How to save OCR text as a plain text file
After Read by Zaweba reads a picture or scan, save the text as a .txt file in UTF-8, or copy it all to the clipboard, including any corrections you made.
In Read by Zaweba, once the text has been read you can save it as a .txt file straight from the window. Read writes plain text in UTF-8 with a byte order mark, so accented letters, currency signs and other languages come through intact. Any corrections you've typed into the text are included in what's saved.
A plain text file is the most portable thing there is. Every editor, every note app and every search tool opens it, and it'll still open in twenty years. When you only need the words, not the look of the page, it's the right choice.
Putting it into practice
Open the picture, scan or PDF in Read by Zaweba. Drop it on the window, paste it with Ctrl+V or open a folder. Read cleans up the page if it needs it, chooses an engine and shows the picture and the text side by side.
Read the text through before saving. You can edit it right there, and those edits are what get saved and copied. If Tesseract did the reading, words under 70 percent confidence are shaded amber in the picture and the text, and F8 jumps to the next one, so you can check the doubtful words without proofreading every line.
Then save as a .txt and choose where to put it. If you'd rather not make a file at all, copy all the text to the clipboard and paste it into an email, a document or a note.
Making the text read well
The layout mode decides how the lines come out, and that's worth a moment's thought for plain text, where there's no formatting to fall back on. Paragraphs and columns joins wrapped lines into paragraphs, rejoins words split by a hyphen, separates paragraphs with a blank line and reads columns one after the other, while keeping short lines such as addresses as they are. Single column, line by line keeps each printed line as its own line. Prose suits the first; lists, receipts and forms often suit the second.
Why UTF-8 with a BOM
UTF-8 can hold every character in every language Read reads. The byte order mark at the start is a small signal that tells programs like Notepad and spreadsheet apps that the file is UTF-8, so they don't mistake an accented letter for two odd symbols. It's an invisible detail, but it saves the "why are my letters garbled" moment.
When a different format is better
Plain text drops everything but the words. If the page has tables, Read can find rows of words lined up across wide gaps and save them as CSV or as real tables in a .docx. If you want the layout with headings, columns and page breaks, save a .docx. If you want the original page to stay as it looks but be searchable, save a searchable PDF.
You can tick more than one on the Output tab when using the Batch tab, which reads a whole folder in the background and saves each picture's text, or one combined document, in a Read by Zaweba folder beside the pictures or a folder you choose.
Quick text from the screen
If the text is on screen rather than in a file, the screen text grab hotkey copies it to the clipboard directly, and you can paste it into a text file yourself. Either way the reading happens on your PC with Read by Zaweba, with nothing uploaded.
