How to Make a Video Abstract for Your Research Paper

The paper is accepted and the journal's author guide mentions a video abstract. You like the idea, but you have no video editor, a full teaching load, and a paper you have already explained once in several thousand careful words.

The good news: a video abstract is small. Most publishers want a few minutes at most, and you already have the hardest part, the result. Here is a walkthrough you can follow this week.

Step 1: Pick the One Finding the Video Is About

A video abstract is not the paper read aloud. It is one finding, explained so that someone outside your subfield sees why it matters. Write that finding as one sentence first. If you cannot, the video will wander.

Stay inside the paper. Elsevier says a video abstract must be "within the conceptual scope of the article and directly support its conclusions", and it is subject to peer review.

Say your paper shows that a cheap soil sensor predicts crop yield about as well as a lab assay. The video is about that, not the six robustness checks. A shape that fits most papers:

1. Title card with the paper title and authors (5 seconds).
2. The problem, in plain words (20 seconds).
3. What you did (30 seconds).
4. What you found, shown on one or two figures (60 seconds).
5. Why it matters (30 seconds).
6. Where to read the paper (10 seconds).

Step 2: Write a Script Sized to the Length

Your journal's limit sets the word budget. Elsevier asks for no more than 5 minutes, IOP Publishing for no longer than four minutes, and Sage for less than three minutes.

My working assumption, and it is only an assumption, is that calm explanatory narration runs at about 130 words per minute, so a 3-minute video holds about 390 words. Time yourself reading a paragraph aloud and use your own number.

You want fewer than that, because figures need a moment of silence to land. Sage's guidelines suggest about 250 words for a final video under 3 minutes. Start there.

Write the way you talk. BMJ asks for a tone that is "relaxed and friendly", and Elsevier suggests speaking "as if you are addressing a class of students". Cut acronyms, and read the script aloud twice before you record.

Step 3: Turn Your Figures Into Visuals, and Check You May Reuse Them

You do not need animation. Sage suggests you "consider narrating over one or more of your article's figures". Pick the figure that carries the finding, crop to the panel that matters, enlarge the labels and highlight what you are talking about. IOP Publishing warns against small text that will be difficult to read, and a journal figure squeezed into a video frame is mostly small text.

Then check the licence. Under CC BY, anyone may share and adapt the material for any purpose, with appropriate credit, a link to the licence and an indication of changes, so put that credit on screen under the figure.

If you signed a copyright transfer, your agreement decides. Elsevier lists use at conferences and in classroom teaching among the rights authors keep, but a public YouTube video is a different use, so check or ask the permissions team. IOP Publishing states that material from previously published work requires permission from the copyright holder. Redrawing the chart from your own data is another option; ask if you are unsure.

Skip music. IOP Publishing does not allow music soundtracks, open source music included.

Step 4: Narration Over Figures, or You on Camera

Voice-over on figures is the quickest format and hides nerves. On camera puts a person behind the paper. A short on-camera opening followed by narration over figures gives you both.

Sound matters more than picture. Taylor & Francis advises somewhere quiet and a plug-in microphone, and BMJ warns against touching the microphone or hitting the table it sits on.

On camera, Sage recommends a tripod, with the main light source behind the camera rather than behind you, and BMJ says to keep the camera at eye level. Record one take per script section, so a stumble costs you 30 seconds, not 3 minutes.

Step 5: Add Captions and a Transcript

Plan for people watching with the sound off: put your one-sentence finding on screen as text at the moment you say it.

Captions also carry the content to people who are Deaf and others who cannot hear the audio. Sage adds closed captions to all video abstracts on publication, while Taylor & Francis requires a written descriptive transcript and IOP Publishing a transcript as a .txt file. Your script is most of that transcript already.

If you use automatic captions, check every line. W3C's guidance is that automatically generated captions do not meet user needs unless confirmed to be fully accurate, and your field's vocabulary is the first thing to check.

Step 6: Export to the Right Specs

Make one high-quality 16:9 master and export each version from it.

• Journal: MP4 with H.264 video and AAC audio is the safe default. Elsevier prefers MP4 (H.264+AAC, max target 720p), with 150 MB as the recommended upper limit; IOP Publishing caps files at 100 MB and Sage asks for below 10 MB.
• YouTube: MP4, H.264, AAC-LC audio, 16:9, with 8 Mbps recommended at 1080p.
• Vertical: YouTube treats square or vertical videos up to three minutes as Shorts, so cut the finding alone into a vertical version.
• Thumbnail: Elsevier asks for a representative still image from the video.

Step 7: Put It Where People Will See It

1. The journal, if it accepts one. Taylor & Francis medicine and health journals want a storyboard at submission, while BMJ asks only after acceptance. Our publisher guide to journals that accept video abstracts has the specs side by side.
2. YouTube, with the title, authors and DOI link in the description.
3. Your lab page and faculty profile, next to the paper.
4. Conferences, through a QR code on your poster.
5. Social posts, using the vertical cut.

Doing This Without a Video Editor

If that is a week you do not have, the alternatives are an agency or an automated tool. We compared what a video abstract costs each way.

Amplify is the automated option we built. It makes a 16:9 video abstract of about 3 minutes and a 9:16 vertical short of 45 to 60 seconds from the PDF you upload and nothing else, so there is no filler from the open web. Each comes as three variants you can edit line by line or regenerate, a person reviews every format before release, and nothing publishes until you approve it. Every video carries the title, the authors and a link back to the paper of record, and you keep ownership.

A video abstract will not replace the paper, and it should not try. What it can do is give the right reader a three-minute reason to open the paper. Start with the one sentence from Step 1; everything else follows from it.

See what Amplify makes from one paper