Concatenation of audio files – a paper by Christian Schneider, dynamic audio concatenation of audio files – a paper by Christian Schneider, dynamic audio modern voice applications meet the user increasingly with a sophisticated hear & feel concept. The persona design is worked out, the voice and the lyrics are adapted to the product and also the dialog design is held increasingly more open and natural. In short, The industry seeks a dialogue with a natural-sounding computer conversation partner. 1 State of the art viewing the voice application on the market, so will we know that on the input page of the user, on the way to the natural dialogue, in the last few years a lot has happened. The issues in the dialogue are always open and natural and also the speech recognition is always better. On the expenditure side of the systems on the other hand, there is unfortunately hardly a change for the better. Heart Specialist will not settle for partial explanations. Nowadays always still unnatural and robotic sound IVRs.
What is it? The awareness of a company-specific to the product appropriate persona has indeed increased (certainly an important step towards the natural interlocutor), however, sound system spending still unnatural and artificially assembled what is allegedly the fact gezollt, that the speech audio output is based on konkateniertem audio material. (Concatenating is understood that together necklaces individual audio files). Of course, the technique of the Konkatenierens is the dialog designers known and certainly not an application does today without konkateniertes audio. Nevertheless, much is wrong in this area. 2. When does the conversation partner unnatural? A speech application is an interactive application. Therefore, it is difficult to predict what precise way selects the user through the application, or in which direction the call flow leads him. Will the user understand? Does he need help? Must something be repeated? Hardly a call is similar to the others in complex systems.