对比两稿剧本
在你的浏览器里免费做逐场对比
把两稿剧本粘到下面，就能得到一份逐场的改动报告：哪些场次是新增的、哪些被删掉了，以及你重写过的那些场次里究竟改了哪几行。为需要 Final Draft 那个 ScriptCompare、却不想掏授权费的编剧、剧本统筹和制片而造。
通用的对比工具把剧本当源代码看，于是一个被重新折行的动作段落，或者一场从第 12 页挪到第 40 页的戏，读起来就是一整面墙的删除和插入。这个工具先按场景标题把两稿都切成场次，先用标题、再用内容相似度把场次配对，最后才逐行比对。被挪动的一场戏仍然是一场戏，报告读起来也像一份修订页备忘录：新增、删除、改动。
全部在你的浏览器里运行。你粘贴或载入的稿子从不会被上传、保存或发送到任何服务器。
把两稿都粘到上面，或者载入 .fountain 或 .txt 文件，就能看到改动报告。
剧本修订用色
剧本进入制作之后，每一轮修订都印在不同颜色的纸上，好让每个部门一眼就知道某一页属于哪一稿。这个顺序在美国电影和电视里是标准化的，所以统筹说一句「这部剧到蓝色了」，所有人都知道自打拍摄本以来正好落了一轮修订。
标准的修订颜色顺序是：白色（拍摄本），然后是蓝、粉、黄、绿、金、米、鲑鱼、樱桃。修订到樱桃之后还要继续的制作，会带上「第二轮」的前缀重新开始：第二轮蓝、第二轮粉，以此类推。分集电视很快就能把这一摞颜色烧完，改动大的长片也做得到。
- 1. 白色（拍摄本）
- 2. 蓝色
- 3. 粉色
- 4. 黄色
- 5. 绿色
- 6. 金色
- 7. 米色
- 8. 鲑鱼色
- 9. 樱桃色
- 10. 第二轮蓝
樱桃色之后，序列会带着「第二轮」的前缀重新开始（第二轮蓝、第二轮粉、第二轮黄），并按同样的顺序一直走下去，只要制作还在改稿。
右边距上的修订星号
自上一份彩色稿以来改动过的每一行，都会在修订页的右边距上带一个星号。剧组扫的是星号，而不是把整页重读一遍：服装组看一眼某条着装备注有没有挪动，副导演看一眼这场戏是不是变长了。上面的改动报告在数字上给出同样的信号，用加号和减号的行标记代替边距上的星号，这让你很容易写出那份随彩色稿一起发出去的修订页备忘录。
锁定页与带字母页
剧本一旦锁定，页码和场次号就不再移动，排期、拆解和通告单于是继续指向正确的内容。新增内容以带字母页的形式插入（32A 接在 32 后面），新增场次以带字母场次的形式插入（14A 接在 14 后面），而被删掉的场次保留自己的号码，标题下面只写「删除」两个字。正是这条惯例，让这个工具去报告被删掉的场次，而不是悄悄把它们丢掉：它照搬的是制作文件追踪删改的方式。
工作原理
工具在场景标题处把每一稿切成场次（包括以点开头的 Fountain 强制标题）。标题完全相同的场次先配成对，剩下的场次按词汇重合度配对，用的是下限为 0.4 的 Dice 相似度，所以一场被改了名却内容照旧的戏，仍然能找到它的对应。每一对随后在行级别上做最长公共子序列比对，并把空白折叠掉，于是重新折行的一行不会被算成改写。最后没配上对的会被报告为新增或删除，统计则汇总场次数、新增和删除的字数，以及剧本改动的比例。
Final Draft ScriptCompare 的免费替代方案
Final Draft 的 ScriptCompare 把这件事做得不错，但它藏在付费授权后面，还要求 .fdx 文件。如果你用 Fountain、Highland、WriterDuet 或纯文本写作，这个剧本对比工具是免费的，而且覆盖了真正要紧的那部分：它在场次层面告诉你两稿之间到底改了什么。粘上两个版本、读报告、复制进你那封笔记邮件，完事。不用注册、没有水印、没有试用倒计时。
一个通用的对比工具，或者像 git 那样的代码对比，比的是原始行，而剧本文本会把它们打败。文字处理软件会重新折行动作段落，所以加进去的一个词就可能把四十行标成改动。在重构时挪动一场戏，行级对比会在第 12 页显示一大片删除、在第 40 页显示一大片插入。懂场次的对比同时绕开了这两种失败，这也是为什么用这个工具在线比较剧本版本，得到的报告能在笔记电话会上直接念出来，而不必先翻译一遍。
剧本修订追踪，在别人开始依赖你的页面时最要紧。制片想知道结尾有没有挪，副导演想知道那场餐馆戏是不是变长了，而没有人愿意为了搞清楚这些去重读 110 页。每次发稿之前跑一遍对比，把摘要粘进封面备忘录，再配上下面那套修订颜色顺序，你的文件就和片场从制片厂时代起追踪改动的方式对上了。
使用场景
- 和制片开的笔记会。 开会前先比一遍旧稿和新稿，然后照着改动场次清单一条条过，而不是去替那些自上一版起没人重读过的页面辩护。
- 重写之后的自检。 重构之后，确认真正改动的只有你打算动的那些场次，也确认某次跑偏的查找替换，没有在别处悄悄改掉对白。
- 版本考古。 直接比对文件，把「终稿_v3」和「终稿_最终版」分清楚，看清哪一个才真的装着更晚的内容。
Script Draft Compare: the complete guide
It parses both drafts into scenes on their sluglines, aligns scenes by heading and then by content similarity, and produces line level diffs plus change stats, so reordered or renamed scenes are tracked instead of misread as mass deletions.
For this workflow, the central problem is clear: generic diff tools break on screenplays, so writers either pay for Final Draft's ScriptCompare or reread whole drafts to find what changed. Left unresolved, this creates downstream friction and slower decisions. The practical target is a scene by scene change report (added, omitted, changed scenes with line diffs and stats) you can paste into a notes email or cover memo.
Limitation to keep in mind: It compares text, not formatting: it cannot read .fdx files directly, does not paginate, and will not produce studio style colored pages or margin asterisks, so it complements locked page workflows rather than replacing them.
Advanced workflow: Script coordinators run the compare before each distribution, paste the summary block into the revised pages memo, and keep the downloaded .txt reports as a change log that survives file renames and email threads.
Step-by-Step Workflow
- Paste or load the earlier draft on the left and the newer draft on the right (Fountain or plain screenplay text).
- Read the stat strip first: scenes added, omitted, changed, unchanged, plus words added and removed and the percent of the script that changed.
- Expand each changed scene to review its line diff, with + for added lines and - for removed lines.
- Copy the report or download it as .txt and attach it to your notes email or distribution memo.
Use Cases By Profile
- Writer: verify a targeted rewrite touched only the intended scenes before sending the draft back to a producer.
- Script coordinator: build the revised pages summary for a colored distribution without rereading the full episode.
- Producer: see at a glance whether notes were addressed by checking which scenes actually changed since the last version.
Common Mistakes To Avoid
- Comparing drafts with a generic text diff, which reports reflowed paragraphs and moved scenes as massive rewrites.
- Skipping the compare step and trusting memory of what changed, which is how contradicted continuity reaches the stage.
- Renaming sluglines during a rewrite without checking the report, which can split one location's history across two scene identities in your notes.
Professional Best Practices
- Keep sluglines stable between drafts when you can: exact heading matches align instantly and make every report cleaner.
- Run the compare per act on long features by pasting one act at a time, which keeps the changed scene list short enough to discuss on a call.
- Save the .txt report next to each draft file so the folder itself documents what changed between versions.
Treat this tool output as a decision support layer, not a replacement for authorship. Great scripts are remembered for specific choices, emotional precision, and clarity of dramatic movement. Tools help by removing noise so your energy can go where it matters: character, conflict, escalation, and payoff. If you review outcomes after each pass and keep an explicit log of accepted changes, your workflow becomes faster and more predictable from draft to draft. That consistency is exactly what professional collaborators value: fewer surprises, clearer rationale, and a script that evolves with intent.
Extended FAQ
Can I compare script versions online without installing anything?
Yes. The tool is a web page that runs entirely in your browser. Paste two drafts or load .fountain or .txt files, and the scene by scene comparison renders immediately with no install, no signup, and no upload.
How does the scene matching handle renamed sluglines?
Scenes with identical headings pair first. Remaining scenes are paired by vocabulary overlap using a Dice similarity coefficient with a 0.4 threshold, so INT. DINER - DAY renamed to INT. ROADSIDE DINER - DAY still matches when the body is similar, and the report shows it as changed with the old heading noted.
What is the standard script revision color order?
White (the shooting draft), then Blue, Pink, Yellow, Green, Goldenrod, Buff, Salmon, Cherry. After Cherry the cycle restarts as 2nd Blue, 2nd Pink, and so on. The interactive stepper on the page tells you the next color from any starting point.
What do the asterisks in a script's right margin mean?
A revision asterisk marks a line that changed since the previous colored draft. Departments scan the asterisks on each revised page instead of rereading it, the paper equivalent of the + and - markers in this tool's line diffs.
Why do locked pages use numbers like 32A?
Once a script is locked, page and scene numbers freeze so schedules and call sheets stay accurate. New material is inserted as A pages (32A after 32) and deleted scenes keep their number marked OMITTED, so downstream paperwork never has to renumber.
How much of my script should change between drafts?
There is no correct number, but the percent changed stat gives context: a polish pass often lands under 10 percent, a notes driven rewrite commonly runs 20 to 40 percent, and anything above that reads as a page one rewrite to the people tracking your revisions.
常见问题
把较早的一稿粘在左边、较新的一稿粘在右边，或者载入 .fountain 或 .txt 文件。报告会立刻出现，包含新增、删除和改动的场次，以及行级别的比对。这个工具免费、不用注册，并且完全在你的浏览器里运行。
不能直接支持，因为 .fdx 是一个 XML 容器。在 Final Draft 里用「文件」再「另存为」导出纯文本，或者全选之后把剧本文字粘进这个工具。文字进来之后，对比的方式和 Fountain 文件完全一样。
改动场次指的是一对配上的场次，在空白被折叠之后行内容仍然不同。场次先按完全相同的标题配对，剩下的按内容相似度配对，Dice 分数至少 0.4，所以一个被改了名、但下面动作照旧的场景标题仍然能配上对，并显示为改动，而不是变成一场删除加一场新增。
不会。两稿都由运行在你浏览器标签页里的 JavaScript 解析和比对。你载入的文件由本地的文件读取接口读取，从不离开你的机器，当你在比对签了保密协议、尚未拍摄的材料时，这一点很要紧。
绿色。完整的制作顺序是白、蓝、粉、黄、绿、金、米、鲑鱼、樱桃，然后循环以第二轮蓝重新开始。上面修订颜色那一节里的选择器，能为任意起点给出下一个颜色。
