3.2. Files
Load files into QualCoder prior to coding.
You can import text from plain text documents, docx, epub, md, odt and pdf documents. Pdf importing is problematic and may need editing as formatting is lost. Large Pdf files take a long time to import. QualCoder extracts text from the Pdf files using pdfminer.six, but you may choose to use another extraction program for the plain text, suggestions include: www.pdf2go.com/ www.pdfmate.com/pdf-converter-free.html https://pandoc.org/ and then copy the plain text across.
The following image files formats can be loaded: jpg, jpeg and png.
You can create and enter text into a file stored in the database, by pressing the 'Create' pencil icon.
If you import really large text files, when coding QualCoder can split the files into 50,000 character chunks to reduce the slowness of the program. The best option when importing large text files is to split them into separate sections and load those smaller sections as separate files instead. Give each filename some ordered logic too, e.g. book_chap01.txt, book_chap02.txt.
If you have enabled AI integration in QualCoder, newly imported text files will be read by the local AI model and stored in its local memory. This happens in the background (status bar shows "AI: reading"). You can continue to use QualCoder, but the AI enhanced features will not be available until the AI has finished updating its memory. This also happens if you edit an already imported text file in QualCoder.
If you ever get the impression that the AI memory is out of sync and has missed some updates to your data (which should not happen under normal circumstances), you can rebuild the memory from scratch by selecting "AI > Rebuild Internal Memory". This will reread all documents, which may take a while.
There are a few sample files in the Examples folder. These can be used to test importing files of different document formats and of importing an image. There are also example files to test importing case attributes, and for importing a survey.
Video (mov, mp4, wmv format) and audio files (wav, mp3 format) can be imported. Loading an audio or video file will also automatically create a blank text transcribed file. This file will have the same name as the audio or video file, but have a '.txt' suffix. Initially, this text transcription will be empty. You can transcribe the file yourself or get the file professionally transcribed and copy and paste the text into this file.
Filename entries can be changed in the database. The original file names of the actual files will not be changed.
A right-click context menu allows you to view, export, delete a current file.
The trashcan button is an option to select multiple files for deletion. The question mark '?' opens this help page.
Files can be linked to rather than imported into the QualCoder project folder. The diagonal link icon performs this task. The two other icons with links allow you to import a linked files into the project folder, and to remove a file from the project folder but keep a link to the file.
Attributes are variables that can be used to describe or classify the files. These can be added using the (x) button or through the Manage Attributes menu option.
Each file is preceded with an icon that describes the type of file. Other columns show the file name, date of creation/upload, association with cases and other attributes. The screenshot above shows a 'source of data' attribute and a reference entry.
Files can be viewed either by right-click menu on the file, or by clicking the magnifying glass icon. This is where transcripts for audio and video can be entered, when viewing an audio or video file.
As a practical example: Open the Manage Files dialog. In the Examples folder import the following files: ID1.docx, ID2.odt, transcript.txt and the miguel-henriques.jpg.
The right-click context menu also allow you to re-order the files according to alphabet, date, and file-type and when right clicking in the name or date columns. You can show only selected attribute types: Choose Show this value option You can also select Show values like. Show values like can be applied multiple times to further shorten the list of displayed rows.
Sometimes a lot of columns can be displayed which can be overwhelming. You can hide columns by right-click on the table header row for options to hide that column or to hide columns beginning with specific text. Or to show columns where the column name starts with specific text.
Right-click on a URL in an attribute and open the URL with the default web browser. This is only if the attribute is a URL and it passes URL validation.
Use the right-click menu option for reference columns (Ref_Authors, Ref_Title, Ref_Journal, Ref_Type, Ref_Year). This allows copying reference to the clipboard as a Vancouver or APA style reference.
You can assign a file to a case via right-click menu when in the cases column.
Manage Files > Create text file or View text file. Manage files View Audio/Video (with the transcribed text file shown).
Text files can be edited providing no coding or annotations or case assignment have been performed with the text file. Copying and pasting text from elsewhere (e.g. web page) may show formatting from the copy/paste (e.g. bold, italic, foreground and background colours) until the text file is re-opened. BEST PRACTICE: Save a copy of the project before editing the text of a coded text file.
Sections of the text file or audio/video transcribed file will have sections of text underlined in green (case assigned), yellow (annotation) or red (coded). Text can be edited even after the text is coded or annotated or assigned (fully or partially) to a case.
You can select text that is not underlined and copy/replace without problems. You can click on a position (without selecting a section of text) to then type, delete, or paste text. This can occur in underlined (coded, annotated, case-assigned) or not underlined (not coded/annotated/case-assigned) text locations. You will see the underlines shift as text is added or removed.
There are some limitations: It is best to avoid selecting sections of text to delete (or to type or paste over) if those sections have a combination of not underlined (not coded/annotated/case-assigned) and underlined (coded, annotated, case-assigned). The reason is that positions of the underlying codes/annotations/case-assigned may not correctly match as intended. If you have made a change that you think has affected these coded/annotated/case-assigned positions badly, exit the text editing window by pressing the Cancel button.
You can open an audio or video file to view. for video, this opens two windows, one for viewing the video and one for the controls and a transcript, shown below. When an audio or video file is loaded into QualCoder a blank text transcription file is automatically created. The transcription file name defaults to videoname.txt.
The transcribed text file is created and stored within the database, and can be exported to a text file. The file can be edited. Important note: The edits occur within the database. An original text file is NOT changed. Transcriptions cannot be linked as an external file.
If you have a .srt file (a translation file that is read by VLC) you can place this alongside the video inside the project folder, in the video folder, shown in the image below. When the video is played, the translation wording will be shown as subtitles in the video. Also, if you open the .srt file in a text editor, copy and paste this into the video.mp4.transcribed text file. Then this text will be shown as the transcription for the video.
If you have ffmpeg installed a waveform is shown. If the audio files has multiple audio tracks, only the first track is shown for the waveform.
To transcribe open the Manage files menu option, then view the audio or video file. The audio/video will load and there will be a text area to enter transcribed text. QualCoder does not have an automated audio to text feature. Other services such as otter.ai may assist you. Transcriptions should ideally contain timestamps indicating when the text is being pronounced during the video. The following formats are recognised by QualCoder, where SSS are milliseconds:
[hh:mm:ss]
[mm:ss]
[hh.mm.ss]
[mm.ss]
For the above, brackets can be [] or {}. These can be changed in the Settings menu.
#hh:mm:ss.SSS#
hh:mm:ss,SSS --> hh:mm:ss,SSS
The video file might contain multiple audio tracks. There is a drop down box that allows you to choose another audio track.
Transcriptions may contain speaker names indicating who is speaking. Speaker names are bracketed in this format: [name] or {name}. Dots ‘.’ and colons ‘:’ cannot be used in speaker names.
Manually transcribing audio and video is helped with some keyboard shortcuts. Transcribing and adding or editing text can only occur if the existing text has no codes or annotations. Copying and pasting text from elsewhere (e.g. web page) may show formatting from the copy/paste (e.g. bold, italic, foreground and background colours) until the text file is re-opened.
Transcribing and adding or editing text can only occur if the existing text has no codes or annotations.
The shortcuts available are:
Ctrl R Rewind 5 seconds
Alt R Alt Minus (release 3.3+) Rewind 30 seconds
Alt F Alt Plus (release 3.3+) Forward 30 seconds
Ctrl S or Ctrl P Stop/Start toggle audio/video. From stop to play will rewind 2 seconds.
Ctrl Shift > Increase play rate up to 2 times
Ctrl Shift < Decrease play rate down to 0.1
The above controls are also available from the toolbar icons.
Ctrl T Insert timestamp in this format: [hh.mm.ss]
Ctrl + N Add a speaker name. This also pauses the audio/video.
Ctrl D Delete one or more speaker names.
Ctrl 1 to 8 Insert speaker name in this format: [name]
Approaches to transcribing speech to text are:
Microsoft Word Transcribe
Zoom transcription
Otter AI: https://otter.ai/
VOSK. https://alphacephei.com/vosk/install
audiopolis https://github.com/audapolis/audapolis
noScribe is a fully offline transcriber that can transcribe 99 different languages. See their gitHub to download, install and use: https://github.com/kaixxx/noScribe
Sections of the text file or audio/video transcribed file will have sections of text underlined in green (case assigned), yellow (annotation) or red (coded). Text can be edited even after the text is coded or annotated or assigned (fully or partially) to a case.
You can select text that is not underlined and copy/replace without problems. You can click on a position (without selecting text) to then type, delete, or paste text. This can occur in underlined (coded, annotated, case-assigned) or not underlined (not coded/annotated/case-assigned) text locations. You will see the underlines shift as text is added or removed.
There are some limitations: It is best to avoid selecting sections of text to delete (or to type or paste over) if those sections have a combination of not underlined (not coded/annotated/case-assigned) and underlined (coded, annotated, case-assigned). The reason is that positions of the underlying codes/annotations/case-assigned may not correctly match as intended. If you have made a change that you think has affected these coded/annotated/case-assigned positions badly, exit the text editing window by pressing the Cancel button.
If your project contains externally linked files, these files can be moved, renamed, or deleted. Linked files are shown with a red link icon.
The Mange bad links feature allows you to edit the existing link and replace it with a new one, by finding the correct file and it’s location. Bad links do affect images, audio and video files. Text is not impacted unless you need to review an original text document. This is because the plain text, on which coding occurs, is imported into the QualCoder database.
If you are importing an REFI-QDA project with external links, you will likely need to update the links to files using this function.
There might be a couple of seconds delay. There is an automated search that looks through the user's home directory for up to 2 matching file names. This intended to speed up finding the link to files where the link is currently pointing to nothing.
Note: The twitter function has not been reviewed recently, so it may not actually work.
Import twitter data from a fully quoted csv file. There is an example file - rtweet_judo_tweets_data.csv - in Examples folder.
The csv file requires these exact column names: id and full_text column names for tweet data and screen_name for the column for user data.
Additional tweet fields can be these exact column names: created_at, coordinates, retweet_count, favorite_count, lang Additional user fields can be: location, url, description, followers_count, friends_count, listed_count, favourites_count, statuses_count
The twitter import page also has an example of how to use R with RTweet to extract tweet data. This was tested and worked before twitter applied a fee for service. Note that currently there is a cost associated with accessing twitter data. I cannot give advice on any problems you may have getting twitter data.
The tweet data is loaded into individual database files (Manage Files), The user data is loaded into cases (Manage cases). Multiple tweets (stored as files are assigned to the matching user (case).
