Main Menu

Powered by <TEI:TOK>
Maarten Janssen, 2014-

TEITOK Help Pages

Waveform view

The waveform view presents a split view on a TEITOK file, with on the top a waveform display of the sound file, and on the bottom the transcription. The waveform is generated by the wavesurfer Javascript library. Both the transcription and the waveform will stay centered around the current time-index while the sound is playing. On the bottom of the page is a small explanation video on how to use the waveform view for transcribing files.

The waveform shows both a detailed cut-out of a couple of second around the current time-index, and a smaller display with the entire waveform. You can scroll the waveform left/right and the transcription up/down, and you can click anywhere in the waveform to jump to the corresponding time-index, or on any utterance to play the corresponding time-slot.

Below the waveform are some control buttons that not only allow you to start/stop, but also zoom in or out, and speed up or slow down the playback rate. Instead of the buttons, you can also use the following keys, where there are some options only available via the keyboard, most of which have to do with the option to use a highlighting timeslot (henceforth HL). HL is a selected timeslot on the timeline that can be used to create new utterances in the transcription. It can be set using the a and e keys. Clicking on an existing utterance with set HL to the corresponding timeslot. And HL can be modified by simply dragging the yellow box, moving or resizing it. Apart from using the shortcuts, you can also create a HL by dragging clicking and shift-clicking on the timeline.

aSet the start of HL to the current time index
bWind back to the start of HL
cCreate a new utterance for HL (in edit mode only)
eSet end of HL
eSet the start HL to the end of the previous HL (if still selected)
pPlay HL
nSet playback rate to 100%
sSlow down the playback rate
qRemove HL
zZoom in
[left]Skip back 2 seconds
[right]Skip forward 2 seconds

Edit mode

TEITOK is not meant as a full-fledged editor for audio transcription, but it does come with a basic function to add transcriptions to a sound file. As a staff member, you can enter the edit mode by clicking the edit link on the bottom of the screen. XML files without a transcription are opened in edit mode automatically. In edit mode, the shortcuts would interfere with typing text, and therefore, you have to hold down the ALT key to make the shortcuts work. This does disable the normal function of functions like ALT-P. The edit mode in wavesurfer uses the DOM editor of the browser, which affects your XML - it will for instance turn self-closing nodes into empty nodes. If you do not want this, do not use this edit function.

In edit mode, utterances that are not (yet) time-aligned are made opaque to make it clear which utterances lack a timeslot. You can create a timeslot for those utterances by creating an HL, and then simply alt-click on the utterance to assign the timeslot to the utterance. You can see it has been linked by the fact that the utterance is no longer opaque.

In edit mode, you can click alt-c to create a new utterance for the HL (after creating one on the timeline). Before actually creating the utterance, an edit window pop up, that allows you to type in the transcription for the selected timeslot. The transcription is the inner XML of the <u>, in which you can use TEI/XML to mark pauses, truncations, gaps, etc. The most useful XML codes for transcriptions are shown in the next section. New utterance are always added to the end of the transcription, you cannot insert an atterance into the timeline. To get the utterances in the right order, you can click the "sort utterances" button on the bottom, which will order the utterances by their start time.

You can also correct existing utterances by clicking on them. First click on the utterance, which will set that utterance as HL. When an utterance is selected, the highlight color will be red rather than the default yellow so that you know you are working with the region corresponding to an existing utterance. Then you can edit the time alignment of the utterance by simply resizing the red box. Or you can hit alt-c to open the annotation window to edit the source code behind the utterance.

You can even modify the transcription itself by just changing the text, or even delete an entire utterance. Although an easy way to change things, you edit a textual version of underlying XML, which means you do not really know what you doing. To help a bit in determining what you are editing, the xpath of the node where the cursor is is shown below the transcription field. And the position on the timeline will follow the cursor, automatically selecting the corresponding HL. Nevertheless, it is recommendable to use the preview edit only in simple cases, and use the "raw XML" button on the bottom for serious editing.

On the timeline, there are three different colours used to indicate active regions: yellow is a newly generated region using the a and e shortcuts; while green is not a HL - bt used to highlight the region corresponding to the "active" timeslot, which show the regions the sound goes through while playing, or the regions corresponding to the utterance the mouse is moving over when the sound is not playing. And finally, red is a HL, but one that corresponds to an existing timeslot; you get a red region by clicking on an utterance and dragging the edges of a red region will change the underlying timeslot, while clicking ALT-c on a red HL will allow you to edit the XML of the utterance.

Recommended XML codes

XML files in TEITOK follow the TEI/XML guidelines. The section on spoken data in the TEI guidelines is very sparse compared to the rest of the framework, and the description not always clear. Below is a short of of the codes established as best practice amongst the spoken project in TEITOK. The codes with a slash behind it such as <pause/> are self-closing tags without any content inside, while the other tags say something about whatever is between the opening and closing tag, so <unclear>Betty</unclear> represent a segment where the speaker probably says "Betty". Because of how the browser treats XML, self-closing elements are converted into empty elements: <pause/> becomes <pause></pause> in wavesurfer.

<pause/>A pause - to differentiate between short and long pauses you can use type="long" and type="short"
<gap/>A gap in the transcription - you can put the motivation behind the gap in the reason, for instance reason="unintelligible"
<unclear>A segment that has been transcribed, but for which the transcription is uncertain
<del>A segment that has been "deleted", which is, corrected by the speaker. There are three typical types:
<del type="truncated"> - a truncated word / false start
<del type="repetition"> - a repeated segment where the speaker "stutters"
<del type="reformulation"> - where the speaker reformulates what he started
<vocal>An extralinguistic element - the description goes inside: <vocal><desc>Uhm</desc></vocal> stands for an "uhm" sound.
<kinesic>A kinesic element - the description goes inside: <kinesic><desc>Clap</desc></kinesic> stands for a clapping sound.
<anon>An anonymized segment - with an option type, say type="person". Can be used later to blank out the corresponding segments in the audio file semi-automatically

How-to video

Back to index