Twelve free browser tools for a documentary workflow: what each one actually does

You can reduce small production mistakes before choosing a full video service: estimate narration time, repair a constant subtitle delay, convert plain caption formats, check chapter timing and organize source credits. Topicut provides twelve browser utilities for those bounded jobs. They do not transcribe audio, align individual words, generate footage or create a finished video.

The tools do not require signup or upload the supplied input to a model service. That makes them useful when you already have a script, caption file or source list and need a deterministic transformation. Public Topicut self-serve rendering and checkout are not open yet. This guide describes the working utilities, not an early launch of the video engine.

Choose the task before choosing the tool

Start by identifying the input you actually have. A script can support a timing estimate; it cannot establish the length of an unrecorded performance. An existing subtitle file can be converted or shifted; it cannot tell the utility what the speaker really said. A list of source fields can be formatted without proving permission.

Keep an original copy before every transformation. Name the output according to what changed, and inspect the result before using it in an edit. A successful calculation or download means the tool completed its stated operation, not that the entire video is ready to publish.

For a reproducible first run, use the tutorial centre. Its examples are deliberately synthetic. They let you compare expected output without mistaking practice data for reviewed client content, licensed footage or a completed production project.

The tool map

These are separate jobs. Combining their outputs can support a workflow, but it does not turn them into a research agent, speech recognizer or rendering application. Use the tool whose stated assumptions match your current input.

1. Words to minutes: plan, then calibrate

Words to minutes estimates a range using rates you choose. Non-Han word tokens and Han characters use separate rate assumptions. It is not a general language-aware tokenizer and does not measure pauses, pronunciation or a generated voice.

The script runtime tutorial uses ten words at 120–150 words per minute, yielding four to five seconds. Those are arithmetic assumptions, not a speaking recommendation. Read a representative passage aloud, time it and adjust the rates before applying the estimate to a longer script. Add visual holds and silence separately.

2. SRT ↔ VTT: convert supported plain captions

The converter handles a deliberately restricted plain-text subset. It refuses unsupported styling and richer features rather than pretending they survive the conversion. Keep the original if the caption file contains information beyond simple cue text and times.

After conversion, inspect the header, timestamp punctuation, cue text and numbering. Open the result in the intended player before replacing a production caption track. Two formats can carry similar content while the destination application interprets them differently. A syntactically valid output is not a complete playback acceptance test.

3. Subtitle time shifter: fix one constant offset

The time shifter applies the same offset to every cue. Positive values move captions later; negative values move them earlier. A cue that would cross the supported time boundary stops the export instead of disappearing silently.

Use the shift-and-convert tutorial to reproduce a positive 1.250-second change. Then check a phrase near the beginning and another near the end of your actual video. If the required corrections differ, you may have drift or a different edit. One offset cannot repair that reliably, and the utility does not listen to the audio.

4. Reading-speed checker: find review candidates

The reading-speed checker calculates cue-level findings from supplied captions and adjustable thresholds. Its CPS count includes spaces and punctuation, excludes line breaks and counts Unicode code points. That is a defined counting method, not a universal measure of human reading difficulty.

A flag asks for review. Check sentence meaning, vocabulary, timing and competing visuals before changing the text. A caption below the threshold can still be hard to follow. If you extend its duration, inspect the next cue for overlap; if you shorten the wording, verify that you have not changed the speaker’s meaning.

5. Subtitle to text: extract, not transcribe

Subtitle to text removes the caption timing structure from supplied cues and produces a readable text draft. It does not hear the recording, correct recognition errors or know whether repeated lines are intentional. Those limitations matter when using captions as research material.

Read the output alongside the relevant recording before treating it as an accurate transcript. Speaker changes, quotations and meaningful pauses may need manual annotation. Keep the original caption file so you can locate a passage again rather than losing its relationship to the source timeline.

6. Timing summary: inspect coverage without guessing runtime

The timing summary reports cue count, covered intervals, gaps and overlap information. Overlapping intervals are merged for coverage calculations. The last cue’s end is the end of the caption timeline, not a measured video duration.

Use the report to ask useful questions. A large gap might be an intentional silent sequence or missing captions. An overlap might be valid presentation or an editing error. The tool cannot decide which explanation applies. Compare the report with the actual video and record what you reviewed instead of automatically “repairing” every finding.

7. Line breaker: change layout while preserving cue times

The line breaker reflows text at word boundaries using the selected width. It retains cue timestamps but normalizes whitespace. An unbroken word exceeding the chosen width is rejected rather than truncated, so unusual URLs or identifiers may require manual attention.

Compare words, cue count and timing before using the new file. A more balanced shape does not prove a caption is comfortable to read or synchronized with speech. Follow layout review with a reading-speed check and playback. Do not use a formatting tool as a substitute for checking the wording.

8. Chapter formatter: check the final segment too

The chapter formatter checks the supplied list against basic timing conditions, including a zero start, enough chapters and minimum segment duration. The final segment uses the video duration you enter, so an incorrect runtime can produce an incorrect conclusion.

The chapter tutorial provides a passing 180-second example and a failing 125-second example with the same starts. Measure the final export before writing a production description. The tool does not fetch YouTube videos, inspect your account or guarantee the platform will display chapters.

9. Description builder: assemble your own material

The description builder combines your supplied summary, optional chapter list and source notes. It is not an AI writer and adds no new factual research. It also does not independently establish whether the pasted chapter list or source information is correct.

Use it after the final script and cut are stable. Review the opening summary, verify any claims and compare credits with the original records. If you change the video later, revisit the description instead of copying a previous version just because the title is similar. Formatting helps the handoff; it does not certify it.

10. Attribution builder: keep supplied uncertainty intact

The attribution builder accepts a title, creator, HTTPS source URL, stated license and optional HTTPS license URL. It turns those fields into a draft. It does not browse the links, identify the actual rights holder or decide whether commercial use is allowed.

Follow the source-credit tutorial with its fictional example before using real records. Preserve the check date, reviewer and asset version in your underlying source log. A readable credit line should never replace an unresolved field with an invented answer, and attribution is not automatic rights clearance.

11. Aspect-ratio calculator: plan geometry

The aspect-ratio calculator reduces dimensions to a ratio and helps plan a target size. It does not crop, resize or render an uploaded video. Use it to communicate the intended frame geometry before an editor changes the file.

When adapting a landscape project to another shape, check what the composition loses. Important labels, faces and diagrams may need a new layout rather than a mechanical crop. The arithmetic can be correct while the visual result is unsuitable, so a preview of the actual design remains necessary.

12. File-size calculator: estimate before measuring

The file-size calculator uses supplied duration, audio/video rates and an overhead allowance. It is a planning estimate, not a probe of your final export. Variable bitrate and different encoder behavior can change the measured result.

Keep units explicit. Decimal megabytes and binary mebibytes are not interchangeable labels. After export, record the actual byte length and use an appropriate hash comparison if your delivery process requires it. A plausible estimated size cannot prove that a large download completed successfully or that its content is correct.

A practical sequence for one video

Before recording, estimate the reviewed script and identify where visual holds or difficult names could change its pace. After captions exist, preserve an original, inspect timing coverage, correct only a demonstrated constant offset, and check the wording and reading speed. Convert formats only when the destination requires it.

When the cut is final, measure its runtime and validate chapter timing. Assemble a description from reviewed text and source records, then compare it with the delivered version. Check frame geometry during planning and actual file size after export. These are different checkpoints; completing one does not silently approve the others.

Keep the workflow lightweight. You do not need to run every utility on every video. Use the smallest useful tool and retain an understandable record of what changed. When a finding cannot be resolved by a deterministic transformation, hand it back to an editor rather than applying increasingly large automated corrections.

FAQ

Are the tools free without signup?

The twelve utilities described here are available through the free tools directory without registration. They perform bounded browser tasks on your input. Check the visible input limits and supported format subset before processing a production file.

Can they create captions from an audio recording?

No. They operate on existing text or caption data. Audio transcription, automatic speech recognition and individual word alignment are different capabilities and are not provided by these free utilities.

Can the attribution builder tell me an image is safe to monetize?

No. It formats supplied metadata into a draft and does not verify rights. Review the actual source and intended use. Neither the tools nor source records promise views, monetization or YouTube Partner Program acceptance.

Can I generate a finished documentary in Topicut self-serve?

Not yet. Public rendering and checkout remain closed. If the utilities are not enough, request a sample to discuss the brief, current availability and deliverables. An enquiry is not an order or a guaranteed production slot.

Start with one small task

Open the tool directory, follow a tutorial or look up an unfamiliar term in the glossary. The Blog covers broader workflow decisions. Use a real input, inspect the complete result and keep a copy of the original; a useful small check is more valuable than a promise that every part of production has been automated.

Related workflow

Keep a production brief with your decisions, and explore the Topicut guides for the next review step. The cover is an original Topicut editorial graphic, not a vendor interface or a measured comparison.