How to make a video of a sketch turning into finished art

Last updated 2026-08-16 · 6 min read

What is the short version?

Artists already photograph their process and almost never post it, because assembling the pieces into something watchable is a separate skill from making the art. The material is usually fine. It is the ordering, the cropping and the pacing that stop it becoming a video.

1. Go to Vizard Agent and upload your sketch photos, progress shots and any clips.
2. Say the format — an Instagram story, vertical, fifteen to twenty seconds.
3. Say how it should feel: calm and satisfying, or fast and energetic.

What do you need before you start?

The photos and clips you already took. Vizard Agent sorts, orients, crops and orders them itself, so what you supply is the destination and the mood, plus the finished piece at the best quality you have, since that is the shot the whole video exists to arrive at.

• The progress material. Sketch, mid-stage, finished. More stages make a better sequence.
• Any video you filmed. Even shaky handheld footage of cutting or inking is useful movement.
• The finished piece, properly photographed. This is the payoff frame.
• The format. A vertical story is the usual target and the crop matters more than people expect.
• The mood. This is what picks the music, and the music sets the pace of the whole edit.

What do you type into Vizard Agent?

Say what the video shows and where it is going. Vizard Agent orders the stages from looking at the images rather than from filenames, so a short brief with the right destination beats a long one describing shots.

Prompt

Variants worth knowing:

• With or without your own audio. Vizard Agent checks whether your clips have usable sound before deciding to score over them.
• A longer cut. Thirty seconds gives each stage room; a story-length cut is tighter and usually stronger.
• Text on screen. Ask for a title card if the piece has a name or a commission story.

What does Vizard Agent actually do?

Here is the order Vizard Agent worked in on a real process video. Two steps stand out: it distrusted the orientation of the extracted frames and checked, and it chose the music by measuring the tracks rather than by reading their descriptions.

1. Inspects everything uploaded and probes the video's format and audio.
2. Builds contact sheets of the images and of the video frames, and looks at both.
3. Checks the audio levels and whether there is any speech in the source clips.
4. Searches the music library and downloads several candidates.
5. Plots the waveforms of the candidates and analyses their tempo and beat timings.
6. Verifies how ffmpeg is auto-rotating the phone footage and checks a frame is really upright.
7. Extracts correctly oriented shots and tests vertical framings of the stills.
8. Reviews a montage of candidate close-ups to pick the best cutting moments.
9. Calculates the transition offsets against the beat grid and composites the crossfades.

Step 6 is the unglamorous one that saves the video. Phone footage carries a rotation flag that tools handle inconsistently, and Vizard Agent confirmed which way was up before cropping rather than discovering a sideways shot at the end.

What does the result look like?

From the run this page is written from, probed on the delivered file: 1080x1920, H.264, 30fps, 16.9 seconds, AAC audio. A vertical story just under seventeen seconds, scored, with the crossfades landing on the music's beats rather than at even intervals.

Ask for a longer version in the same conversation and Vizard Agent re-times against the same beat grid, so the second cut is a re-pace rather than a fresh build.

Seventeen seconds is short on purpose. A process video works by arriving at the finished piece before the viewer decides to scroll, and the stages before it are there to earn that last shot rather than to document the work in full.

This category was not separately measured, so treat the timing as a range rather than a promise: comparable work runs a median of 28 to 38 minutes end to end. Across all projects the median cost by tier is Flash 47, Pro 55, Max 242, Ultra 263 credits.

When does this not work well?

Everything here comes from the pictures you took while working, and most people photograph the beginning and the end and nothing in between. Vizard Agent sequences and paces what exists beautifully, and it cannot show a stage of the work that was never photographed.

• Two stages is not a process. A sketch and a finished piece is a before-and-after, not a video.
• Changing light between shots shows. Photos taken across several days at different times of day jump visibly.
• A moving hand-held camera fights the crop. Vertical framing of wide handheld footage loses the edges.
• The finished piece deserves a proper photograph. A bad final shot undoes everything before it.
• This is not a time-lapse. If you want the drawing to appear stroke by stroke, that has to be filmed that way.

How do you fix a result that came back wrong?

Name the stage or the pacing. Vizard Agent keeps the oriented frames, the beat analysis and the chosen shots, so reordering or slowing a section works against the timing grid it already measured rather than being re-derived by ear.

• "Hold on the finished piece longer." A timing change on one shot.
• "Use the other music." Re-timed to the new track's beats automatically.
• "Start with the sketch, not the materials." A reorder, not a re-edit.

How does Vizard Agent compare to doing it yourself?

By hand this is a mobile editing app, twenty photos in the wrong order, a music track that does not quite fit, and transitions placed by eye. It is doable in an evening and it is the evening most artists would rather spend making work.

Common questions

How many photos do I need? Enough to show change. Five or six stages gives Vizard Agent a real sequence to build from.

Can it use my video clips too? Yes, and it will check them for usable audio before scoring over them.

Will it fix sideways phone footage? Yes. Vizard Agent verifies the true orientation before cropping anything.

Can I choose the music? Yes. Upload a track or describe the mood, and Vizard Agent measures its tempo either way.

Can it add a title card? Yes. Ask for one and Vizard Agent styles it against the fonts it has available.

Can I get a longer version for another feed? Yes. Ask in the same conversation and Vizard Agent re-paces against the same beat analysis rather than starting again.

Do the photos need to be in order? No. Vizard Agent reads them as contact sheets and works out the sequence from what the pictures show.

Can it include a commission or client credit? Yes. Say the wording and Vizard Agent places it as a title card at the start or the end.