.10-0.47+0.70+3.27
春分 / Spring Equinox-1.40-2.10-0.70+0.70+3.50
清明 / Clear and Bright-1.17-1.87-0.00+0.93+2.80
谷雨 / Grain Rain-0.93-1.63+0.70+1.17+2.10
立夏 / Summer Begins-0.70-1.40+1.40+1.40+1.40
小满 / Grain Buds-0.47-1.17+1.17+1.63+1.63
芒种 / Grain in Ear-0.23-0.93+0.93+1.87+1.87
夏至 / Summer Solstice+0.00-0.70+0.70+2.10+2.10
小暑 / Minor Heat+0.47+0.23+0.47+1.17+0.93
大暑 / Major Heat+0.93+1.17+0.23+0.23-0.23
立秋 / Autumn Begins+1.40+2.10+0.00-0.70-1.40
处暑 / End of Heat+1.17+2.10-0.23-0.47-0.70
白露 / White Dew+0.93+2.10-0.47-0.23-0.00
秋分 / Autumn Equinox+0.70+2.10-0.70+0.00+0.70
寒露 / Cold Dew+1.17+1.87-0.47-0.47-0.23
霜降 / Frost Descends+1.63+1.63-0.23-0.93-1.17
立冬 / Winter Begins+2.10+1.40+0.00-1.40-2.10
小雪 / Minor Snow+2.57+1.63-0.23-1.63-2.57
大雪 / Major Snow+3.03+1.87-0.47-1.87-3.03
冬至 / Winter Solstice+3.50+2.10-0.70-2.10-3.50
小寒 / Minor Cold+2.10+0.70-0.47-1.17-1.40
大寒 / Major Cold+0.70-0.70-0.23-0.23+0.70

Units: dB. These are full-colour, pre-adaptation filter targets, including the source’s 1.4 profile factor. They are neither the response at default 50% Colour nor measured output gains at these frequencies; broad bands overlap. Mix, Colour, adaptation and level matching affect the result. Harmonic shares are omitted.

节气 / Term120 Hz300 Hz1 kHz3 kHz9 kHz
立春 / Spring Begins-0.70-2.10+0.00+0.70+2.80
雨水 / Rain Water-0.93-2.10-0.23+0.70+3.03
惊蛰 / Insects Awaken-1.17-2.10-0.47+0.70+3.27
春分 / Spring Equinox-1.40-2.10-0.70+0.70+3.50
清明 / Clear and Bright-1.17-1.87-0.00+0.93+2.80
谷雨 / Grain Rain-0.93-1.63+0.70+1.17+2.10
立夏 / Summer Begins-0.70-1.40+1.40+1.40+1.40
小满 / Grain Buds-0.47-1.17+1.17+1.63+1.63
芒种 / Grain in Ear-0.23-0.93+0.93+1.87+1.87
夏至 / Summer Solstice+0.00-0.70+0.70+2.10+2.10
小暑 / Minor Heat+0.47+0.23+0.47+1.17+0.93
大暑 / Major Heat+0.93+1.17+0.23+0.23-0.23
立秋 / Autumn Begins+1.40+2.10+0.00-0.70-1.40
处暑 / End of Heat+1.17+2.10-0.23-0.47-0.70
白露 / White Dew+0.93+2.10-0.47-0.23-0.00
秋分 / Autumn Equinox+0.70+2.10-0.70+0.00+0.70
寒露 / Cold Dew+1.17+1.87-0.47-0.47-0.23
霜降 / Frost Descends+1.63+1.63-0.23-0.93-1.17
立冬 / Winter Begins+2.10+1.40+0.00-1.40-2.10
小雪 / Minor Snow+2.57+1.63-0.23-1.63-2.57
大雪 / Major Snow+3.03+1.87-0.47-1.87-3.03
冬至 / Winter Solstice+3.50+2.10-0.70-2.10-3.50
小寒 / Minor Cold+2.10+0.70-0.47-1.17-1.40
大寒 / Major Cold+0.70-0.70-0.23-0.23+0.70
23

HUB 激活、试用与离线使用HUB activation, trial and offline use

Hare 1.0 使用与其他 Budgie 产品相同的机器绑定签名许可证。请安装 Budgie Hub 0.8.10 或更高版本,在 HUB 中登录账号,再为 Hare 激活已购许可证或开始 14 天试用。插件本身不收集账号密码,也不提供激活码输入框。

  1. 在 HUB 的「我的软件」找到 Hare,下载安装包并安装所需格式。
  2. 在许可证页面选择 Hare,激活当前电脑;也可在插件的授权提示中点击「打开 Budgie Hub」或「开始 14 天试用」。
  3. 完成后回到插件,授权状态会自动刷新,无需重新创建工程。

正式授权每个账号最多激活两台电脑,14 天试用绑定一台电脑。需要更换电脑时,先在 HUB 释放旧设备席位。永久授权需要定期通过 HUB 联网续确认;有效缓存许可证允许短期离线使用。不要通过回拨系统时间延长试用,时间异常时请恢复正确时间并联网打开 HUB。

未激活、试用结束、许可证无效或超过有效离线期限时,Hare 直通原始音频,不输出处理后的效果。工程参数仍可保存。重新激活后,原来的设置继续生效。

Hare 1.0 uses the same machine-bound, signed licensing system as other Budgie products. Install Budgie Hub 0.8.10 or later, sign in, then activate your Hare license or start the 14-day trial. Hare never asks for your account password or activation code.

  1. Find Hare in My Software in HUB and install the formats you need.
  2. Activate this computer on the Licenses page. The plug-in’s Open Budgie Hub and Start 14-day trial buttons open the same workflow.
  3. Return to the plug-in. License status refreshes automatically; you do not need to recreate the session.

Paid licenses allow up to two computers per account; the 14-day trial is bound to one computer. Release an old device in HUB before replacing it. Perpetual licenses need periodic online confirmation through HUB; a valid cached license supports short periods offline. If the system clock is incorrect, restore the correct time and open HUB online.

Without a valid activation, after trial expiry or after the offline license deadline, Hare passes the original audio through without its effects. Session parameters remain saveable and take effect again after activation.

24

授权信息与云端预设License details and cloud presets

从 Hare 1.0.1 起,点击右上角齿轮可查看授权状态、当前设备、缩略许可证编号、联网确认期限;试用授权显示剩余天数、小时或分钟。「重新检查授权」重新读取本机签名许可证;「在 BUDGIE HUB 中管理」打开许可证页面。

云端预设需要 Budgie Hub 0.8.9 或更高版本。先在 Hare 的「我的预设」中保存音色,再点击齿轮里的「同步到云端 · 打开 HUB」。登录 HUB 并启用 Hare 同步后,每 30 秒自动检查,也可点击「立即同步」。只有已保存的 .budgiehare 用户预设参与同步;未保存的旋钮改动、DAW 工程和 24 个内置节气不会上传。

另一台电脑登录同一账号并启用同步后,重新打开插件的「我的预设」即可刷新列表。同步不会改变正在播放的声音。自动同步需要 HUB 保持运行;断网时文件保留在本机,联网后重试。

本机与云端均有修改,或切换账号时,HUB 会暂停覆盖,由你选择「保留本机版本」或「恢复云端版本」。保留本机会上传本机整套用户预设;恢复云端会替换本机整套预设,包括同步删除。恢复前保存本机备份,云端保留最近 10 个版本;可以使用「本机备份」和「恢复历史版本」找回。总容量上限 8 MiB。

From Hare 1.0.1, the gear menu shows activation status, this computer, an abbreviated license number and the online renewal deadline. Trials show remaining days, hours or minutes. Check license again rereads the signed local license; Manage in BUDGIE HUB opens license management.

Cloud presets require Budgie Hub 0.8.9 or later. Save a User Preset, then choose Sync presets · Open HUB. Sign in and enable Hare sync. HUB checks every 30 seconds; Sync now starts an immediate check. Only saved .budgiehare User Presets are synced. Unsaved knob changes, DAW projects and the 24 factory solar terms are excluded.

Sign in to the same account on another computer and enable sync, then reopen User Presets in Hare to refresh its list. Synchronization never changes the sound currently playing. Keep HUB running for automatic sync. Files stay local offline and retry when connected.

If both copies change, or you switch accounts, HUB pauses overwrites and asks whether to Keep this computer or Restore cloud copy. Keeping this computer uploads its entire User Preset set; restoring cloud replaces the local set, including synced deletions. Restores save a local backup first, and the service retains the latest ten cloud versions. Use Local backups or Restore history to recover. Total data is limited to 8 MiB.

25

简体、繁体与语言记忆Simplified Chinese, Traditional Chinese and language memory

Hare 1.0.2 在右上角齿轮中提供「简体中文」「繁體中文」「English」。选择后会记住本机最后一次选择;下次打开界面、新建插件实例,或在其他 DAW 中打开 Hare,都会沿用这个语言。旧工程、用户音色、A/B 对比或节气默认不会把语言改回去。

繁体模式包含界面、节气与物候名称、提示、授权信息和云端预设说明。你自己命名的预设、电脑名称、文件内容和授权编号保持原样。语言属于本机偏好;换电脑后可以在当地重新选择,不会改变音频参数或声音。

Hare 1.0.2 offers Simplified Chinese, Traditional Chinese and English in the gear menu. Your last explicit choice is saved for this computer and reused when you reopen the editor, create a new instance or open Hare in another DAW. Older sessions, User Presets, A/B comparison and factory term resets do not override it.

Traditional Chinese covers controls, solar terms and phenology, tooltips, license details and cloud-preset instructions. Your own preset names, computer name, file contents and license identifiers remain unchanged. Language is a local preference; choose it again on a different computer. Audio parameters and sound are unaffected.

矢量图 / Vector figure · 可横向移动查看