Time stretching technology is used for time scale modification (changing the speed or duration) of an audio signal without affecting its pitch.
The algorithm can be implemented both in frequency-domain (FD) using a Fourier transformation, and in time-domain (TD) using Synchronized Overlap-Add (SOLA) technology.
The time-domain modification was chosen for implementation, due to a number of factors:
· time domain algorithm is easier to implement;
· it offers better transformation quality;
· it can be optimized to increase transformation speed;
· the code is easily converted from float point code to any DSP, ARM or RISC platform.
The principle of functioning our algorithm is based on time-domain modification using SOLA technology, such as Waveform Synchronized Overlap-Add (WSOLA).
With Time Stretching speech & audio signals you can easily modify the speed and duration of audio files, avoiding pitch alteration in the same time.
Time stretching technology can be effectively used:
· With audio players for changing the duration of music or speech;
· With audio-video recorder devices or video post synchronization;
· For voice messanging services (voice mail);
· For reduce the rate of bit transmission (jitter control) in devices of IP telephony.
Here are some key features of "Time Stretching speech and audio signals":
· High sound quality;
· High robustness;
· Time-stretch (time-scale) factor: 50%, ..., 200% (normal speed: 100%);
· Guarantees fixed speaker position for stereo signals;
· Dynamic mode changing of time stretch factor;
· Full synchronization with coders of packet length 20 ms and 30 ms;
· Easy integration with target applications.
Requirements:
· Pentium-III class CPU;
· 64 Mb RAM;
· Soundcard;
· Headphones/Speakers;
· Microphone.
Signal requirement (Supported format):
· Windows PCM, 8, 16, 32 bits linear;
· 8 bits A/mu-Law PCM encoder;
· Microsoft ADPCM (MS ADPCM);
· Intel ADPCM (IMA ADPCM);
· Microsoft ACM GSM 6.10 (ACM Waveform).
Limitations:
· 30 days trial
· 20 seconds time limitation for the audio files