SIS II and plug-ins

Overview

Voice recordings play an important role in forensic investigations as it may sometimes be the key or even the only evidence of a security threat or a crime. Therefore the need in powerful tools of sound processing was always crucial for audio forensic experts all over the world. For the last 20 years information technology made a big step forward giving an opportunity to create software products helpful in various aspect of life including sound processing as well. However the development in audio industry didn't only make it easier to analyze voice recordings but also raised the requirements and standards to a higher level.

Speech Technology Center (STC) as a world leader in developing technology for processing of audio data always had a goal to create products with an optimal potential for audio analysis that could meet all possible experts requirements and stand any sophisticated challenge. Having made a breakthrough in developing unique algorithms in audio processing, voice matching and noise supression STC have released the newest version of its forensic audio analysis software – SIS II.

SpeechPro, the US subsidiary of STC is proud to present SIS II – a unique software that embraces not only the latest advancements in forensic audio processing capabilities but also includes new analysis and processing tools as well as features designed to improve productivity and plug-in modules for biometric matching and audio recording authentication.

SIS II is the base application for the SpeechPro Forensic Audio Workstation (FAW) which can be provided as a software only or turnkey software & hardware solution and include additional audio processing, analysis, filtering, matching and voice acquisition modules depending on the particular needs of the end-user.

Features

SIS II is the best solution for anyone doing speech analysis. It supports all audio formats including extraction of audio track from .AVI video files and gives you a wealth of options for editing, visualization and analysis.

One of the notable features of SIS II is a set of powerful visualization tools that makes it easier to do all sorts of audio analysis:

• Waveform
• FFT and LPC sonograms
• FFT power spectrum
• Cepstrum
• Autocorrelation
• Pitch tracks
• Formants tracks
• Energy
• Histogram and histograms correlation

SIS II offers multi-window interface and tools for convenient signal comparison:

• Signal overlaying
• Signal synchronization in time and spectral domains
• On-the-fly adjustments of visible speech for better clarity

Besides rich functionality of visualization and analysis SIS II provides a wide choice of editing capabilities geared for forensic purposes:

• Signal normalization
• Signal balancing
• Speakers separation
• Mixing
• Modulation
• Slowing/speeding up of speech tempo (without pitch distortion)
• Re-sampling
• Signal synchronization of several signals
• Splitting stereo signals and separation Stereo or merging of two mono signals into stereo.
• Phase change
• Waveform reversing
• Service real time marks decoding

What’s new

Automatic Speaker Identification Automatic calculation of speech signal characteristics: assesses the suitability of a phonogram for identification purposes. Automatic marking of speech sections of a phonogram, plus the capability of indicating different types of noise. The high reliability of SpeechPro’s automatic identification algorithms has been confirmed by NIST* test results	Working with Projects Create projects and add related files (audio, video, text). Open project files from the SIS II interface.
Easy Text Transcription and Export Selected fragments can be easily assigned to particular categories (e.g. different speakers or sounds) and speech fragments can be displayed as corresponding text. This output can be exported to Word as a text document.	Work with Layers and Transparency Signals can be layered for easy comparison. Colors and transparency can be changed to ease readability. Optimal Window Arrangement Signal windows are arranged conveniently according to task: vertical for identification, horizontal for authentication and noise filtering, or customized according to user preference.
Optimal Window Arrangement Signal windows are arranged conveniently according to task: vertical for identification, horizontal for authentication and noise filtering, or customized according to user preference.	Convenient audio signal navigation Functional navigation oscillogram allows to see the whole signal. The visible section of a waveform is seen as a rectangle in the navigation oscillogram that you can easily move or choose other sections of a waveform.
Quick Text Search Quick search of text transcriptions finds words appearing in two or more waveforms, which is useful for comparison purposes.	Working with Spectrum Filters One of three filters (inverse, harmonic and spectrum-saving) can be created and applied to the signal spectrum, modifying its shape accordingly (e.g. to enhance speech, improve clarity, remove tonal noises, etc.). Filter processing is important in noise filtering and identification.

SIS II Plug-ins

EdiTracker

EdiTracker designed for audio authenticity analysis is a unique software plug-in for SIS II. Paired with SIS II functionality it significantly increases its capabilities in terms of audio authentication. EdiTracker provides six different methods of investigation, automatically searches for suspicious areas of phonogram fragments, visualize all the steps of research and generates text reports.

• Сalculation of a recorder's parameters
• Finding the traces of previous digital processing
• Finding the traces of tampering by the harmonics phase shift
• Background noise scanning
• List of auditory analysis features of tampering
• Built-in instructions for each type of analysis

AIPlugins

Software set of plug-ins that allow for automated speaker identification. AI Plug-ins enables the calculation of speech signal characteristics to determine the suitability of a recording for identification purposes. All plug-ins are automated which makes speaker identification convenient and easy-to-use.

• Automatic calculation of speech signal characteristics: assesses the suitability of a recording for identification purposes
• Automatic marking of speech sections of a phonogram. Plus the capability of indicating different types of noises
• Automatic comparison of two speakers using STC automatic identification algorithms

Videos