My name is Fredrik Larsson and I am the CTO of ReadSpeaker. My first contact with TTS (short for Text To Speech) was in the mid 80s. I was 14 years old, doing some period of practical occupational experience in school, together with a blind researcher in history. He used it together with software that could read the screen aloud to him. It was a quite robotic voice but I could hear what it was saying and I got some idea that I maybe could use computers like my friends.

My next experience was a kind of Nim game played with a computer as opponent, and the computer talked to me. It was the same kind of synthetic voice but it was almost like playing a game with a human being. This meant that I was even able to play those text adventure games and some more simple ones not requiring graphics. It was now time to do some serious investigations and try to get hold of a computer with Text-To-Speech for use in school.

I got my first IBM computer, about 2 dm high, and 5 dm wide and deep. It was not an ABC80 or C64, but my friends thought this was even cooler – and it could talk. Back in that time, a TTS system was a 4 dm long ISA card that one plugged into a computer and connected it to a standard car speaker. It was controlled by a separate box with thick knobs for volume and on/off. Some years later, external Text-To-Speech systems came to market. They were boxes, about the size of an iPad, but 2 inches thick, and you could carry them between different computers and connect them, without the need for a screwdriver.

I used TTS quite heavily in the early 90s, mostly at home, since my university studies were more convenient to do using braille. However, I used quite a lot of talking books, mostly in science, so I realized that having something that could talk is not enough. It also matters how the information is read to make it understandable. Reading everything from left to right, top to bottom, is often not the best way to convey something aloud to someone.

I then stayed away from TTS until 2001, when my work with ReadSpeaker technology started. TTS had then made a dramatic change. The voices sounded almost like humans, but unfortunately they also made some of the mistakes as people tend to do, such as speaking a bit more sloppy. The old classic voices appeared very clear, you could even hear spelling errors in the original text. This was not as simple to do with the more natural sounding voices, but they attracted new target groups, both people that listened to text while reading it themselves simultaneously, and people that wanted to consume information by listening to it, instead of reading it on paper or screen.

TTS in itself is not enough to produce a nice user experience when listening to complete tables, formulas, and image descriptions. It must also be complemented with good automatic processing of documents and web pages to produce a high-quality narration. Listening to all the audio books and teachers describing what they wrote on the blackboard during all these years in school and university has definitely made it clear to me that quality is more than voice and audio. One of our greatest challenges for the future is to make new Text-To-Speech systems sounding as clear and exact as the classic ones, still having as nice sounding voices as real people, but without the sometimes too sloppy reading.

When developing web based products that are supposed to be a piece of the accessibility puzzle, such as ReadSpeaker, the product itself must obviously not introduce new difficulties for the users. They need to be made to work everywhere and for everyone, always. How is that achievable? Our approach to this is that we put as much of the intelligence as possible on the server side, and only put what is absolutely needed on the client side (in our case the “client side” is on the website where ReadSpeaker is implemented – not the end user side). And any JavaScript components that are used to increase the user experience for visitors using browsers that support it, we host in the cloud. This way we can make sure that the user experience is the same, or in the worst case similar on all platforms, and we can easily update the cloud scripts as the web evolves. Since most web browsers have support for different techniques and have different abilities to view embedded media (such as audio streams) we can adapt what is sent from the server-side fully depending on what kind of device and browser the end user uses. And by using obtrusive scripts with graceful transformation, we make sure that even the simplest technology used on the client side can also get the speech back. Basically, it needs to be able to provide the core functionality even without requiring JavaScript support or possibilities to embed rich media. You never know what the visitor might use to browse your site or use the online product. Even though the majority uses the Top 5 browsers I would say stopping there just isn’t enough. It does not make sense to lock people out for no or the wrong reasons. A reason should never be “If you can’t have it all you shouldn’t have anything”.

By developing the core functionality of a web based product using the simplest possible pieces and following the specifications, you can ensure maximum possible visitors/users. Additional UI features and user experience bells and whistles should be added on top of it, like the ketchup on a hot dog. The hot dog is totally eatable also without the add-ons, even though it wouldn’t be as nice, it makes a very small difference if you are really hungry!

Apart from developing a platform and browser independent solution, there are also a number of others important aspects of creating an accessible web product. The user interface should be keyboard accessible, as intuitive as possible and in general be very easy to use. In the ReadSpeaker case, we make sure that the speech is always only ‘One Click Away’.

Web products and web applications that provide flexibility for the customer on how they are implemented can sometimes have some drawbacks. With flexibility also comes responsibility. It can become difficult for a company developing these kind of product to always guarantee the functionality if implementers with low skills in web accessibility get too creative. The implementation instructions we provide to our customers guarantee that they will be as end user friendly in general and as accessible as possible in particular.

Apart from the technical side of things, nothing beats end user tests. We try to gather a relevant number of people from the intended target groups and observe them while using the service and following the tests ask them number of questions. Then, we often go back to the drawing table and incorporate the missing pieces.

One of the difficulties is always to combine the requirements from very different user groups to get to a “design for all” solution at the end, but once there, If you make sure that it is functional for the users with the greatest difficulties; it is then usually the case that it is very usable for everybody.

ReadSpeaker has traditionally been strong in speech-enabling websites. We are the world market leader in that domain. Over time, we have become involved in speech-enabling (news) apps. More recently, we have seen that clients have requested our services in a variety of ways, where the text-to-speech technology is baked into their own product.

Besides the adoption of text to speech in online teaching curricula, ReadSpeaker has also been embedded into online therapy. In this way, patients are able to receive therapy at home behind their computer screen and therapeutic texts are read to the patient.

Another way of embedding ReadSpeaker into other products where speech can play a major role is to use text-to-speech technology to build audio tours for cities, museums, and other tourist attractions. Recent discussions with clients across Europe, the US, and Australia have given us very positive feedback with regard to the feasibility of this concept.

The concept is rather simple: embed the text-to-speech engine into an editing platform for mobile internet applications. For the content owner/museum, this will deliver an easy-to-use, flexible platform to create audio tours for its visitors on their mobile devices (Android, iOS, etc.).

The museums and cultural institutions that received a presentation were very interested: it provides them with a flexible tool that they can maintain and author themselves. They can publish tours, change and update them on the fly, and provide the visitors of the museum with additional services without having to invest in hardware devices with difficult-to-maintain and inflexible prerecordings.

Some of the institutions that I visited in both NYC and Australia complained about the expense of the current tour apps that they had in place. The costs were astronomical, the flexibility was nonexistent, and the process to create them had left some frustrating scars in the organization because of the difficult cooperation with external agencies.

The nice thing about this new service is that the power is completely in the hands of the cultural institution. You build your own tour, you choose which elements to include, which texts to be read, which pictures to be shown, what sequence to follow, etc. And very important, the user interface to both build the tours (as the content owner) and to play the tours (as a visitor) is completely intuitive and is built by following the “users as designers” principle.

With the arrival of Siri, the voice recognition service of the iPhone, it has become clear that voice services as credible man-machine interactions have become a reality. Yet it is clear that the built-in voice technology of today still is not perfect. Online services can therefore be a major step forward. Especially with the fast development of online connectivity of networks, the cloud will become more and more a place for these solutions.

ReadSpeaker has been cloud-based from the start. The development of networks and the need for voice services are going hand in hand with our development path. It is clear that online voice services will play a major role in the future of man-machine interactions.

How do you find a speech button on the website? That is a question that people often ask us. Of course, we try to promote a consistent way of providing the button on a website and a way of implementation that ensures the highest chance that people will find the button. So a big button! Preferably one that made up of both a text (Listen!) and a graphical part ().  And preferably as close to the content as possible. For example, next to the header of the article. Or just below the article header.

Remember: a speech button on a website is not a function that replaces reading! It supports the reading, and listening to the text while reading at the same time greatly enhances the understanding of many people who have difficulty reading texts.

In our broad experience of supporting customers that have implemented our speech functionality on their websites, we have seen many interesting variations on the same theme. So let’s have a look at one such example. UMC Groningen, the largest hospital in the northern part of the Netherlands, has added our speech functionality to their website. Here you see the speech button implementation:

What is really nice is that they have decided to have an audio introduction message the first time you visit their home page. It will tell the visitor something like “welcome…you find a speech button on this website if you want to listen to the content of the webpage”. To have such an introduction message is a real help and at the same time is hardly intrusive for the web visitor! And the button is well designed, consisting of both a text and a graphical part.

Still, we feel there is also room for improvement because the placement of the button is quite far from the text; a spot next to the print button or just below the article header would be better. Even if that might be so, if we look at usage statistics then we clearly see that the audio introduction of the speech function is a huge success! Usage of the speech function is many times more than on other comparable websites!

Looking at examples like this gives you a good feel for what options and choices there are and to determine for your own specific situation which design is best to help your visitors find their way to the speech function on your website. And an audio introduction clearly seems the way to go!

Be sure to contact us when you’re ready to add speech to your website!

In our Q&A series, we are very pleased to have had some time to discuss with Bert Zahniser, Vice-President of Technology of the American Institute of History Education, and Laureen Hungo-Brady, Instructional Design and Curriculum Editor of CICERO Kids™ .

The American Institute for History Education’s (AIHE) mission is to provide substantive, engaging historical content and activities for teachers to use in their classrooms that will dramatically increase students’ comprehension of historical events, personalities, issues, and trends. The American Institute for History Education was founded to provide history teachers with high quality professional development programs, firmly anchored in rich historical content, along with substantive historical lessons, activities, and resources to use in their classrooms.

AIHE recently launched CICERO Kids™ which is a convenient and cost-effective, anytime classroom tool that is comprised of five online units (sold separately or in modules). Each unit is a complete and comprehensive program designed modularly so that it may be added to other units to expand the series. An interactive museum is the setting for students in Pre-K through 5th grade to learn about American history and social studies from the first explorers to the early 20th Century. CICERO Kids™ lets students explore a great lobby and meet a cast of interactive characters who are waiting to share their “Did you Know” facts.

Since CICERO Kids™ lessons are available on a subscription basis, they created a special demo page where you can see and listen to ReadSpeaker in action.

Roy Lindemann (ReadSpeaker): Why did you add ReadSpeaker to your CICERO Kids website?

Laureen Hungo-Brady (AIHE): I’ll discuss the educational reasons first and then hand over to Bert. There are 2 reasons for which we believe ReadSpeaker is a useful tool for the children who use our site. We have a wide range of students from pre-kindergarten to 5th grade with different levels of reading comprehension going from non-readers to more advanced readers. The children use our website with the help of teachers and at other times by themselves. With the help of ReadSpeaker’s listen button, all children can now listen actively to the different text contents we have in our interactive musuem. ReadSpeaker provides the right breadth so that all our students can use it easily. The second reason is that ReadSpeaker helps struggling readers validate what they are reading. The dynamic conversion of text into high quality speech enables these readers to hear if they are reading the text correctly from a phonetic point of view.  The synchronized highlighting of the words as they are being read enables a better comprehension of the written content. This helps empower non readers, beginner and struggling readers as well as students with English as a second language.

Bert Zahniser (AIHE): From a technical point of view, the 2 fundamental reasons that made us choose ReadSpeaker are the synchronized highlighting of the words being read and the quality of the text to speech voices. We also like the various voices that are available.

Laureen Hungo-Brady: Can I add that another reason of importance in adding ReadSpeaker, is the current focus on literacy, in conjunction with the National Common Core Standards. The synchronized highlighting as the text is being converted to speech improves literacy by helping the eye track the words that are being spoken thereby associating sound and meaning.

Roy Lindemann: What has the feedback been so far?

Laureen Hungo-Brady: It’s been absolutely great. Teachers really see the value in the highlighting and tracking of the words as they are being speech-enabled.

Roy Lindemann: How easy has it been to add ReadSpeaker to the CICERO Kids website?

Bert Zahniser: It went very smoothly. We implemented ReadSpeaker on our development, staging and production sites and it worked nicely from the start. We got good support from your technical team to deal with the sign-in aspect of our site. Our pages were speech-enabled in a couple of hours and it took only about 2 weeks to fully test and fine-tune aspects of the player such as the look and the positioning until we were ready to release to the public.

Roy Lindemann: How did ReadSpeaker cope with the pronunciation of your written content?

Bert Zahniser: We were pleasantly surprised how accurately ReadSpeaker interpreted the pronunciation of our content including names. As a history provider, we have lots of names of people and places that can be hard to pronounce.  I expected quite a few pronunciation issues but thus far have had only a couple that we needed to have corrected.

Roy Lindemann: To what type of content owners would you recommend our service?

Laureen Hungo-Brady and Bert Zahniser: Any website that targets struggling readers, beginning readers or language learners would greatly benefit from adding a service like yours.

Recently I visited some of our clients  in Australia. We had several meetings with existing long term clients and some engaging sessions with new prospects. It is clear that accessibility is high on the agenda of the Australian public. There is a major effort to implement WCAG2.0 and the knowledge level of the Australian webmasters about the topic is impressive. A few things were striking during the long meetings that we had; online text to speech is an excellent tool to help the accessibility of a website, and most of our existing and new clients understand that and see that as one of the reasons to implement ReadSpeaker on their website. The usage statistics for the ReadSpeaker services are excellent and I was pleased to see such a large group of keen ambassadors of our solutions.

During the meetings, we discovered another striking new way of using ReadSpeaker for accessibility purposes. A way that we had not envisaged before. In several of our meetings with the Australian Federal government departments, it became clear that many of the PDF and Word documents that are displayed on the government’s website are not accessible. They have been produced in times when accessibility was not an important attribute. Lay-out and design took the forefront. This has created a major task for government institutions, as all information needs to be accessible by year end 2012. Some of the more innovative departments have been using our ReadSpeaker docReader tool on their websites to read PDF documents. The perfect reading of these documents depends on their level of accessibility;  the better they are tagged, the better they are read.

This fact has led some departments to use ReadSpeaker in a slightly different way. It has become a gate keeper to test a document on its accessibility before it is published on a website. When someone wants to publish a document (from inside the department or from an external agency), ReadSpeaker docReader is now being implemented as a checker on the staging/testing server. Any PDF, ODF, Word document can be checked on its correct accessibility tagging by running it through the docReader tool. If it reads correctly, it is correctly tagged. This proves to be a great help for the creators of the content. In many cases these are external agencies that create these documents for the Federal government. Our clients have found an interesting way to use the ReadSpeaker products to help them test the accessibility of their own content. That is probably one of the best proofs of the alignment of ReadSpeaker to the online accessibility goals of WCAG2.0.

While I was in Australia, Julia Gillard, Australia’s prime minister, launched the National Year of Reading. Australia is facing 4.5M people with literacy issues. That is a stunning 20% of the 22M population in Australia. These percentage numbers are similar in other developed countries across the globe. The ReadSpeaker technology offers a great help for these struggling readers, not only by assisting them to read and understand the content of the online texts, but also by helping them to learn to read better. Our technology of synchronized text highlighting while listening offers the capability to learn to read better over time.

Those are the words of the Australian Prime Minister, Julia Gillard, in a speech given to launch the National Year of Reading campaign on February 14. About 4.5 million working-age adults Australians do not have the necessary higher reading and numeracy skills to succeed in work or study, Ms Gillard said. According to the The National Year of Reading 2012 project, nearly half the Australian population struggles without the literacy skills to meet the most basic demands of everyday life and work. There are 46% of Australians who can’t read newspapers; follow a recipe; make sense of timetables, or understand the instructions on a medicine bottle. This is just the latest statistical example of how illiteracy is affecting countries in all parts of the globe.

According to the UNESCO, 793 million adults suffer from illiteracy in the world. Over half of the adult population of Benin, Burkina Faso, Chad, Ethiopia, Gambia, Guinea, Haiti, Mali, Niger, Senegal, and Sierra Leone are illiterate. 21% of women in the world are illiterate.

In France, 9% of the adult population suffers from functional illiteracy. Over half of the people who are functionally illiterate in France have a job.

In the United States, a study by the U.S. Department of Education, National Center for Education Statistics, showed that 22% of adults were below basic in quantitative literacy in 2003 (indicating they possess no more than the most simple and concrete literacy skills). 63 million adults - 29% of the United States adult population – over age 16 don’t read well enough to understand a newspaper story written at the eighth grade level. An additional 30 million - 14 % of the country’s adult population – can only read at a fifth grade level or lower.

From February 13 to February 15, 2012, we participated at the Tools of Change for Publishing conference of O’Reilly in NYC. O’Reilly’s TOC Conference is where the publishing and tech industries converge, as practitioners and executives from both camps share what they’ve learned from their successes and failures, explore ideas, and join together to navigate publishing’s ongoing transformation. This year’s conference was focused around the rapid technology changes that are facing the education business. The fast acceptance of the tablet as a delivery vehicle of learning content has changed the industry profoundly. Textbooks are morphing into interactive apps, which create a complete different learning experience.  These apps provide a perfect platform for experimental and interactive learning and teaching. I was very impressed with the new interactivity tool provided by Wolfram Alpha for interactive math learning, as was demonstrated in a keynote by Theodor Gray.

The rise of electronic textbooks is impressive, as was showcased by Safari Books Online, whose innovative business model is based on usage and not just on the number of users. Their rapid growth shows that the model is successful and accepted, probably not too dissimilar from the ReadSpeaker model.

The conference’s major theme this year was the rise of big data in ePublishing, from the importance of metadata to the interpretation of usage and interaction data. It has become evident that tagging the content is almost as important as producing the content in the first place, as the task of making it findable has become very important. One of the side effects of creating eContent is the disappearance of pages, which creates a difficulty in sequencing content. Finding specific content becomes even more important (and difficult).

Clearly accessibility continues to play a major role in the conversations about technology development and roadmaps for all major publishers at the conference, especially since the electronic era offers a new dimension to provide that access. Most publishers are engaged in deep conversations of how to do this in the best possible way, especially around their mobile delivery platforms. The ReadSpeaker solution offers several benefits : it guarantees a high level of speech quality/voice regardless of what device, operating system and browser the user happens to use, it delivers exact usage statistics to our clients, it works on-the-fly thereby allowing publishers to change their content at any time, it provides quality control over pronunciations of words that are of extra importance to our customers, the SDK – our audioMobile API – is super light for mobile applications and the packaged solutions for websites or mobile websites offers the same user experience on desktop PC as for mobile/portable devices. This explains the rapid growth of ReadSpeaker technology among the major publishers.

Overall, the conference provides a perfect platform for publishing and technology providers to discuss the further integration of technology into the curriculum delivery and into the teaching methods. All major publishers and providers meet and mingle at the conference and the exhibit floors on Times Square. I am sure however that there is always time to enjoy part of the buzz in the Big Apple when you are there.

Different organizations choose to provide a talking version of their website for different reasons. Some have a true wish to reach out to an audience that may have difficulty reading; they just want to help. Others might want to enable the visitor “on the go” to consume the content when moving from A to B. In the public sector, where the information is really for large target groups that are extremely diverse and where there are requirements to be accessible to them all, a service such as ReadSpeaker makes perfect sense. For online newspapers and other online media sites, the more people who can come to the site and use the content, the more visitors and the more revenue from advertisers.

Some organizations are most interested in the actual return on investment (ROI) and profitability of adding such a service. Organizations that I believe would profit the most (financially) are the eCommerce players. I mean the more people who can shop in your store, the more customers you’ll get and the more money you’ll make. The math is simple. If you had a physical store where the customers needed to climb a steep ladder to get in, you would effectively exclude a lot of potential customers. That could be one reason you simply don’t have store entrances like that. Widen the doors, get a wheelchair ramp for strollers, wheelchairs, and walkers, and everyone can get in and buy stuff! Makes sense in the physical world. But the same goes for eCommerce. You shouldn’t build your e-shop just for people who use a certain web browser and a certain kind of computer. If people want to shop, they should be able to do it from any kind of device anywhere and anytime.

But you can do more! A large percentage of consumers (your potential customers) has some sort of reading difficulty. In that case, it doesn’t matter how device- or browser-independent your website is. Once in, your customers need to be able to read about the products, read comments from other customers, compare different products, or simply read whatever text content is there. A consumer needs to feel secure and confident to click the “add to basket” button. Speech enabling can make all the difference to a lot of people, enabling them to learn about your products, to feel confident, and to buy.

Just as you choose whether to take the wheelchair ramp or the stairs, the choice is yours; you do what is most convenient given the situation.

Contact us to learn more about how you can make a difference as an eCommerce player.

We’re asked all the time if we prerecord audio and play it back when the user clicks the Listen button. Some people even ask if we then keep these recordings available on our server. The answers are no and no. All speech is generated on the fly from your latest content.

The most explicit example of this concept can be seen on media and news websites and mobile apps where content may be added or updated several times a day, hour, or minute.

Why should you add online text to speech to media and news sites or mobile apps?

• It increases the accessibility of your news content for your current readership and attracts new users who appreciate the choice to read or listen to your online content.
• Your content becomes mobile. Giving your mobile users the ability to listen to your news while on the go is a very useful way to deliver content to them.
• You can provide your advertisers with a new audio- and/or banner-based advertising space focused on users who enjoy audio content.
• It enables automatic redistribution of your written content in audio format.
• Multitaskers appreciate being able to do other things while listening to the news.

What about live coverage? Why should you add online text to speech to live coverage, such as news, sporting events, stocks, tweets, etc.?
The internet is a real-time medium, with more and more content added or updated faster and faster. ReadSpeaker converts live coverage of all types of news into speech on the fly enabling your users to listen to the most up-to-date news and information.

Have a media/news site or mobile app? Contact us to make it talk!