Video editor: narration, visuals, captions and music

Magic Video, Shorts and Voiceovers share one editor, which keeps narration, audio, visuals and timing in sync. To see which Product fits, or what differs between them, read Choosing a video product.

The preview sits beside the narration editor, split by a draggable divider. On narrower windows the preview moves above the narration. Changes save automatically, and the preview updates without rendering an export. Generating or regenerating narration and visuals does not start a render. To share or download the result, see Video library and sharing.

Original prompt

Choose Prompt beside the video title to read the original brief, then Copy prompt to copy it. Videos and Shorts also offer Use again, which opens the brief in the creation page. Choose Use prompt to replace your current draft's text and sources, or Keep current draft to leave it alone. Your selected creation settings stay in place, and generation starts only when you choose Generate.

Videos made from an uploaded script do not show a prompt. Voiceovers show a prompt only when you supplied narration guidance.

Edit Narration

The narration editor holds the whole script in paragraphs that match the audio segments. Text edits save on their own but never silently replace finished audio.

• Split or delete: Press Enter partway through a paragraph to split it, or delete a paragraph's text. Neither marks anything stale: the paragraphs keep their audio and the visuals over them. Deleting a paragraph removes its audio and closes the gap in the timeline. Undo brings it back.
• Reword: Rewording a paragraph marks only that paragraph stale. Use its Regenerate Audio button to update it. Each stale paragraph regenerates on its own.
• Change the Character: Every paragraph shows its Host, the Character speaking it. Choosing a different one changes that paragraph alone, and it stays stale until you regenerate it in the new Voice.
• Pronunciation: Use the pronunciation control in the narration editor to fix how a word is spoken, as described in the pronunciation guide.

When you render, download or share with stale narration, Jellypod asks whether to export the current audio. To update it first, cancel and generate audio in the editor. The dialog is described in Choose which audio to render.

A paragraph with spoken text but no audio, including one the voice provider refused, shows a highlighted Regenerate Audio button (Generate Audio if it has never been voiced); hovering it names the reason when one is known. If the provider refuses the wording again, edit it or switch the Host before retrying. Empty paragraphs can stay in your draft and do not need audio. You can keep editing visuals meanwhile. You can generate the missing audio in the editor or choose to export with the audio already on your timeline. Rendering with existing audio leaves the missing lines unvoiced. A previously rendered version stays downloadable. If audio is already generating, wait for it to finish and try again: Jellypod does not queue the request.

You can type audio tags such as [pauses] or [laughs] in square brackets, the same way as in the episode script editor. In Magic Video and Shorts, Jellypod also writes tags for you, but only when every Host's Voice performs them. If you change a paragraph's Host to a Voice that does not support tags, leftover tags turn into warning-colored chips. Click one to open Audio Tags Unsupported, where you can edit the Character's Voice or clear the tags from that block.

Edit Visuals

The Visuals lane, the first lane in the timeline, shows each shot as a strip of thumbnails. Click a visual to select it and seek to that moment. Seeking never starts or stops playback.

Hover over the edge between two visuals to reveal a resize handle. Drag it, or focus it and press Left or Right (Shift moves one second), to shift time between neighbors. The first visual's start is fixed. Every other boundary moves, including the end of the last visual: drag it past the end of the narration to hold that shot longer, which lengthens the Video. Dragging back stops at the end of the narration, so the closing words always have a visual.

Right-click a visual, or select it and choose Edit visual, to open one dialog:

• Regenerate: Choose Regenerate to open the prompt the visual was drawn from. Edit it or write a new description, then confirm. The button shows the cost, and how credits are charged is covered in Understanding credits. The text you submit becomes the visual's prompt, the new visual saves itself when it arrives, and undo restores the previous one. A visual you uploaded opens an empty prompt. A visual with no generated shot behind it has no Regenerate and can only be replaced.
• Replace: Drop an image or video onto the visual in the dialog, or click it or Choose file. The file previews in place, and Save image or Save video saves it. The button stays disabled until a file is staged.

If the image model declines a shot for its content, Jellypod says so instead of retrying. Edit the description, then regenerate.

Drag and drop on the lane

Dragging an image or video over the Visuals lane previews the result before you release. Over the middle of a visual, it is outlined and marked Replace. Near either end, a vertical line snaps to the nearest boundary and dropping there inserts a new visual. The new visual takes its time from the neighbor on that side, so the rest of the lane stays put and the Video keeps its length. An image splits that neighbor in half (a 6 second visual becomes two of 3 seconds). A video keeps its own length, trimmed only when the neighbor cannot spare that much. Neither side drops below one second, so a visual shorter than two seconds offers no boundary to drop on. A visual inserted between two that dissolve into each other picks up the same transition. Dropping onto an empty lane adds the first visual.

A GIF works like an image. A browser that cannot decode a long GIF one frame at a time (Safari) shows a no-preview notice or a warning when you add one, and rejects a GIF too large to open. Use Chrome or Firefox, or upload an MP4. Rendering and exports are not affected.

Reframe and delete

Select an image and choose Reframe to open a larger editor. Drag the image beneath the fixed frame, adjust zoom, fill the frame, or fit the whole image without stretching. Voiceover visuals can also return to Jellypod's automatic framing. Saving changes only that image's appearance, even if another visual uses the same source. Reframe and effects are not available for the styles that draw each visual on their own canvas.

Right-click a visual and choose Delete visual. A confirmation appears first, the neighboring visual extends to fill the gap, and undo restores it.

Regenerating, replacing or deleting a visual leaves narration and timing unchanged. While one visual regenerates you can keep editing, but rendering waits until it lands.

Transitions

Click + between two visual cards to add a transition, then choose its style and duration. An existing transition shows an icon there. Styles are Fade, Push cut, Slide, Wipe, Flip, Clock wipe and Iris. Magic Video and Shorts add some transitions during generation, matched to the visual style, and every Voiceover visual after the first starts with one. Edit, replace or remove any of them the same way.

A transition belongs to the incoming visual on its right and moves with it when you reorder. Moving a visual to the first position removes its transition. Transitions never shift narration, visual timing or total duration.

Visual effects

A selected generated visual shows a pill at its bottom left naming its effect: Zoom in, Zoom out or No effect. Click the pill, or right-click and choose Edit effect, to pick one and set Strength, how far the zoom travels, from 25% to 300% in 25% steps. Strength is disabled with No effect.

An effect overrides the automatic alternating zoom for that one visual. It survives regenerating the rest of the Video and clears only if you replace that visual. Effects are not available on animated shots or on the five drawn styles above.

Timeline

Narration and music sit in their own lanes beneath the Visuals lane, on the track-based timeline shared with podcasts: drag, trim, delete, adjust volume and fades, and zoom in or out. Music segments can also be cut, copied and duplicated. A generated narration segment cannot, because each paragraph has one take, so its menu offers only Regenerate Audio, Edit Audio and Delete. Visual timing reflows automatically when narration audio changes, so visuals stay lined up with the words.

Reflowing does not redraw pictures. If the narration under a visual changes enough that it likely no longer fits, the visual gets a Needs regenerating badge. It keeps playing and exporting until you regenerate it.

Captions, music and watermark

Controls for these sit in the workspace beside the preview and timeline.

• Captions: Click Captions to preview the caption styles, pick one, or choose No Captions. The change applies to the preview and the final render. New Shorts and Magic Videos start with the caption style linked to their Visual Style.
• Music: Use the music control (Add Music) to set, swap or remove the background music. Shorts can start with a generated instrumental score. Magic Video and Voiceovers start without music, so add your own here. Music then edits like any other timeline clip, as described above.
• Watermark: Click Add logo to place a logo in a corner of the preview and final video. Plan availability, image format, size and corner options are in Video creation settings.

Unfilled visual slots

If generation could not produce a visual for a shot, for example when the image model declined the prompt, the rest of the Video is still committed. The empty shot shows a dashed Missing visual placeholder. Click it to open the dialog above and regenerate or replace it.

A banner above the preview counts missing shots and offers Review visual (or Review visuals), which jumps to the first one. Rendering waits until every shot has a visual, and a previously rendered version stays downloadable. If the whole generation failed rather than one shot, see Video generation failures.

Missing files

If a clip's stored file has been removed from your media library (image, video, GIF or music), the timeline badges it File missing and names the file. The Video keeps playing and skips that clip, but rendering, downloading or sharing fails and names the same file rather than publishing without it. Replace the file to clear the badge.

Open Help & Support in the Studio sidebar to chat with the team, or email support@jellypod.com.

Copy a Video to a Podcast

1. Open the Video and click the ••• menu in the top bar, beside Share and Download. This menu also contains Import script and Export script, which work as described in Use your own script.
2. Choose Copy to Podcast, then select the destination Podcast.
3. Click Copy to Podcast to create a separate, editable Episode.

The Episode keeps the Video's script, saved narration, speech timing and music. Copying does not regenerate audio or charge generation credits. Changes to either copy do not change the other.

To go the other way, create a Magic Video from an Episode.