Analyzing Orphaned Transcripts in TransanaNicolas Sheon, PhD, UCSF Center for AIDS Prevention StudiesAn orphaned transcript is one that has lost its source media file, such as the recording of the interview from which it was transcribed. Transana software (www.transana.org) is primarily designed for transcription and analysis of multimedia data. The main idea of Transana is to synchronize the transcript to the audio or video recording using time codes. Coded segments of a transcript are bounded by these time codes.While working with media files in this way immerses one in the source data, this is not always possible or desireable. Sometimes the source media are no longer available. If you are analyzing written texts, a spoken version never existed. Even when the source media are available, inserting time codes is time consuming, especially if you are new to Transana's keyboard short cuts. For users who have transcripts but no source media or patience to insert time codes, the following steps provide an alternative method that inserts time codes between each word, thereby creating an artificial time line that is synchronized to the text. The section entitled, "Creating a Foster Episode File" describes how to create media file based on the transcript using a text-to-speech engine. A computer generated audio file can never fully replace the original parent media file. However, by creating a foster home for your orphaned transcripts, the method outlined in the following steps allows you to take advantage of Transana's unique coding, visualization and collaboration features that rely on the time line. The goal of the following steps is to add a time code between each word. Time codes in Transana transcripts consist of a time code symbol, ¤, and a number of miliseconds between two brackets, <4000>. To make the time code symbol: Hold down the Alt key and type 0164 on your number pad, and then release the Alt key. You may need to hit Num Lock to activate your number pad on your keyboard, or Alt+Fn if you are using a laptop keyboard that lacks the number pad. For example, if we wanted to insert a time code every second between the words: "one mississippi, two mississippi, three mississippi" it would look like this: one mississippi¤<1000>, two mississippi¤<2000>, three mississippi¤<3000>Normally, each spoken word takes only a fraction of a second. The whole point of adding a long word like "mississippi" is to slow the rate of speech down to one second. If we were to find the average length in miliseconds for each word in a segment of talk, we would simply divide the length of the segment by the number of words. In normal conversation, each word would only take an average of around 400 miliseconds. If an interview transcript originally looked like this: After adding a time code every 440 miliseconds, it would look like this. If we put this time coded text into Transana, the numbers would disappear and we would see: However, there is a way to make even the red time code symbols (¤) disappear. Just click on the time code button at the top of the transcript window and the time codes will disappear from view, but still function as an index for the time line.
Adding time codes between each word will increase the size of your transcript considerably, resulting in slower loading speeds in Transana. Also, work with a duplicate, not your only copy of the transcript. Use this Macro to Insert Time Codes in your Transcript with Microsoft WordDavid Woods has written a macro that inserts time codes between each word. A macro is simply a series of instructions to automate a repetitive task. To use this macro, copy the entire text from this text file and paste it into your Visual Basic Editor that can be opened in Microsoft Word. Go to Tools->Macros->Visual Basic Editor or type Alt-F11. Right-click on the "Normal" project in the tree on the left side of the screen and choose Insert > Module. Then paste the text from the Macros.txt file into the new module, save, and close the Visual Basic Editor. (What you need to do may vary, depending on your platform and version of Word.) The screen shot below shows the Visual Basic Editor with the macro script pasted into it.
If you would like to change the time code interval, within the Visual Basic Editor, simply change the time code value from 450 to the desired length. In Word, open the document into which you would like to insert time codes between each word. Hit Alt-F8 to launch the macro. Select "AutoTimeCode" and click the Run button.
The macro will then insert the time codes automatically. To test it out, save the document as an RTF file and assign it to an episode in Transana. See "Creating a Foster Episode Audio File" below for guidance on creating a silent episode file or a text to speech audio file that will serve as the episode for this auto-time coded transcript. As you can see in the above screen shot, there are actually several macro scripts in the editor, one to remove time codes, one to insert time codes, and several others to hide and/or reformat time codes. See David Woods' at description of these macros. Should you edit the time coded transcript in Transana, the time codes will no longer be in synch with the episode. In that case, from Transana, go to File->Save Transcript As and save it on your computer. Open it in Microsoft Word. Run the "StripTimeCodes" macro to remove the old time codes and then insert new time codes, create a new episode file and load them back into Transana. Creating a Foster Episode Audio FileEven if you do not want to hear the audio, you will still need an audio file to provide a timeline for the creation of clips. The audio file needs to be a certain length which is the number of words multiplied by the time code interval, which in the above example is 440 miliseconds. For example, if you have an interview with 10,000 words, the audio file would need to be 4,400,000 ms which is equivalent to 4,400 seconds or 73 minutes, 20 seconds). The free audio editing software, Audacity (available for download at http://audacity.sourceforge.net) can be used to create an audio file of a certain length. In the Menus choose File->New. Then select Generate->Silence then enter the number of seconds you need. Screen shot of the Audacity silence generator.
To save the file to your computer, Choose File->Export as wav or mp3. This file can then be assigned to Transana as an episode. Creating a Zombie Episode Audio FileI you want to hear the words of the interview spoken and synchronized (more or less) with your text, then you can use TextAloud to generate an audio file based on the text of your interview. TextAloud is a text-to-speech engine that will produce a computer generated voice reading the text of your interview. This is nothing like the original recording, hence the name "Zombie Episode Audio File," but if no audio recording exists or you don’t want to take the time to synch the original with the transcript, this is a quick and dirty way to make an audio file to help navigate your transcript.
You may want to remove the speaker ID’s from the file first so that the reader does not say "I:" or "P:" before each utterance. TextAloud can asign different segments of text to different voices to produce the effect of two or more voices in a dialogue. Simply replace the I: and P: with text commands that tell Text Aloud which voice to use. Also, certain acronyms and transcription short hand may give the text to speech engine some trouble. For example, it helps to replace “HIV” with “aech ivy” and "{laughs}" with “heh heh heh.” Also, you can insert pauses as needed by changing the punctuation in the text that will be read in TextAloud. For example, when I read the phrase, "Would the following people please stand up: Person A, Person B, etc." I would pause at the colon. To add the pause in TextAloud, simply change the colons to periods. The software has a built-in search and replace feature. TextAloud can be purchased here: http://www.nextup.com/TextAloud/index.html. To see and hear an orphan transcript being read by "Charles" and "Graham" in Transana, see http://www.palmpal.org/PIdemo/PIdemo.html. Compare the text in the Transana Transcript to the text in the TextAloud window in the above screenshot. Questions or comments about this document email nicolas.sheon@ucsf.edu. For questions about Transana, visit www.transana.org. |
