Skip to main content

Processing Times & Queues

Learn what to expect from CHAMELAION's video translation times, how Lip-Sync affects speed, and how to work efficiently while your videos are translating.

K
Written by Konstantin Dorndorf
Updated over a month ago

What to expect

Processing time scales with your video length and the options you choose. On average, base translation takes about 1 - 2 minutes per minute of video. If you enable Lip-Sync, plan for roughly 6Γ— - 10x the video length per language.

For example, a 2-minute clip, translation only, about 2 - 4 minutes, but with Lip-Sync it will take about 12 - 20 minutes πŸ“Ή.

Why does it take time?

Behind the scenes, CHAMELAION has real work to do for each language you request:

  • Transcribe your source speech for all speakers,

  • Translate the transcript,

  • Synthesise new audio in your cloned voice,

  • If Lip-Sync is on, track faces and align mouth movements to the new audio so it looks natural.

Typical processing times

  • Base translation, no Lip-Sync: about 1Γ— - 2x video length.
    A 10-minute video usually takes about 10 minutes per target language.

  • Translation with Lip-Sync: about 6Γ— - 10x video length.
    A 10-minute video with Lip-Sync usually takes about 60 - 100 minutes per target language.

  • All other options, like translation style, background sound retention and so on, barely affect the processing times 😎.

Tip: For the first pass, translate one language without Lip-Sync, review and fix any transcript issues, then re-generate and add Lip-Sync. This saves tokens and avoids re-processing lines you would change anyway. For more information, visit our Best Practice collection of articles πŸ’―.

Queues, at a glance

When many jobs are running at once, your task may wait briefly in a queue before it starts. You will see progress move from waiting to active processing, then to finished, indicated by the progress bar ⏱️.

FAQ

Can I close the tab while it processes?
Yes, processing continues server-side. Come back later to see the status πŸ‘.

Why does Lip-Sync take so much longer?
This is where CHAMELAION has to use all its muscles πŸ’ͺ. It adds computer vision work on top of audio generation; the model tracks faces and aligns lip movements to the new speech, which is why the guideline is about 6Γ— video length.

Does the number of target languages change the total time?
All the languages you select are translated simultaneously, so the total time needed stays the same. The only thing that affects the total time are added features like Lip-Sync.

How do I know when my Video is translated?

No one wants to look at one page for an extended period and just wait for a progress bar to finish, we understand that. Therefore, you will get a confirmation email once CHAMELAION has successfully translated your Video πŸŽ‰.

Done! You now know how long things usually take, why they take that time, and how to work with queues for a smooth run πŸ‘.

Did this answer your question?