DECtalk[4] was a speech synthesizer and text-to-speech technology developed by Digital Equipment Corporation in 1983,[1] based largely on the work of Dennis Klatt at MIT, whose source-filter algorithm was variously known as KlattTalk or MITalk.[5][6]
Uses ranged from interacting with the public to allowing those with speech disabilities to verbalize, include giving a public speech.[7][8]
Announced December 1983, a trickle came February 1984; larger DECtalk quantities were delivered in March.[9]
They were standalone units that connected to any device with an asynchronous serial port. These units were also able to connect to the telephone system by having two telephone jacks. One connected to a phone line, the other to a telephone. The DECtalk units could recognize and generate any telephone touch tone. With that capability the units could be used to automate various telephone-related tasks by handling both incoming and outgoing calls. This included acting as an interface to an email system and the capability to function as an alerting system by utilizing the ability to place calls and interact via touch tones with the person answering the phone.
Later units were produced for PCs with ISA bus slots. In addition, various software implementations were produced, most notably the DECtalk Access32. Such implementations began as explorations into real-time software synthesis on general purpose CPUs,[10]: 2 subsequently delivering a DECtalk Software product for Digital Unix and for Windows NT on Alpha and Intel processors.[11] Certain versions of the synthesizer were prone to undesirable characteristics. For example, the alveolar stops were often assimilated as sounding more like dental stops. Also, versions such as Access32 would produce faint electronic beeps at the end of phrases.
In the final years, early/mid-2000,[12] the DECtalk IP was sold to Force Computers, Inc. In December 2001, the IP was sold[13] from Force Computers, Inc, to Fonix Speech, Inc. (now SpeechFX, Inc.), which offers DECtalk as a small-footprint TTS system and in a computer program form.[14]
The New York Times wrote: "like a scratchy recording of a person with a lisp" but added "usually understandable."[4]
DECtalk had a number of built-in voices which were identified by the following names: Perfect Paul (the default voice), Beautiful Betty, Huge Harry, Frail Frank, Kit the Kid, Rough Rita, Uppity Ursula, Doctor Dennis and Whispering Wendy. Each of the voices were editable by adjusting various parameters (such as throat size, crossover frequencies, etc.).
DECtalk understood phonetic spellings of words, allowing customized pronunciation of unusual words. These phonetic spellings could also include a pitch and duration notation which DECtalk would use when enunciating the phonetic components. This allowed DECtalk to sing.
the audience heard the DECtalk, voicing words that the educator typed into his computer.
{{cite web}}
Lokasi Pengunjung: 18.219.229.160