SceneAnvil Mark
SCENEANVIL
Clipping

The method

Turn a long video into short clips.

The job starts from speech. SceneAnvil transcribes a video you own, keeps up to 8 ranges, and burns one short caption line onto each vertical clip.

Upload a video
Short answer

Upload one MP4, MOV, or WebM you own. SceneAnvil transcribes the speech and returns a zip of up to 8 vertical MP4s. Each clip is 20 to 45 seconds. You post the files.

8Clips at most
20–45sLength of each clip
SpeechRequired in the file

Who this page is for

This page is for someone holding a long video and wanting short pieces from the words in it. A podcast recording, an interview, a webinar, a lecture, and a commentary stream all land here when the value is the speech.

Those sources share one job on SceneAnvil. The site does not run a different pipeline for each name. A separate page for each name would answer the same question.

What the file has to contain

The file needs an audio track and speech the transcript can read. A video with no audio track stops with “This video has no audio track.” A video with audio and no speech stops with “No speech was found in this video.”

The file also has to be one you own or have the rights to use. MP4, MOV, or WebM. Up to 1 GB. Up to 30 minutes. A longer file stops with “This video is longer than 30 minutes.” A larger file stops with “The video is over 1 GB.”

How a range is chosen

The transcript is a set of timed phrases. SceneAnvil keeps ranges of 20 to 45 seconds from those phrases, up to 8 of them. A range that cannot meet that length is dropped. If nothing usable remains, the job stops with “No usable clips were found in this video.”

There is no editor in this version. You do not drag the in and out points. The words on the clip come from the phrases inside the range.

What the caption line is

Each clip gets one short uppercase line, burned into the picture. The line is a few words from the transcript, and one of those words is marked. A transcript can mishear a name, a number, or a term. You cannot rewrite the line inside SceneAnvil.

The line is a caption of what was said in that range. It is not a title you type, and it is not a promise about how the clip will perform.

An example

Example. You recorded a 22-minute interview. Two people talk the whole way through, and the file has an audio track. SceneAnvil transcribes that talk and can return as many as 8 clips, each between 20 and 45 seconds, with one caption line from the words in that stretch.

You download the zip from the job page and post the MP4s yourself. This is an example of the job. It is not a customer story, and it does not predict views.

What you still do

One job runs at a time. The next file waits until the current one finishes or fails. The first file is free when it is 10 minutes or shorter, once per account. After that, a minute of the source costs one source minute, rounded up. The price for Pro is on the pricing page.

A job that fails before a finished zip returns the minutes, or the free file, once. A finished zip does not. The layout of the picture, frame or fill, is a choice you make before upload. Opening the clips in an editor is a separate decision.

The job stops in these cases

The job stops in these cases
What is in the fileWhat the job says
No audio trackThis video has no audio track.
Audio, and no speechNo speech was found in this video.
Longer than 30 minutesThis video is longer than 30 minutes.
Larger than 1 GBThe video is over 1 GB.
Nothing in the 20–45 second rangeNo usable clips were found in this video.

The steps

1

Start from a video you own

Use an MP4, MOV, or WebM with speech, up to 1 GB and 30 minutes.

2

Choose frame or fill

Frame keeps the whole picture with a blurred surround. Fill crops to the center.

3

Upload and wait for this job

One job runs at a time. SceneAnvil transcribes the file and keeps up to 8 clips of 20 to 45 seconds.

4

Download the zip

The clips stay on the job page. You post the files.

Questions

Does a silent screen recording become clips?

No. The job needs speech. No audio track, or audio with no speech, stops the job before a zip.

Can I choose the exact in and out points?

Not in this version. Ranges come from the transcript and stay between 20 and 45 seconds. There is no editor on the job.

Why might the words on the clip be wrong?

The line is taken from the transcript. A transcript can mishear. SceneAnvil does not give you a way to rewrite that line.

How many clips should I expect?

Up to 8. A short or sparse transcript can return fewer. Each kept clip is 20 to 45 seconds.

Does a failed job spend the free file?

A job that fails before a finished zip returns the minutes, or the free file, once. A finished zip does not.

Will SceneAnvil post the clips?

No. The zip is the end of the job. You choose the account and the time.

Who can open the stored file?

The job page is for the signed-in owner. The stored file address is a public storage URL, so anyone with that address can open the file. Source files, clips, and the zip are deleted about 14 days after upload. Cleanup runs once a day, so a file can remain a little longer.

A long video becomes short clips here only when it contains speech you have the rights to use. The zip holds up to 8 captioned vertical MP4s. You post them.

Related

SceneAnvil returns the zip. You post the files.