How to Train Your iOS Device: Dragon Dictation

This article reviews Dragon Dictation, a speech recognition and text conversion app for the iPhone, iPad or iPod Touch. With Dragon Dictation, users can supplement the standard on-screen keyboard found on iOS devices with natural language, spoken word input.

The Smartphone war between Research In Motion’s Blackberry devices and Apple’s iPhone center around the keyboard. Long before the iPhone was released, RIM paired their large screen all-in-one mobile phone and email devices with sturdy, easy to use physical keyboards. The iPhone was designed with media and simplicity in mind so Apple opted for a virtual, on-screen keyboard. Both physical keyboards and virtual keyboards have drawbacks and as any Star Trek fan will tell you, voice input is the way of the future. Created by text-to-speech software creator Nuance, Dragon Dictation hopes to become the bridge between flawed mobile keyboards by evolving the way that people input text on their iOS devices.

Read on to find out if Dragon Dictation can improve the way that you enter text on your iPhone, iPad or iPod Touch.

Speech To Text on the iPhone and iPad

Converting speech to text is a processor intensive task. It takes a lot of effort for a computer to record what you say, slice the audio into small pieces that can be statistically compared to samples and then reassemble that information into written, plain text words.

While iOS devices have advanced, optimized processors, they do not compare to the strength of desktop and laptop components. As a result, people are often surprised to hear that full speech-to-text functionality is available for the iPad, iPhone and even iPod Touch iOS devices.

The Dragon Dictation App overcomes the limitations of the iOS CPU by using the Internet. When thought of in context of replacing processing power with the resources of the Internet, the Dragon Dictation App is somewhat revolutionary. iOS programmers are confined to the hardware specs that Apple has laid out. Similarly, end-users can’t upgrade internal components to improve the performance of Apple’s devices. In order to overcome these limitations, Nuance records the voices that are input by the user on the device and then transmits portions of the recording to Nuance servers across the Internet. When the data arrives to powerful Nuance servers, time slicing, comparison and statistical analysis are preformed before the plain text is returned to Dragon Dictation’s App.

This process is optimized and surprisingly fast. While there is a noticeable delay between spoken input and plain text output, understanding the processing requirements and the solution Nuance has created makes this delay bearable if not impressive. Beyond the conversion delay, this process also requires that the Dragon Dictation App be connected to the Internet to function. Also, slow or throttled Internet Service Provider upload speeds can increase the delay between input and output. This limitations can’t really be described as programatic issues or problems because they are a simple side effect of the ingenious solution Nuance has created to overcome the limitations of iOS devices.

Features and Functions

The benefits of using this App are somewhat obvious. By supplementing the standard iOS keyboard with your voice, input speeds increase and generally, the device is easier to use. Using speech-to-text software, especially on the iPad, seems natural and easy. I could see this functionality being built in to iOS devices in the future as a default input method to accompany the often tedious virtual keyboard.

The Dragon Dictation App has three primary states. Each one has been created to preform the seemingly simple task of converting your voice into text.

Recording

The Dragon Dictation user interface was designed to be clean and easy to use. The basic design elements clearly contrast against each other and as a result, your eyes lock to the most important elements. The recording screen has a simple, one-click function. This comes in handy, especially when driving. Without looking at the screen, it is easy to load the app and start recording an email, note, or status update on the go.

Conversion

While Dragon Dictation is converting text, it shows you the plain text output. One of the best ways that I found to use Dragon Dictation was in combination with the app’s built-in keyboard. Making quick edits to text that wasn’t identified correctly or simply switching back-and-forth between speech to text and inputting with the keyboard is a fast, efficient way to compose long or detailed messages.

Output

After the text has been recorded and converted, it can be copied to the iOS system clipboard for use in any app or it can be sent directly to a new e-mail message, Facebook or Twitter. While copying to the clipboard allows universal access to the text, it would be nice if the Dragon Dictation app allowed users to define additional social networks. For example, being able to replace Facebook with Baidu would make the app as friendly to users in China as in North America.

Why You Should Use Dragon Dictation

This is a professionally created, high quality application. The user interface is clear and easy to understand. Inline documentation provides tips, tricks and suggestions about how to get the most from the app. Beyond the design and documentation, this speech-to-text app is surprisingly accurate. While background noise and personal pronunciation can sometimes cause words to be mis-identified, post-recording editing makes it easy to proof-read and correct errors.

Day-to-day, this free app should be part of your standard iPhone, iPad or iPod Touch library of apps. Personally, I’ve been using this app in the car to transcribe e-mail replies and it has been doing a fantastic job. In short, this app overcomes some, but not all of the limitations imposed by the on-screen keyboard found on iOS devices. With improvements to the time it takes to analyze speech and a decreased error rate, this app has the capability to finally move Blackberry owners away from their beloved hard keyboards.