Chinese remains the source record of what was said.
PRODUCT PAGE
Source-Preserving Translation
Source-preserving translation means English helps the viewer understand Chinese speech without replacing the Chinese line that can be checked against the original video.
Sign in to uploadsource-preserving translation
OUTPUT PREVIEW
你吃饭了吗？
Have you eaten?
WHAT YOU GET
A focused path from speech to subtitle file.
English works as a meaning note rather than a replacement.
The user can catch fluent but inaccurate translation by comparing layers.
The same file can support comprehension and learning.
HOW IT WORKS
How this capability fits the SubChinese workflow.
- 01
Start from a Chinese video the user already has. The product begins with the source rather than assuming a useful platform caption track exists.
- 02
Create timed Chinese segments. Each sentence becomes an addressable unit that can be replayed, checked, corrected, and exported.
- 03
Connect the reading and meaning layers. Pinyin belongs to the Chinese characters, while English remains an optional interpretation layer.
- 04
Choose the output that matches the job. SRT carries portable timed language. ASS carries more deliberate learning layout.
SPECIFICATIONS
The capability in context.
| Layer | User-facing role | What it helps protect |
|---|---|---|
| Chinese source | The written record of speech | Accuracy and accountability |
| Pinyin | The reading path | Pronunciation-to-character connection |
| English | The meaning check | Comprehension without source replacement |
| Timing | The playback structure | A phrase’s place in the video |
| Export | The portable result | Reuse beyond the browser |
Why English should open the source instead of covering it.
Translation is useful because it reduces uncertainty. It becomes dangerous when it is the only visible version of the sentence. A fluent English line can hide a recognition error, smooth over a cultural reference, or choose a natural phrase that no longer shows what the Chinese actually did.
SubChinese treats English as a meaning layer. The user can hide it during focused watching, reveal it when needed, and keep it in exports when the file is meant for comprehension. The Chinese remains the line that carries accountability.
This also protects editing. If the Chinese source is corrected, the English can be reviewed as an interpretation of the corrected line. Without the source, the user is editing a translation without knowing whether the underlying sentence was right.
Why the capability matters in real content
Real Chinese videos are not arranged like beginner material. They contain names, unfinished sentences, cultural references, casual reductions, repeated phrases, and changes in speaking speed. A useful product has to support understanding without pretending that every line can be reduced to a perfect textbook example.
That is why SubChinese keeps the source, timing, reading support, and meaning support connected. The user can decide which layer deserves attention at a particular moment: listen first, read the Chinese, check pinyin, reveal English, or return to the same phrase later.
The result is a more respectful kind of assistance. It does not replace the content that made the video valuable. It gives the viewer a clearer way to approach it.
Why this is a product capability, not a marketing label
A capability should be visible in the user workflow. It should explain an action the user can take, a problem the user can avoid, or a result the user can keep. The SubChinese capability pages are designed around that standard: each concept appears in the preview, editor, playback, history, SRT export, or positioned ASS export.
This distinction also keeps the product language honest. We can describe a language layer, a reading relationship, or a frame-aware layout without claiming that automatic output is perfect. The value is in the structure that makes understanding and correction easier.
Why serious viewers need both source and meaning.
A casual viewer may only need the gist. A serious learner needs to notice how the Chinese expresses the idea. Source-preserving translation keeps that door open. It lets the user understand now while still being able to learn from the source later.
This is also useful for tutors and language schools. A teacher can ask students to compare the Chinese and English, discuss why a phrase was rendered a certain way, and focus class time on meaning rather than manual transcription.
USE YOUR OWN VIDEO
Sign in to uploadStart from a file, not an existing subtitle track.
Sign in first, then upload a local video, review the generated transcript, keep English optional, and download a subtitle file.
WHEN THIS IS NOT WHAT YOU NEED
When this capability is not the right emphasis.
- If the user only needs a one-time summary, a full subtitle language layer may be more structure than necessary.
- If the source does not contain Chinese speech, the Chinese-focused workflow is not the right fit.
- If the destination player cannot support the chosen layout, the user should choose the more compatible output.
The product should make the useful layer available without forcing every viewer to display every layer at once.
QUESTIONS
Questions about source-preserving translation.
Is this an implementation document?
No. It explains the user-facing product theory and capability. Internal infrastructure and provider details are intentionally not part of the public page.
Does automatic output still need review?
Yes. Names, context, homophones, segmentation, and translation nuance can require human judgement. The product makes review structured and reusable.
How do the capabilities connect?
They share the same timed language segments, so Chinese, pinyin, English, playback, editing, and export stay connected.