FrameCut 0.1.2

中文

这一版补齐了 vlog 常用的文字和声音：标题、多条音频轨、旁白、背景音乐自动避让人声。

新功能

• 文字与标题：标题、地点、日期、章节、片尾等 8 款预设，点一下加到播放头处；可调字体、颜色、描边、背景，有淡入、上浮、弹出、打字机 4 种动画；剪辑视频时跟着画面走
• 在预览画面里直接改字：暂停时双击文字或字幕，原地输入修改
• 多条音频轨：配乐、旁白、音效可以同时播放，时间重叠时自动放到新轨，上下拖动换轨；整条轨可静音、调音量
• 录旁白：从播放头开始边看画面边录，录完自动放到音频轨上
• 人声避让：有人说话时背景音乐自动调低。说话的时间段可以按字幕或声音生成，也能在时间线上手动添加、拖动；调低多少、渐变多快都能调
• 分离音频：把视频原声拆到音频轨上单独剪辑、调节
• 字幕样式移到右侧，选中字幕即可调整
• 最近项目可以右键移除、在访达中显示，或一键清空（只清记录，不删工程）

修复

• 拖动修剪后又拖回原位时，可能丢掉一条字幕
• 拖动修剪的途中识别完成，新字幕会被旧字幕覆盖
• 导出 SRT 时会覆盖同名文件；视频导出失败也会写出 SRT
• 预览里字幕框的缩放手柄会闪烁

界面目前只有中文。

升级说明

第一次录旁白时，系统会请求麦克风权限。录音保存在工程文件里；从时间线和素材库删掉、不再使用的录音，会在关闭工程时移到废纸篓。

安装

1. 下载下方的 FrameCut-0.1.2.dmg，打开后把 FrameCut 拖进「应用程序」（覆盖旧版本即可）。
2. 第一次打开时，系统会提示无法验证开发者，点「完成」。
3. 打开「系统设置 → 隐私与安全性」，点「仍要打开」并输入密码。

系统要求

macOS 14 或更新版本，Apple 芯片（M1 及更新的 M 系列芯片）。安装包 2.6 MB，安装后 8.3 MB。

校验

SHA-256：13ac57d465048df9526a5783e789280ebd5544fed3145c89947f4f2346b5bae5

---

English

This release fills in the text and sound a vlog needs: titles, multiple audio tracks, voiceover, and music that ducks under speech.

New

• Text and titles: 8 presets for titles, locations, dates, chapters and end cards, added at the playhead in one click; choose font, color, outline and background, with fade, rise, pop and typewriter animations; text moves with your video edits
• Edit text right on the preview: while paused, double-click a title or caption and type
• Multiple audio tracks: music, voiceover and sound effects play together; overlapping audio goes to a new track automatically, drag up or down to change tracks; mute or set the volume of a whole track
• Record a voiceover from the playhead while watching the picture; it lands on an audio track when you stop
• Ducking: the music lowers automatically while someone speaks. Speech ranges are generated from captions or from the audio, and you can add or drag them by hand on the timeline; set how far the music drops and how fast it fades
• Detach audio: move a clip's original sound to an audio track to edit it on its own
• Caption style now lives in the inspector on the right; select a caption to adjust it
• Recent projects: right-click to remove one or show it in Finder, or clear the list (projects themselves are not deleted)

Fixed

• Dragging a trim and then back to where it started could lose a caption
• Captions finished during a trim drag could be overwritten by the old ones
• Exporting SRT overwrote a file with the same name, and wrote the SRT even if the video export failed
• The resize handle on the preview caption box flickered

The interface is currently available in Chinese only.

Upgrade notes

The first time you record a voiceover, macOS asks for microphone access. Recordings are stored inside the project; recordings no longer used anywhere in the project are moved to the Trash when you close it.

Install

1. Download FrameCut-0.1.2.dmg below, open it, and drag FrameCut into Applications (replacing the old version).
2. On first launch, macOS says it cannot verify the developer. Click Done.
3. Open System Settings → Privacy & Security, click Open Anyway and enter your password.

Requirements

macOS 14 or later, Apple silicon (M1 or later). 2.6 MB download, 8.3 MB installed.

Checksum

SHA-256: 13ac57d465048df9526a5783e789280ebd5544fed3145c89947f4f2346b5bae5

FrameCut 0.1.1

中文

这一版主要改进字幕：识别更准，编辑更顺手，剪辑视频后字幕不再错位。

新功能

• 识别更准：语音模型换成 Whisper large-v3 turbo，中文错字率从 7.2% 降到 2.2%（实测）；可选识别语言（中文 / 英文 / 自动）
• 选择声音来源：可以识别视频原声、音频轨里的旁白，或两者一起
• 在源处编辑字幕：时间线上的字幕块和字幕列表都能右键分割、与下一条合并、删除，也可以一键清空；选中字幕按 ⌘B 分割
• 字幕跟随视频剪辑：删除、修剪、变速、调整顺序或加转场后，字幕跟着画面走；可在字幕页菜单关闭
• 在预览画面上调字幕：暂停时直接拖动字幕调整位置（靠近中线自动居中），拖右下角或双指捏合调整大小
• SRT 字幕文件：导入、导出 SRT；导出视频时可选择字幕是否画进视频、是否同时导出 .srt
• 时间显示带上小时，如 00:00:24.35

修复

• 识别长视频时偶尔整段没有字幕（如开头 20 多秒缺失）
• 时间重叠的两条字幕会叠在一起显示
• 59.996 秒显示成 00:60.00

界面目前只有中文。

升级说明

语音模型换了新的，第一次识别字幕时会重新下载约 600 MB，加载模型需要一两分钟；之后每次加载约 10 秒。旧模型会自动删除。

安装

1. 下载下方的 FrameCut-0.1.1.dmg，打开后把 FrameCut 拖进「应用程序」（覆盖旧版本即可）。
2. 第一次打开时，系统会提示无法验证开发者，点「完成」。
3. 打开「系统设置 → 隐私与安全性」，点「仍要打开」并输入密码。

系统要求

macOS 14 或更新版本，Apple 芯片（M1 及更新的 M 系列芯片）。安装包 2.4 MB，安装后 7.4 MB。

校验

SHA-256：b2c5d0dad01c304c7022aaca188106c98d14e88336eb28babcc409cbc2e0e3c0

---

English

This release is all about captions: more accurate recognition, easier editing, and captions that stay in sync when you edit your video.

New

• More accurate captions: the speech model is now Whisper large-v3 turbo; the Chinese character error rate dropped from 7.2% to 2.2% in our tests. Choose the spoken language (Chinese, English or auto-detect)
• Choose the audio source: transcribe the original audio, a voiceover on the audio track, or both
• Edit captions where they are: right-click a caption on the timeline or in the list to split, merge with the next one or delete, or clear them all; press ⌘B to split the selected caption
• Captions follow your edits: delete, trim, speed up, reorder or add transitions and captions move with the picture; can be turned off in the Captions menu
• Adjust captions on the preview: while paused, drag a caption to move it (it snaps to center), drag its corner or pinch to resize
• SRT files: import and export SRT; when exporting video, choose whether to burn in captions and whether to save an .srt alongside
• Time display now includes hours, e.g. 00:00:24.35

Fixed

• Long recordings could occasionally lose whole stretches of captions (for example the first 20-odd seconds)
• Two overlapping captions were drawn on top of each other
• 59.996 seconds was shown as 00:60.00

The interface is currently available in Chinese only.

Upgrade notes

The speech model has changed, so the first time you generate captions FrameCut downloads about 600 MB and takes a minute or two to load the model; later loads take about 10 seconds. The old model is removed automatically.

Install

1. Download FrameCut-0.1.1.dmg below, open it, and drag FrameCut into Applications (replacing the old version).
2. On first launch, macOS says it cannot verify the developer. Click Done.
3. Open System Settings → Privacy & Security, click Open Anyway and enter your password.

Requirements

macOS 14 or later, Apple silicon (M1 or later). 2.4 MB download, 7.4 MB installed.

Checksum

SHA-256: b2c5d0dad01c304c7022aaca188106c98d14e88336eb28babcc409cbc2e0e3c0

FrameCut 0.1.0

中文

FrameCut 的第一个版本：为 Apple 芯片 Mac 打造的 vlog 剪辑软件。免费、无水印、全部本地运行。

亮点

• 剪辑：分割、排序、修剪、变速（0.25×–4×，变速不变调）、转场、背景音乐、本地语音识别字幕
• 调色：10 款风格滤镜或导入 .cube LUT；高光、阴影、自然饱和度、HSL 分色、锐化、暗角；复制 / 粘贴调色、对比原图、保存预设
• HDR：自动识别 iPhone 拍摄的杜比视界 / HLG，全程按 HDR 处理，可导出 HLG、HDR10 或 SDR
• 速度：编解码用媒体引擎、合成用 GPU、语音识别用神经网络引擎；M2 实测 3 分钟 vlog 导出 1080p 约 26 秒、4K 约 95 秒
• 体积：安装包 2.3 MB，语音模型第一次生成字幕时才下载

界面目前只有中文。

安装

1. 下载下方的 FrameCut-0.1.0.dmg，打开后把 FrameCut 拖进「应用程序」。
2. 第一次打开时，系统会提示无法验证开发者，点「完成」。
3. 打开「系统设置 → 隐私与安全性」，点「仍要打开」并输入密码。只需要做一次。

系统要求

macOS 14 或更新版本，Apple 芯片（M1 及更新的 M 系列芯片）。

校验

SHA-256：dd0199cf68ac975b2888ed3179b5ebb47f12f0e3ba5b0f3c42c820ea07d14ef9

---

English

The first release of FrameCut, a vlog editor built for Apple silicon Macs. Free, no watermark, fully on-device.

Highlights

• Edit: split, reorder, trim, speed (0.25×–4×, pitch preserved), transitions, background music, on-device captions
• Color: 10 looks or your own .cube LUTs; highlights, shadows, vibrance, HSL, sharpen, vignette; copy/paste grades, compare with the original, save presets
• HDR: detects iPhone Dolby Vision / HLG and keeps it HDR end to end; export HLG, HDR10 or SDR
• Speed: codecs on the media engine, compositing on the GPU, captions on the Neural Engine; on an M2 a 3-minute vlog exports in about 26 s at 1080p and about 95 s at 4K
• Size: 2.3 MB download; the speech model is downloaded only when you first generate captions

The interface is currently available in Chinese only.

Install

1. Download FrameCut-0.1.0.dmg below, open it, and drag FrameCut into Applications.
2. On first launch, macOS says it cannot verify the developer. Click Done.
3. Open System Settings → Privacy & Security, click Open Anyway and enter your password. You only need to do this once.

Requirements

macOS 14 or later, Apple silicon (M1 or later).

Checksum

SHA-256: dd0199cf68ac975b2888ed3179b5ebb47f12f0e3ba5b0f3c42c820ea07d14ef9