+++ /dev/null
-README for the ATutor 1.5.2 ATalker Module
-
-ATalker is a text-to-speech utility for reading text out loud over the Web with synthisized speech, available as an addon for ATutor. With the ATalker Theme installed, ATalker also provides an audio interface for ATutor, and reads feedback and error messages out loud.
-
-This version of ATalker (the second version) was developed on a Linux based system, but should, in theory, work on other systems that support the Festival Text-to-Speech system. It differd from the previous version on;y in that it include the ATutor module files. ATalker can currently produce English and Spanish voices. The potential for many other languages is available through the Mbrola project for those who would like to contribute. This version of ATalker is based on the Festival 1.4.3 Speech Synthesis System, though should function properly with Festival 2.0.
-
-1. Requirements
- *Server
- *Client
- *Limitations
-2. Installation and Integration with ATutor
- *ATalker Install
- *ATalker Integration
-3. Using ATalker
- *Students
- *Instructors
- *Administrators
-
----------
-
-**1 Requirements** (In addition to a working ATutor)
-*Server*:
- 1. Festival 1.4.3 or greater. Festival is a text-to-speech engine developed at the University of Edinburgh. Festival needs to be installed and functioning properly on your system before you attempt to install ATalker.
-
- Festival:
- http://www.cstr.ed.ac.uk/projects/festival/
-
- The ATalker reader will function with just the base Festival system, though it will only produce WAV audio files, which are very large, take a long time to download, and essentially make ATalker not very useful. You will want to install and MP3 or OGG encoder to convert WAV files to a more manageable size. MP3 or OGG files are about 1/10th the size of a WAV file. At least one of the following encoders should be installed on the system, and be available in the system's PATH variable (so it can be started from anywhere on the system). To determine if you have an encoder installed, type "lame", "oggenc, or "bladeenc" at the command prompt. If you get something other than "does not exists", chances are the encoder is installed. ATalker will try to determine which encoders are installed and automatically make those audio formats available.
-
- To create an audio interface for the ATalker theme, one of these encoders must be installed. WAV files are otherwise too large to be used for interface speech.
-
- 2. MP3 & OGG Audio Encoders (optional but reccommended)
- 1. LAME MP3 encoder (Unix/Linux and Windows versions)
- http://sourceforge.net/projects/lame/
-
- 2. Bladeenc MP3 encoder (Unix/Linux, Mac, and Windows versions)
- http://bladeenc.mp3.no/
-
- 3. Oggenc OGG encoder (available through vorbis-tools for Unix/Linux and Windows)
- http://www.vorbis.com/
-
- 3. Mbrola Binary and Voices (optional additional voices)
- http://tcts.fpms.ac.be/synthesis/mbrola.html
- http://www.cstr.ed.ac.uk/projects/festival/mbrola.html
-
-*Client*:
- A user's browser must be setup to read ATalker speech files. On Windows systems, default settings are often sufficeint, or when opening a speech file in a browser for the first time, users can choose which application to read the file with, and there after all such file will be read with that application. On Linux/Unix based systems it is sometimes necessary to configure browser helper applications to read these files. The mime types to add to the helper applications list are audio/wav, audio/x-ogg, and audio/x-mp3.
-
- The following are potential audio players that might be used to read ATalker audio files:
-
- Windows
- IE
- recommended: winamp (mp3,ogg,wav)
- alternates: windows media player (mp3/wav)
- Firefox
- Quicktime for Windows.
- Linux/Unix
- Mozilla
- recommended for 2.4 kernel: xmms (ogg/mp3/wav), mplayer (ogg/mp3/wav)
- recomended for 2.6 kernel: mplayerplug-in for mozilla/netscape (ogg/mp3/wav), netscape-plugger(ogg/mp3/wav)
- alternates: xmms
- Opera
-
-
-*Limitations and known issues*:
-
- The ATutor content directory must exists inside the ATutor installation. ATalker will not work with ATutor where the content/ directory exists outside the ATutor installation.
-
- Any current browser can be used to generate speech with ATalker, and when the appropriate helper application or plugin is installed, be used to read the files outloud.
-
- Linux mozilla users will require a version 2.6 kernel with a working version of ALSA (Advanced Linux Sound Architecture) installed, to take advantage of the *audio interface* features. All other features will function with 2.4 versions of the kernel (and earlier). The audio interface has been working on a Linux 2.6 system with the mplayerplug-in installed for Mozilla 1.7, but only for mp3 voice files.
-
- The audio interface works by default using Internet Explorer on Windows, and works with Netscape, Mozilla, Firefox, and Opera with the appropriate plugin installed (an audio player that plays mp3 or ogg files).
-
- Javascript is required for the popup ATalker, and for the audio interface reader, and some administrator voice manager functions.
-
- Though it is poreferable to create ogg files for the audio interface, in general mp3 files will be better supported across systems.
-
- Embed vs Object: Though we should be using the HTMLOBJECT element as the preferred method of preloading speech files, EMBED is still more broadly supported, so for now we have choosen to use the EMBED element. Support for the OBJECT element will come as it becomes more widely supported.
-
- Bandwidth: Unless you are on a T1 or higher network, you will likely notice that pages take longer to load when the ATutor theme is enabled. It will also take a little while to generate speech files. On DSL/Cable the delays are noticable, but not excessive. If you are using a modem, the ATalker theme, and using ATalker in general, is going to be slow.
-
- Server Power: The processing power of the server will also have an effect on the time it takes to generate new speech files. The speech generation process is still somewhat time consuming. Where speech files already exist, server processing power will not have much effect. Connection speed will. Creating and serving speech files works best with more cpu power, and with a faster internet connection speed.
-
- Where an error or feedback message is new, or contains dynamic information such as a file name of a file being deleted for example, ATalker will generate a new copy of the speech file, otherwise it will try to retrieve a copy of an existing speech file. When new files have to be generated, page loading speeds may not be as fast as usual.
-
- title, alt, input values: It is currently not possible to create interface voice files for language that appears within HTML attributed. In the future a parser will be developed to recognize attribute values, and handle speech files for those values without breaking the surrounding HTML.
-
-**2. Installation with ATutor 1.5.2**
-(see: ATalker .01 for ATutor 1.5 and 1.5.1)
-
- Like most other ATutor modules, the source code of the module is unzipped into the ATutor/mods/ directory, then the module installer is run. If you plan on using the ATalker theme to add speech to the ATutor interface, see the section below for additional installation details
-
- ATalker Install:
-
- *Note*: For now ATalker will only function if the ATutor content/ directory is located inside your ATutor installation, and the AT_FORCE_GET_FILE constant is set to FALSE in include/config.inc.php. For future versions we will try to handle speech files so they can be located outside the ATutor installation, or in the MySQL database.
-
- 1. First upload the atutor_talker.0.2.tar.gz into the mods directory, then unzip the file. This will create a directory called "atalker" (i.e mods/atalker/)
-
- cd ATutor/mods/
- tar xzvf atalker.x.x.tar.gz
-
- 2. Login to ATutor as the admistrator and run the module installer. Select the Modules tab, choose Intall module, then select ATalker from the list of available not yet installed modules. Read through the Details then press Yes to confirm the installation
-
- 3. Enabled the ATalker module after it has been installed.
-
- ATalker should now be available as a TTS reader for students and instructors, and as a speech generation tool for Instructors and administrators. Read on if you want to add/create a speech enabled theme.
-
-
-**3.Creating an ATalker Theme**
- Optionally you can create an ATalker enabled theme, or download and install the Default ATalker theme with each of the following features already available (see below).
-
- A) Install the ATalker theme (the ATutor default theme with speech enabled), which includes mouseover/onfocus speech, and audio feadback and error messages. See atutor.ca for the most recent version of the ATalker theme.
-
- http://atutor.ca/atalker/downloads.php
- (If you have installed the ATalker theme jump to Step 5 in section B below. If you are creating a new speech enabled theme, begin at Step 1)
-
-
- AND/OR
-
- B) Create your own ATalker enabled theme. .
-
- 1. To create your own ATalker enabled theme, copy the following Javascript into the theme/theme_name/include/header.tmpl.php file near the Jump Menu to allow users to open ATalker into a new window and have it available while navigating through ATutor.
-
- <script type="text/javascript" language="javascript">
- <!--
- document.write(" <a onclick=\"window.open('<?php echo $this->base_path; ?>mods/atalker/index.php?popup=1','newWin1','toolbar=0,location=0,directories=0,status=0,menubar=0,scrollbars=1,resizable=1,copyhistory=0,width=640,height=400')\" style=\"cursor: pointer; text-decoration:underline;\" >Open ATalker</a> |");
- //-->
- </script>
-
- Or. you can just insert a link somewhere in one of the other theme templates something like
-
- <a href="<?php echo $this->base_path; ?>mods/atalker/index.php?popup=1" target = "newWin1">Open ATalker</a>
-
-
-
-
- 2. Add the reader Javascript to the theme header template themes/theme_name/include/header.tmpl.php. Place the following script in the head area of the HTMLin the file.
-
- <script language="JavaScript" type="text/javascript">
- //<!--
- function evalSound(soundobj) {
- var thissound=document.getElementById(soundobj);
- thissound.Play();
- }
- // -->
- </script>
- 3. Add the ATalker controls themes/theme_name/include/header.tmpl.php, immediately after the Global $msgs line near the bottom of the file, but before the $msg->printAll(); line. Like the following
-
- global $msg;
- require(AT_INCLUDE_PATH."../mods/atalker/reader_controls.php");
- $msg->printAll();
-
-
- 4. Turn on ATalker by adding $atalker_on to the globals line at the top of themes/theme_name/include/header.tmpl.php, and assigning a value 1 "1" to it, as follows
-
- global $system_courses, $atalker_on;
- $atalker_on = '1';
-
- 5. Add the following lines to the _AT() function in include/lib/output.inc.php just before the last line "return $outString". This adds speech output from ATalker to the ATutor language output function, and collects a list of language items to preload.
-
-
- require($_SERVER['DOCUMENT_ROOT'].$_base_path.'mods/atalker/template_reader.php');
-
-
- You will also need to add the $embed and $_base_path variables to the globals statement for the _AT() function in include/lib/output.inc.php, like so:
-
- global $_cache_template, $lang_et, $_rel_url, $_base_path, $embed;
-
-
- You can create voice files using ATalker Administrator so ATutor interface feature will read outloud. You may want to manually create a list of language to generate speech for, then use the ATalker Administrator to create a voice for ATutor. See the administrator section in Using ATalker below.
-
-
-**4. Using ATalker**
- The ATalker addon provides speech output utilities for ATutor, for Students, Instructors, and Administrators:
-
-*Students* (and instructors and administrators):
- ATalker includes a text reader that can be used to read text out loud, copied then pasted into the reader. ATalker also includes a SABLE/SSML reader that allows users to add their own text-to-speech markup to control various speech features such as pronunciation, volumn, emphasis, pitch etc. For more about these easy to learn markup language see the following links. The SABLE/SSML reader can also be used to read text that contains HTML (so the HTML is ignored). If available in the ATutor interface, ATalker has a popup version so it can be opened alongside ATutor as you navigate through courses.
-
- **Note that markup elements in the SABLE/SSML reader must be typed in uppercase fonts. See the following references for more about SABLE and SSML text-to-speech markup languages.
-
- SABLE Markup
- http://www.bell-labs.com/project/tts/sable.html
-
- SSML Markup
- http://www.w3.org/TR/speech-synthesis/
-
- If the *ATalker Theme* is enabled, a mouse pointer, or the Tab key, can be used to read interface elements outloud when they recieve focus. Feedback and error messages will also be read outloud.
-
-*Instructors*:
- Instructors and course members with ATalker privileges have the same features available to them as those described above for students, with the add ability to save files to the course content/ directory. A directory called "speech/" will be created to contain course speech files, and can be accessed through the course file manager. The Save tool can be used to generate audio files that can be linked into ATutor content pages, providing instructions, summaries, or perhaps audio copies of content pages etc. Using the SABLE/SSML reader TTS output can be formatted with markup to fine tune the audio output (see the references above). You could potentailly create TTS speech clips of two people talking, in different language, at different speeds, and different volumns, for example. The SABLE/SSML reader should be used to read text that contains HTML, otherwise the text reader will read the HTML tags along with the text.
-
- Instructors should use the MP3 or OGG audio formats for generating speech files. You may also generate files in WAV format, though they are very large, will take up a lot of space, and take a long time to download when students are listening to the files. MP3 is the more common format of the other two, supported by most current audio and multimedia application. OGG files are becoming more common, and are the preferred format. OGG is an open audio format that should be used where ever possible.
-
-*Administrators*:
- If the ATalker theme has been installed, the system's administrator will want to generate voice files for ATutor so users can mouseover the interface to have it announced outloud. If the ATalker theme is not installed, then there isn't much to manage for an administrator, except the installation.
-
- Course size limits:
- You may find that the *file size limits* placed on courses becomes a problem when instructors are generating lots of audio content, particularly if they save their content in WAV files. You might want to discourage using WAV files, and encourage using MP3 or OGG formats which are about 1/10th the size of WAV files. Set the following default values in include/config.inc.php to something larger.
-
- $MaxFileSize = 1048576; /* 1 MB */
- $MaxCourseSize = 10485760; /* 10 MB */
-
- Also in course properties you can increase space available on a course by course basis.
-
- Generatong an Interface Voice:
- Using the Manage ATutor Voice tool administrators can choose which interface items to make available with speech. If you choose to make the entire interface speech enabled, it can take a very long time to generate all the voice files. You may want to select a subset of interface items to add speech too, or generate speech files a few at a time.
-
- Use the text reader or the SABLE/SSML reader to adjust the voice to your liking. Then select the items from the list below the reader to generate their speech files. Voice items that contain HTML must be generated with the SABLE/SSML reader, otherwise the HTML will be read along with the text. You can also create other speech files that can be manually linked into the ATutor interface through theme templates to add various speech enhanced instructions, help,etc.
-
- Once your speech files are generated, they can be managed through the Voice Files tab in the ATalker Administrator
-
projects / atutor.git / blobdiff