Uploading documents
Which files work, how big they can be, and what happens to a scanned document.
On this page9 sections · 6 min read
Which files can I upload?
You can upload a PDF, Word (DOCX), PowerPoint (PPTX), Markdown (MD) or plain text (TXT) file. A file can be up to 50 MB, and a PDF up to 300 pages.
We go by what is inside a file, not by its name. If the name and the contents disagree, we tell you. For an old .doc Word file, save it as .docx in Word and upload that.
In PowerPoint, the slide text, speaker notes, tables and charts are read, and the page in a source line is the slide number. Hidden slides are not read. Text that sits in a picture is not read (PowerPoint has no text recognition), so add the text to the slide as real text. For an old .ppt file, save it as .pptx in PowerPoint and upload that.
How do I make a video?
You upload your file, and Zıpır reads the document and prepares a scene plan. Every scene in the plan comes with its quote and its source. Nothing is made until you approve it; once you do, the video is prepared and we tell you when it is done.
Can I decide how a term in the video is pronounced?
In the panel, open the pronunciation dictionary from the account menu, write the term (for example NASA, YouTube or a brand name) and spell out how it should be read. Only the narration changes; the caption stays the exact quote from your document. The dictionary spends no minutes. A new pronunciation applies to the next video you make; videos that are already finished do not change. Up to 200 terms are kept per account. At the approval step of the scene plan, words whose reading is unclear come with a suggestion that you can add with one click.
The dictionary has no audio preview. Next to each term a line shows how it will be read; that is the text handed to the narration engine, worked out as you type. You hear the actual voice in the finished video. There is no spoken preview at the moment.
Can I upload a scanned PDF (one that is like a photo)?
We also try to read the writing on scanned pages (text recognition). It takes a while, and each document has a time allowance. If a long scanned document runs out of time, no video is made and we tell you.
The best result comes from a PDF whose text can be selected (one with a text layer), or from uploading a part of the scanned document, for example the first 20 pages, as a separate file.
Can I choose the narrator voice, the pace and the language of the video?
After you upload a document, the scene plan screen has a folded voice and narration section above the approval card (the phone approval has the same section). If you choose nothing, the default narration is used. There are 10 voices: Defne, Selin, Kerem, Barış, Ada, Iris, Tess, Cole, Finn, Jude. A voice with a listen button next to it plays a short ready-made sample.
Pace: Slow, Normal, Fast. Tone: Calm, Balanced, Warm, Lively, Enthusiastic. The sentences stay exactly as in your document; only the way the voice speaks changes. A slow narration can make the video a little longer and a fast one a little shorter; the minutes are the same at every pace and are counted from the length at normal pace. Choosing a voice, pace, tone or language costs nothing extra.
The language of the video can be Turkish or English. If the document is in another language, the on-screen text and the narration are translated into the language you pick, and under each scene in the plan you see the original sentence from the document. If you change the language on the approval screen, the scene plan is rewritten in the new language.
The voice and language cannot be changed after approval; for a different one, make a new video.
A document that is not in Turkish · Pronunciation dictionary
Can I upload a document that is not in Turkish?
Yes. You choose the language of the video when you upload, in the voice and narration section: Turkish or English. If the document is in a different language from the one you pick, the on-screen text and the narration are translated into it, and under each scene in the plan you see the original sentence from the document.
You get the best result from a Turkish document made into a Turkish video. For a document in another language we do not promise the quality of the result; try a short part first. Your minutes are not used until you approve the scene plan.
Which format should I choose?
The format is chosen by what the document is about. Documents that set out steps, rules and obligations (a safety instruction, an article of a law) suit Clay figures; documents that explain why or how something works (a natural phenomenon, a recipe) suit Whiteboard or Illustrated explanation; dates and sequences of events suit Timeline; texts where the page itself matters, such as a contract or a form, suit Document walkthrough; number-heavy and table-heavy documents suit Summary presentation; documents about the parts of a machine suit the 3D cutaway; documents about places and routes suit the Map.
If you choose no format, Zıpır looks at the kind of document and picks a suitable one. If you chose one, your choice is not blocked. If the document is a poor fit for that format, a one-sentence note appears in the chat when the scene plan is ready (it ends with the Turkish sentence “Yine de devam edebilirsin.”, "You can carry on anyway"), with a link that starts a new video in a more suitable format. You can ignore the note and approve the scene plan.
Can I use more than one document in a video?
Each video is made from one source document. If you want to combine several documents in one video, merge them into one file and upload that, or upload each one as its own video.
How do I split a very long document?
Save a part of the document, for example a chapter or a topic heading, as a separate file and upload it. Each part becomes its own video and goes through its own approval. To see in advance how long a document will run, use the calculator.
Other topics: Length and quota, Payment and invoices, Errors and fixes, Privacy, Account, Sharing. If you did not find the answer, write to the contact address in the footer.
Last updated: