Screen Readers

For people who are blind or visually impaired, working on a computer is extremely difficult. However, technical de­vel­op­ments in the past few years have meant that blind people can now use computers in their daily work without en­coun­ter­ing major problems. So-called screen readers provide the written text as an (text to speech) or as braille. The purpose of this software is to enable blind users to work in­de­pend­ently on the computer. This tech­no­logy not only turns PCs into devices that can be operated by blind people: screen reader apps also enable blind people to use smart­phones.

When choosing a screen reader, it is also important for it to be com­pat­ible with the devices used in the work place. For example, if someone uses a braille display, this must be com­pat­ible with the screen reader. Of course, this also applies to the computer’s operating system and other as­so­ci­ated software. The following questions are also important to ask during the selection process: Does the user only need reading or writing aids at the workplace? What content do they have to work on? What is the budget for setting up the work­sta­tion?

For some tasks you will need expensive software, while free tools are suf­fi­cient for others. This often depends on the user’s range of tasks: Do they only have to read documents or also research them on the internet? How well can the user handle the tools that may be available at the workplace? How does co­oper­a­tion with other col­leagues work? In any case, a com­pre­hens­ive workplace analysis is crucial. The functions and ways in which the screen reader can be used are also important for private users.

Before pur­chas­ing a screen reader, you should make sure to look into whether the state might grant benefits. Check also if your health insurance covers aids such as screen readers. However, since the number of free tools has increased, many health insurance companies simply recommend these to their patients.

Screen readers read displayed text elements and translate the in­form­a­tion via synthetic speech output or send signals to a connected braille display. This display then outputs the cor­res­pond­ing text as tactile braille, for trans­la­tion. Synthetic speech output uses syn­thes­isers such as Elo or eSpeak to produce the sounds. These ap­plic­a­tions use in­teg­rated dic­tion­ar­ies, which also contain the pro­nun­ci­ation of most of the words. However, it’s not easy to create a natural sounding speech this way. After all, words written in the same way are often pro­nounced dif­fer­ently depending on their meaning or syntactic position (e.g. read, wind, tear, dove, etc.). If a syn­thes­iser doesn’t dif­fer­en­ti­ate between upper and lower case and word meaning, the word 'Polish' (a person from Poland) could end up being pro­nounced like 'polish' (making something smooth or shiny by cleaning it). Other problems arise due to the different in­ton­a­tion of in­ter­rog­at­ive sentences. The way the voice elevates and sinks often in­flu­ences the meaning. 'OK' and 'OK?' have different meanings, which only become clear when spoken, since they are em­phas­ised dif­fer­ently. For the speech syn­thes­izer to reproduce the second word correctly – as a question, the voice must be elevated at the end of the sentence, otherwise, the user will mis­takenly interpret the question as a statement. In speech synthesis, the following quality features are of par­tic­u­lar im­port­ance:

• Word stress: The syn­thes­iser should not only pronounce each in­di­vidu­al word correctly, but should also be able to emphasize certain words.
• Syllable trans­itions: If the syn­thes­iser composes the language from syllabic blocks, the trans­itions must be fluid to produce un­der­stand­able words.
• In­ton­a­tion: If elevating and dropping the voice is relevant to the meaning of the sentence (marked in the text), the syn­thes­iser must be able to reproduce it.
• Speech rhythm: The syn­thes­iser should try to imitate the natural rhythm of the speech so that the user gets a natural hearing im­pres­sion.
• Speech tempo: The speed of reading aloud is also important for the user. Ideally, it can be set by the user.
• Breaks: Format elements such as para­graphs and line breaks should be marked with pauses to make it easier to register the end and beginning of a passage.

Computer lin­guist­ic research has made con­sid­er­able progress in recent years. Google’s Tacotron 2 system is already very close to human speech quality. The driving force behind the latest de­vel­op­ments is the in­de­pend­ent learning ability of modern syn­thes­isers. Roughly speaking, the program learns the language like an infant and then 'builds' the language from real voice documents. It is par­tic­u­larly as­ton­ish­ing that Tacotron 2 is re­l­at­ively resistant to ty­po­graph­ic­al errors and deals well with punc­tu­ation and stress in sentences (e.g. with caps lock). However, there is still a lack of emotion in speech synthesis. Foreign words can also cause dif­fi­culties for Tacotron 2. It remains to be seen when this syn­thes­iser and similarly strong competing products will be made available to a wider public.

The ap­plic­a­tion areas for screen readers aren’t limited to helping blind people. Screen readers are primarily intended as a tool for people who cannot see very well or not at all due to a dis­ab­il­ity, or who are maybe il­lit­er­ate, but they are not limited to only these people. In principle, the programs can also be useful for people that don’t have reading dif­fi­culties. For example, if you have a job where you can’t read at the same time as doing another task, it would be helpful to have the text read out to you, and therefore a screen reader would be the best solution. Here are some usage scenarios:

• When driving: Screen readers can read out texts for you while you con­cen­trate on the journey. Be it technical articles, news texts, or guide books.
  
  
• Listen to texts on your smart­phone while on the go: If you feed a text into a screen reader app be­fore­hand, you can also have any text read aloud to you from your smart­phone e.g. during a walk or on public transport. This can be more con­veni­ent than reading a text on a small screen.
  
  
• Radio re­place­ment or guidance during work: If you are pursuing an activity that doesn’t require your full con­cen­tra­tion, then a screen reader can take on the role of a radio or podcast to entertain or inform you. Even if you work with in­struc­tions, a screen reader can be a great support by reading the in­di­vidu­al steps out to you.
  
  
• Give your eyes a break: Hours of reading text on a computer screen can strain your eyes. Sometimes it’s a nice change to sit back and have a text read out to you. Es­pe­cially after a long time of working with written texts, switching to an acoustic medium can increase your con­cen­tra­tion again.

In everyday life, it is important that buildings are designed to be easily ac­cess­ible for wheel­chair users e.g. with ramps and lifts. The same principle applies to the internet. To enable visually impaired users to navigate the internet with screen readers, certain rules need to be obeyed: The Equality Act 2010 explains what website operators need to consider to make their website ac­cess­ible to as many people as possible.

However, since the internet has changed con­sid­er­ably in recent years, it could also mean that this equality act has changed too. The WACG re­com­mend­a­tions can be used as a benchmark.

However, the term 'ac­cess­ible internet' doesn’t only refer to the han­di­capped ac­cess­ible structure of websites, but also to a generally easily ac­cess­ible structure of contents and texts. Ac­cess­ib­il­ity across multiple platforms and operating systems also char­ac­ter­ises good ac­cess­ib­il­ity on the internet. Ideally, websites are also designed in such a way that they can be un­der­stood by people of different social and age groups as well as by non-native speakers. However, depending on the topic of the website, this is sometimes difficult to achieve.

Ac­cess­ib­il­ity on the internet is un­for­tu­nately not a given. Modern, dynamic websites are often just as difficult to read from screen readers as those that in­creas­ingly work with Flash, JavaS­cript, and graphics. Many content man­age­ment systems (CMS) have also been cri­ti­cised for gen­er­at­ing websites that are not very ac­cess­ible. Fre­quently en­countered hurdles include in­suf­fi­cient colour contrasts, lack of scalab­il­ity, and missing alt texts for images. Without these al­tern­at­ives, blind users cannot fully access the websites.

If websites aren’t designed for ac­cess­ib­il­ity, screen readers can’t work properly. They will come out with in­co­her­ent drivel that even those ex­per­i­enced with screen readers will find hard to un­der­stand.

• Graphics: If the web designer does not specify an al­tern­at­ive text (alt text) for graphics, then screen readers have to skip the content of these graphics. However, if a graphic is important for the meaning of the ac­com­pa­ny­ing text or if the text refers directly to the graphic, then screen reader users won’t be able to un­der­stand. Al­tern­at­ive text for images can easily be added using HTML code. Screen readers then read out these texts so that visually impaired people know what would be visible on the cor­res­pond­ing images. If the graphic also contains an ad­di­tion­al link, it is important to indicate this in the al­tern­at­ive text.
  
  
• Tables: If tables are not arranged sys­tem­at­ic­ally and clearly, it will be hard to read them out. Many screen readers make it easier for the user to un­der­stand tables, but the screen reader itself will have no luck if the table has no clear rows or column names, for example.
  
  
• Text design/text code: If certain text formats such as bold format or italics are necessary for the text to be properly un­der­stood, not every screen reader will be able to ad­equately reflect this. Missing headers and sub headers in the code can also make nav­ig­a­tion difficult with screen readers. Generally speaking, if you use pro­gram­ming code such as HTML and CSS properly and create an easily un­der­stand­able text structure, you will make it easier for visually impaired users to read the content with screen readers and navigate a website.
  
  
• Hidden content: For some websites, certain content may only be displayed if the user carries out a certain action. For example, in­ter­act­ive buttons could magnify a text that was pre­vi­ously hidden. Screen readers may not be able to recognise texts like these. Or the text is read out in the wrong order because the screen reader is trying to read it out even though the user hasn’t clicked on anything. In general, screen readers have a lot on their plate when it comes to com­plic­ated text and website struc­tures.
  
  
• Frames: If a website does not assign unique titles to certain frames, only the content of these frames is often read aloud (e.g. nav­ig­a­tion bars). This makes nav­ig­a­tion con­sid­er­ably more difficult for users with limited vis­ib­il­ity. Con­sist­ent texts should be in a single frame. Ideally, a website should not be divided into frames and luckily this is usually the case nowadays.
  
  
• Forms: Input fields must be clearly labeled so that screen readers can display their functions correctly. In addition, they should also be easy to navigate using the keyboard.

NonVisual Desktop Access (in short 'NVDA') is char­ac­ter­ised by its very simple ac­tiv­a­tion and operation, and is therefore par­tic­u­larly re­com­men­ded for beginners. NVDA supports both synthetic speech output that reads texts and trans­la­tion using a braille display. The program is open source and financed by donations, which enable its de­vel­op­ment to continue. NVDA is com­pat­ible with all versions of Windows from Windows 7 onwards (Service Pack 1), but the man­u­fac­turer re­com­mends using it with Windows 10. The features of NVDA at a glance:

• Supports all current browsers, e-mail programs, chat programs, and Microsoft Office programs such as Word and Excel. NVDA also helps with Windows input prompts and console ap­plic­a­tions.
• Consists of a single .exe file (approx. 50MB), which can also be loaded from external data carriers such as USB sticks. The in­stall­a­tion tool is operated in­tu­it­ively.
• Many simple keyboard shortcuts help navigate content such as web pages, turn NVDA on and off, and pause the reading process.
• The speech syn­thes­izer enables the output of texts in 44 different languages. The default syn­thes­iser is the freely ac­cess­ible eSpeak syn­thes­izer. NVDA also enables SAPI 4 and SAPI 5 to be in­teg­rated.
• The screen reader can recognise and indicate special text format­ting.
• Supports mouse functions such as hovering the mouse over certain text.
• Supports many current braille displays.

Job Access With Speech ('JAWS') is a fee-based screen reader from Freedom Sci­entif­ic. The market leader in screen reader tech­no­logy (the first version of which was developed for MS DOS back in 1989) is still the most popular elec­tron­ic aid for blind and visually impaired users with the current version: JAWS 2018. JAWS is available for all current Windows versions and is con­sidered to be extremely stable and powerful. You can use a trail version that will function for 40 minutes, but for anything longer than this and you will need to purchase a license. You can obtain these from various dealers. Depending on the license, it could cost between £850 and £2,600, although the price model is anything but trans­par­ent. Before making a purchase, you should see if your health insurance will cover the costs or at least some of it.

The features of JAWS at a glance:

• Powerful speech syn­thes­iser called Eloquence, capable of reading 7 languages (English, Spanish, German, French, Italian, Finnish, Por­tuguese). Supports the com­par­at­ively natural sounding syn­thes­iser of RealSpeak.
• Manuals in reading and braille make it easy to get started with JAWS. In addition, a large help database and customer support is available online.
• The screen reader can also capture PDFs and text that is embedded in graphics and an­im­a­tions, in addition to the usual text documents.
• Quick nav­ig­a­tion buttons and list functions enable fast and con­veni­ent nav­ig­a­tion on the internet.
• Supports many internal Windows tools and services.
• Supports most braille displays.
• Voice con­trolled in­stall­a­tion enables in­de­pend­ent setup by the visually impaired user.
• Support of numerous programs e.g. Microsoft (Office), Adobe, and the common web browsers. With expert knowledge, you can also align JAWS with specific software.

The Dolphin screen reader was formerly marketed under the name SuperNova before being named after the developer Dolphin Computer Access Ltd. This screen reader is designed to give blind and visually impaired users complete control of their computer. The tool can be tested free of charge for 30 days, but after that you have to purchase a license for the program. The price for the Dolphin license is between £850 and £2,600. It is important to research whether there’s a possible subsidy so that you do not have to bear the costs yourself. Dolphin is char­ac­ter­ized by the following features:

• Powerful, com­par­at­ively natural sounding speech syn­thes­iser 'Vocal­izer Ex­press­ive'. It offers settings for voice, melody, and speed. According to the man­u­fac­turer, it is possible to create your very own voice output thanks to the many options.
• Quick nav­ig­a­tion through keyboard shortcuts, fast switching between elements.
• Search function to navigate even faster with websites.
• The screen reader can check documents with regard to format­ting and then provides 'text style' in­struc­tions so that the documents look good even for sighted readers. Dolphin helps you to define font, font size, bold and italics, and un­der­lin­ing. The screen reader can also read out PDF documents.
• 'Dolphin Cursor' simulates mouse over keyboard commands. This can help users who are unable to navigate with a mouse.
• Supports more than 60 braille displays.
• The Setup Wizard file makes it easier for blind and visually impaired users to install Dolphin on their own.
• Ad­di­tion­al eBook support.

This tool is very simple: It makes it easier to navigate through some keyboard commands and reads texts without any problems. However, the available speech syn­thes­isers are outdated and no longer meet current quality standards. Com­pat­ible syn­thes­isers are not available in all languages. Blind and visually impaired users should therefore buy special software such as NVDA, JAWS, or Dolphin, which are useful in the long run. Windows’ internal tool is only re­com­men­ded for beginners who are trying this simple tool for the first time, users with residual vision, as well as users who do not depend on screen readers all the time.

VoiceOver enables blind and visually impaired people to operate Apple systems using voice control and the keyboard. The tool is pre-installed on macOS, making VoiceOver a very obvious choice for Apple users. Activate VoiceOver in macOS using the key com­bin­a­tion 'command F5'. The software is also available on mobile iPad and iPhone systems (from 3GS upwards). VoiceOver currently supports 26 languages.

Since macOS X 10.5, VoiceOver’s speech syn­thes­iser has been called 'Alex' and is con­sidered as powerful and easy to un­der­stand because, unlike other syn­thes­isers, it manages to imitate human speech flow in a very con­vin­cing way. 'Alex' can also integrate breathing sounds, which make this syn­thes­iser even more authentic. As of macOS X 10.7, VoiceOver also allows the in­teg­ra­tion of RealSpeak voices from the Nuance company. Since the choice of the speech syn­thes­iser is often a matter of taste, the wider range of products and services provides added value for the user.

VoiceOver also supports the most common braille displays and offers control options via both the keyboard and mouse. To simplify nav­ig­a­tion, VoiceOver divides website content into elements that can be read aloud via the program. In­ter­act­ive buttons such as text fields and scroll­bars can be con­trolled with VoiceOver using the keyboard. It also offers suitable functions for those that use track pads and touch pads. How VoiceOver is used, is explained in the video a few para­graphs below.    

Orca is a free, open source screen reader for Linux that is con­stantly evolving thanks to the large community and active developer circle. Orca uses eSpeak as the default voice syn­thes­iser, which generates a rather ar­ti­fi­cial-sounding language. For this reason, Orca is looking for computer linguists who would like to par­ti­cip­ate in the syn­thes­iser’s further de­vel­op­ment. Orca also relies on in­ter­na­tion­al trans­lat­ors to make the program usable in many languages. It’s clear that Orca is a joint project.

The practical audio in­struc­tions, which can be down­loaded free of charge, are what makes Orca extremely popular. The user is in­tro­duced step by step to the program and its functions. In­teg­ra­tion into the Linux operating system also plays a central role.

Orca is part of the GNOME platform (since version 2.16) and comes standard with Ubuntu and Fedora. In addition, the systems usually contain syn­thes­isers, braille display drivers, and mag­ni­fic­a­tion tools that Orca auto­mat­ic­ally in­teg­rates. In some cases, however, the user has to download these com­pon­ents in­di­vidu­ally, which may com­plic­ate Orca’s in­stall­a­tion somewhat.

Some browser ex­ten­sions read text aloud to you. These small programs have been developed es­pe­cially for reading out website content. ChromeVox for Google Chrome and Chirpy for Firefox are two free, easy to install read-aloud programs for the browser.

Both ap­plic­a­tions work via small buttons in the browser window, which can be used to read aloud text that has been selected be­fore­hand. While Chirpy’s speech syn­thes­iser sounds pleasant and quiet, many people often find the speech output of ChromeVox annoying and too loud. Qual­it­at­ively, both syn­thes­isers cannot compete with the voice programs of 'real' screen readers.

However, browser ex­ten­sions are not an equi­val­ent re­place­ment for special screen reader software. Es­pe­cially since visually impaired users depend on good usability. This means that the browser add-ons cannot be used. Rather, they are of interest to those users who do not have severe visual impair­ments and only want to be read aloud to from time to time without much effort.

How to activate TalkBack:

• Use the app menu to access your phone’s settings and scroll all the way down
• Under 'Set­ting­s' you will find the category 'Ac­cess­ib­il­ity'
• TalkBack appears under 'Ser­vices', if it’s installed; If it isn’t installed, it can be down­loaded for free from the Google Play Store
• In the TalkBack window, activate or de­ac­tiv­ate the app using the slider
• Find the 'Read aloud' function under 'Input help', which is useful if texts are to be read out only on your command (If TalkBack is activated, you can ignore the option 'Read aloud')

TalkBack is con­stantly active, which can cause some devices to no­tice­able slow down. The screen reader is therefore es­pe­cially re­com­men­ded when combined with a powerful smart­phone. This is because loading delays can be prob­lem­at­ic for blind users, since they might presume they’ve made a mistake them­selves and that’s why the screen reader isn’t working properly, rather than it being a technical fault. TalkBack is a very im­press­ive program, but requires adequate per­form­ance from the equipment to ensure that it runs smoothly.

If you have problems with TalkBack or just want to try an al­tern­at­ive, the screen reader from Shine (also known as ShinePlus) is a superior choice. The tool is available free of charge from the Play Store and financed by ad­vert­ising. As soon as you have installed it, it will appear in the 'Ac­cess­ib­il­ity' section and can be activated there.

As long as the app is activated, a single tap on a certain area is all that’s needed to instruct the app to describe it to you. Buttons are briefly explained and text elements are read aloud. To able to use and navigate mobile phone functions, each icon must be touched twice i.e. double-clicked on. ShinePlus also provides a mag­ni­fy­ing glass for users with limited vision. However, the feature for re­cog­nising icons could be improved – es­pe­cially with more complex apps. ShinePlus doesn’t recognise in­ter­act­ive surfaces, which throws a spanner in the works if a blind user should come across these. TalkBack enables a better handling of web content in com­par­is­on. However, according to user reports, it is more fluid than Google’s app, which is es­pe­cially be­ne­fi­cial for weaker devices. In addition, the signal tones in ShinePlus are con­sidered less intrusive.