Import a Document

Have you already written a manuscript? Or do you need to merge existing subtitles with your audio description? You can import these texts when creating a document.

How to Import a Text File

You can import a text file while creating a new document. Follow the usual steps and then scroll down to the “Upload Document” section.

1. Select a “Document Type.”
2. Choose the encoding.
3. (Optional) Set the timecode offset.
4. (Optional) Enter a comment to prepend to the imported text.
5. Select the file you wish to import.
6. Click “Import Data.”
7. Click “Create.”

Note: If the import is successful, a preview of the data will be generated.

Detailed Overview of the Import Options

Below is a more detailed list of all options and steps for a successful import, as well as a few comments on different text formats.

Document Type

There are different text formats in which, for example, subtitles can be saved. Select the format you received.

Encoding

Today, texts are typically encoded in UTF-8. However, some programs may use different character sets. You can identify this issue if, for example, German umlauts like ‘ä’ appear as ‘╠ê’. In this case, try importing the document again using a different character set.

Timecode Offset

In most cases, scripts and videos are synchronized, and timecodes burned into the video correspond to the information in the script. However, sometimes the timecodes do not match. For example, “10:00:00.00” may be burned into the video, but the script starts at “00:00:00.00”. The importer will try to adjust this automatically. Unfortunately, this doesn’t always work, but you can manually adjust it later in the editor using batch processing.

Prepending Comments

You can automatically add a comment before any imported text. For example, you can prefix each imported subtitle with “UT:” to mark it.

File

Select the text file to import here.

Preview

Once you select a file, the importer will begin copying the texts and generating speech output as a test. The progress will be displayed as a growing list of small dots. Once the import is complete, the first three imported text blocks will appear.

Click on “Create” to proceed, and you will be taken directly to the editor.

Additional Information about CSV Import

While formats like SRT, VTT, or TTML are standardized, CSV requires some preparatory work. You need to use column names that Frazier understands. You can download an example which contains the English description from our “Big Buck Bunny” demo. This example shows how your CSV should be structured.

You can copy the column names from the example into your own script. The easiest way to open CSV files is with Microsoft Excel or LibreOffice. Make sure to select the correct extension when saving (i.e., .csv).

Details About the Columns

Please note that column names must always be written in English, even if your script is in another language!

Start

Timecode for the start of the text block (TCIN).

End

Timecode for the end of the text block (TCOUT).

Dialogue (optional)

Instruction or dialogue information. Displayed in italics in the editor and not read aloud.

Text

Your description. This will be read out in the editor.

Author (optional)

Your name. Only necessary if you are writing the script with colleagues (to track who made which changes).

Voice (optional)

If you already know which voice should read the text, you can enter it here. Otherwise, it’s best to delete the column in the CSV. Frazier will use the default voice you set when you created the document.

Speed (optional)

If you previously exported the CSV from Frazier, this column will show the speed used. If you’re creating the CSV yourself, it’s best to leave out the column. Frazier will generate speech output at normal speed, and you can adjust the speed later in the editor.

People (optional)

Tag the people. They will be automatically added to the glossary.

Locations (optional)

Tag the locations. They will be automatically added to the glossary.

Things (optional)

Tag the items. They will be automatically added to the glossary.

Problems with the import?

Imports can be… well… complicated. There is no reliable importer for DOCX files or unstructured text. For subtitles, we always recommend using SRT, VTT, or TTML, as these formats contain structured data, including time codes and text blocks. If something goes wrong during the import, please contact support. We can also discuss other formats. Please send us an example, and we’ll see what we can do.

Last Update: 2026/01/23