Every shot in a Didascalies flipbook is read through three lenses. The Making documents how the shot was produced — crew, light, lens. The Frame Speaks reads what the composition argues — axis, poles, thesis. The Bloodline traces the shot's descendants in cinema history.
Your team will apply this method to a film of your choice, one shot at a time, and produce a flipbook indistinguishable from the published Didascalies volumes.
Each team member takes a role. You can divide by shot (each person owns their shots end-to-end) or by register (one person writes all the Making, another all the Frame Speaks).
No shots yet.
Most film assignments let you stay at altitude — themes, arcs, vibes — and never land on anything a reader can check. This one makes that impossible. You stop the film, choose a single frame, and account for it: how it was made, what it argues, where it comes from. The claim sits two inches from its evidence.
What you carry out of this isn't facts about one film; it's a discipline of attention — the reflex of stopping on an image and asking the three questions that never date. It transfers to any image you'll work on later: a shot you direct, a cut you make, a film you program or review.
Worth knowing before you start: the published volumes you'll study as the standard — Potemkin, D.O.A., The General — were written with the help of artificial intelligence, working from extracted frames and the screenplay. The machine never watched the film at twenty-four frames a second, and never heard a note of it.
It reads wide and it never skims — but the rhythm of the cutting, the timing of a gag, the grain of the sound, everything that lives between the frames, is closed to it. That territory is yours. Beat the machine by doing the one thing it cannot: see and hear the film — with the patience it brings to reading. Analysis isn't talent; it's attention.
A Didascalies flipbook reads a film one shot at a time. For each shot you choose, you write three short analyses — the three registers — and the Studio binds them into a book indistinguishable from the published Didascalies volumes. You are not summarizing the plot. You are arguing about how a single frame is built and what it means.
Every shot becomes a two-page spread. The two pages do different jobs, and confusing them is the single most common reason a flipbook fails.
The right page is the argument — the three registers written as prose. This is where you write. The left page is the proof — under the frame, structured evidence rises when you tap a register: the crew and keywords for The Making, the compositional axis for The Frame Speaks, the lineage of descendant films for The Bloodline.
These are separate inputs. The left page is not generated from your prose. If you write a beautiful Making paragraph but leave the crew and keywords empty, the left page stays blank — a half-built book.
The three registers form a ladder of difficulty — which is really a ladder of intellectual demand.
The Bloodline — research. Which films descend from this shot? Concrete, verifiable, sourced. The easiest to fill.
The Making — documentation. Who made it, and with what. Crew, light, lens, set.
The Frame Speaks — interpretation. Name the two poles of a compositional tension and the pivot between them. This is not "describe the image" — it is argue. It is the hardest register, and the heart of the method. Give it the most time.
Open a published volume and read it like a student of the form. Notice how The Making under the frame is who-did-what, never a paragraph. That is the standard your flipbook is held to.
The vocabulary that runs through this Help and the books themselves:
A project holds one section of a flipbook — your team's shots plus the book matter around them. You can run a whole small flipbook from a single project, or split a larger one across teammates (chapter 7).
On the setup screen, fill the fields that identify your volume:
You can change any of these later from More ▾ → Project settings.
Click Open existing project on the setup screen and choose your .json file. This is how you return to work between sessions, or pick up a teammate's project. Your .json is the project — keep it safe (chapter 6).
A finished flipbook is organized into up to five numbered sections, each a thematic movement through the film. In the Studio you usually own one section; its shots are numbered within it — I·01, I·02, and so on — and renumber automatically as you reorder them. Adding and moving sections is covered in chapter 6.
A shot is the unit of work. Everything below is filled on the shot editor; the right-page prose and the left-page evidence are separate fields — see chapter 1 if that distinction isn't yet second nature.
Upload the still you're analyzing — drag it onto the frame zone, or click to choose a file. The image is stored inside your project, so it travels with your .json file.
Enter the timecode as HH:MM:SS:FF — hours, minutes, seconds, and the frame number. The timecode is a citation, not a control: it tells a reader exactly which frame you mean. Standard players stop at the second; to land on an exact frame, step through it in VLC (press E to advance frame by frame), DaVinci Resolve, or ffmpeg. Precise frame extraction is a skill worth learning — it will serve you across your studies.
The frame is your evidence, and it rewards a good source — a DVD, a Blu-ray, a restoration, or a clean digital file gives you far more to read than a low-resolution stream.
Don't worry about file size or format: the Studio takes any standard image (PNG, JPEG, WebP, AVIF) and resizes and compresses it for you, so your finished book stays small — never gigabytes. Just give it the sharpest still you have; don't pre-shrink it to a thumbnail.
Go big — bigger than it looks on the page. A reader can press and hold any frame to magnify it (the loupe), so a still that looks fine in the spread but was saved small will pixelate up close. Extract at 1920 px wide or more; don't match the size you see displayed. If the Studio warns that a frame is small, re-extract it from a sharper source.
Three ways, easiest first. Remember you're quoting the film — using a still as evidence — so work from a copy you're entitled to, and never share the film itself. A flipbook is frames and analysis, never the movie.
# one frame at 00:14:22.5 ffmpeg -ss 00:14:22.500 -i film.mkv -frames:v 1 -q:v 2 shot.png
Not sure which frame yet? Make a contact sheet — a grid of thumbnails to scan — then pull the one you want:
ffmpeg -i film.mkv -vf "fps=1/10,scale=320:-1,tile=5x6" -frames:v 1 contact.png
The title (up to 80 characters) names the shot — "The Trial Chamber", "The Odessa Steps". The caption (up to 240) is one plain sentence describing what is literally visible in the frame, before any analysis.
The three register paragraphs are the prose on the right page. Each prose field has a B / I toolbar — select text and click B for bold or I for italic, or use Cmd/Ctrl+B and Cmd/Ctrl+I. Use emphasis sparingly: a film title in italic, a single load-bearing word in bold.
Prose on the right (300–1000 characters). On the left, fill the structured evidence that rises under the frame:
The hardest register, and the one to spend the most time on (prose 300–1050). Its left-page evidence is the axis: a single compositional tension, named by its two poles and the pivot between them.
Choose an axis type — Depth, Horizontal, Vertical, Motion, or Rotational — then name:
Then the argument (up to 180 characters): one declarative sentence — the frame's thesis, in the shape "By doing X, the frame argues Y." If you can't state it in a single sentence, you haven't found the axis yet.
Prose on the right (200–950). On the left: a source — the shot named as an origin ("The trial chamber, from The Threshold · 1928") — and a list of descendants, later films that inherit this shot's idea. For each: year, title, and a short note on what it took. When you export, the Studio does the rest automatically from the title: it finds the film's Wikipedia link (so the descendant is clickable) and a poster for The Reach. It's right about nine times in ten. Two buttons handle the rest: W finds or fixes the link, and P lets you upload a custom poster — for a wrong match, or a film with no poster online. On sixty descendants, expect to touch about six.
Each prose field shows a live counter (The Making 1000 · The Frame Speaks 1050 · The Bloodline 950; aim for at least ~300 / 300 / 200). It turns amber near the limit. The budget isn't an obstacle — it's the discipline of the form. A register that won't fit is usually one that hasn't decided what it argues. Cut until only the load-bearing sentences remain.
You don't have to wait for a grade to find out whether a shot is working. The tests your instructor uses are ones you can run on yourself — and the moment to run them is now, while you're building. A thin shot is cheap to fix in week three and painful to fix in week ten.
The whole form rests on one idea: the claim sits two inches from its evidence. So test the fit. Cover the frame with your hand and read your right-page prose. Then ask: does this argument need this frame — or would any frame from the film do just as well? Uncover. If the prose floats free of the image, you're writing about the film in the abstract, which is the one thing this form exists to refuse.
Almost every weak spread fails in one of four places — one per register, plus the caption. Naming the place is most of the fix, because you can act on a diagnosis where you can't act on "make it deeper." Read each part of your shot against its thin / real pair.
Don't reach for fancier words — a thin paragraph in better prose is still thin. Instead, name the slack register and rebuild it: "my Frame Speaks asserts; make it account for the axis." Then find one of your own shots that already works, and hold the weak one to that bar. A diagnosis plus a self-made example beats any amount of polish.
Some failures are bigger than any one shot — and cheap to fix early, ruinous to fix late. Watch for these:
If your sections feel like a table of contents rather than an argument, go back to chapter 5 — your sections are your argument.
The shots are the heart of a flipbook, but a volume needs matter around them — a cover, a way in, a way out. You edit these from the sidebar, above and below your shots; the right-hand panel changes to fit each one. Most of these fields take bold / italic and blank-line paragraphs, like the register prose.
The cover carries your film's title (from project settings) over an optional frame. Upload a cover image, pick an accent colour from the swatches, and write a subtitle — one or two sentences naming what your analysis does. With no image, the cover falls back to clean typography.
Your editorial way in: four to six paragraphs on why this film, why these shots, what you're claiming. Separate paragraphs with a blank line. The transition line is a single closing sentence that hands the reader to the film.
A synopsis for a reader who may not know the film — three to five paragraphs on where it sits, what happens, who made it. Same blank-line paragraphs, same closing transition line.
Before you build a single shot, decide how your book is divided. It's the biggest choice you'll make — your sections are your argument. Split the film into "beginning, middle, end" and you've written a table of contents. Make each section a claim and you've started analysing.
There are three good ways to divide a film — look at how the published volumes do it:
Give each section a title that does some work — not "Light" but "Light as Argument." And read about your film before you decide: good research feeds your sections. Just don't copy another writer's structure — borrowed sections fit a borrowed argument, and it shows.
Your instructor sets the number — but here's the shape of a good one. Aim for enough per section to make its case: about 8 to 15 frames per section. Fewer than that and a section can't argue; many more and the book loses focus — it turns into a list, not a reading. For scale, the published volumes run 50–60 frames in total (the montage film The Man with a Movie Camera reaches 90). Fewer and deeper beats more and thinner.
Each section opens with its own intro: a short kicker (a label above the title), a title, and one or two paragraphs on the section's thematic angle. A section intro appears for every section you create, always at its head.
Introduce yourselves: a short paragraph on who you are and what drew you to this film, plus a card per member — photo, name, role, and a one-line bio. This is the page students are proudest to show; spend a minute on the photos.
The colophon is generated from your project settings (film, team, institution, course); you can add a line of notes. The closing studio page is generated for you — it credits Passeurs de sens as the maker of the tool, not as the publisher of your work. Neither needs hand-editing beyond the notes.
In the sidebar, every shot shows ▲▼ arrows on hover — click to move it one step. At the top or bottom of a section, the arrows carry the shot into the neighbouring section. You can also drag a shot: grab it and a gold card follows your cursor, with an insertion line showing where it will land. Shot numbers (I·01, II·03…) renumber themselves automatically.
Add a shot to a section with the + on its header; add a new section with + Add section at the bottom of the sidebar. Drag a section header to reorder whole sections at once.
The Studio auto-saves to this browser every couple of seconds; the header reads "Saved". But the browser is not a backup — clearing site data, switching machines, or a wiped cache loses everything. Download your project file with Save project file after every session and keep the .json somewhere real (Google Drive). The Studio nudges you if you go too long without downloading. That .json is your only true backup.
Click Preview flipbook to open your book in a new tab — the full flipbook, every spread, exactly as it will export. Leave the tab open: it refreshes on its own each time you save, so you can write on one screen and watch the page rebuild on the other.
From More ▾ → Export flipbook (HTML), the Studio builds a single self-contained HTML file — the engine, your text, and your frames, all in one document. It needs no server and no internet: open it in any browser, host it anywhere, hand it in. If you're assembling sections from a team, see chapter 7.
A Didascalies flipbook is built for a computer screen — a 19-inch display or larger, at 1920 × 1080 or finer. It's a two-page spread with small type and a page-turn animation; it needs the room. An iPad works at a pinch. A phone does not — the spread can't fit, and the reading falls apart.
So when you share or hand in your flipbook, tell your audience to open it on a computer. Say it upfront — the book itself stays clean and never interrupts the reader with a warning.
A flipbook is a team project. The Studio is built so several people can work in parallel and one person stitches the result together.
Two ways to split:
The five roles — shot selector, Making writer, Frame Speaks analyst, Bloodline researcher, and assembler — are a starting point, not a rule. Adapt them to your team's size.
When your shots are ready, choose More ▾ → Export my shots only. This writes a partial .json with your shots and the film's basic info, but none of the book matter. Hand that file to your assembler — by Drive, email, whatever you use.
One person is the assembler. They hold the master project — the one with the cover, preface, and book matter (chapter 5). For each teammate's file they choose More ▾ → Import shots from teammate and pick the .json. The incoming shots are added to the project, each landing in its own section.
Two things to watch: importing the same file twice duplicates its shots, and import only ever adds — it never replaces. So import each file once; if someone sends a corrected version, delete their old shots first. When everything is in, the assembler reorders as needed (chapter 6) and exports the final flipbook.
Only shots move between projects. The book matter — cover, preface, film intro, section intros, the team page — lives solely in the assembler's project, and only the assembler edits it. A volume has one cover and one preface; they belong in one hand, for one voice.
So if you want to write a section intro or a line for the team page, hand the text to your assembler — they place it. There's no live, two-people-at-once editing: the Studio runs entirely on your own machine (no accounts, no server, your work stays yours), so collaboration is by handoff, not by simultaneous typing.
In practice, teams settle around one editorial lead — the assembler, who also authors the book matter — while the others own their shots. It's the shape a film set already knows: many hands on the work, one vision holding it together.
Every contributor should still download their full Save project file as backup (chapter 6). The partial export is for handing off — not for keeping.
Most problems have a quick fix — try these before asking for help. Two moves alone solve the majority: reload the page, and re-open your saved .json.
Reload the page. The Studio updates itself, so a reload loads the newest version and clears almost every glitch. Your work is auto-saved to this browser, so reloading won't lose it.
Your browser data was cleared, or you're on a different computer or browser. The Studio saves to this browser, which is not a backup. Re-open your project: on the setup screen, Open existing project → choose your .json. If you never downloaded a .json, the work can't be recovered — so from now on, Save project file every session (chapter 6). The .json is the only real backup.
You filled the prose on the right but not the structured evidence on the left — they're separate fields. Under each register, add the crew and keywords, the axis, or the descendants (chapter 3). The left page never builds itself from your prose (chapter 1).
The Studio doesn't blur anything — it came in soft. Re-extract from a sharper source (a Blu-ray, a restoration, a clean file), at 1080p or finer, and don't shrink it before uploading (chapter 3).
Use a standard image file — JPEG, PNG, WebP, or AVIF. A still saved in any of those works; a PDF or a video file won't.
You're over the budget for that register — that's the discipline of the form, not a bug. Cut until only the load-bearing sentences remain (chapter 3). Amber means you're near the limit; red means over.
Standard players stop at the second. To land on an exact frame, step through it frame by frame — in VLC, press E — and read the frame number there. The timecode is a citation for your reader, not a control (chapter 3).
In the sidebar, hover a shot and use the ▲▼ arrows to move it one step; at a section's edge the arrows carry it into the next section. Or drag it. The numbers renumber themselves (chapter 6).
Before you leave, Save project file and carry the .json. On the other machine, choose Open existing project. The .json is your project — it travels with you (chapter 6).
That's the assembler's job: More ▾ → Import shots from teammate, once per file. Importing the same file twice duplicates its shots (chapter 7).
The file was probably damaged in transit — pasted into a message, or cut off mid-upload. Re-export and send the whole file (zip it if your channel mangles it). Opened on a computer, a fresh export always works.
Links and posters are found automatically from each descendant's title — right about nine times in ten. For the rest, two buttons in The Bloodline register fix it: W finds or corrects the Wikipedia link, and P lets you upload a custom poster (handy for an ambiguous title, or a film with no poster online). On sixty descendants, that's about six to touch. The entry reads fine either way (chapter 3).
You toggled the theme — the sun / moon button. Light or dark is just a preference; toggle it back any time.
Yes. Open More ▾ → Project settings and edit any field — film title, year, director, team name, section title and number, institution, course — then Save. The change flows through the whole book, cover and colophon included.