Multi-Speech, Model 3700

Multi-Speech, Model 3700, is a low-cost, Windows®-based, speech analysis program, which uses standard multimedia hardware (e.g., Sound Blaster™ boards) to capture, analyze, and play speech samples. Multi-Speech brings KayPENTAX quality to the Windows-based software product category.

Applications

Multi-Speech is a comprehensive speech recording, analysis, feedback, and measurement software program. It includes the same analysis features as CSL, the most widely used speech analysis system. Multi-Speech also has numerous application-specific software options. It is only limited by the specifications, features, and S/N limitations typical of audio devices in the host computer. If you are using a professional-level device (e.g., CSL, DAT, ADAT, etc.), or your application does not require robust acoustic measurements, Multi-Speech is ideal. For measurements, or when budgets allow, consider a CSL system. Multi-Speech is also ideal as an option for CSL so that multiple users can analyze signals acquired by CSL.

Multi-Speech, with the caveats noted above, can be used for teaching, research, voice measurements, clinical feedback, second language articulation, and forensic work.

The screen above is from a sample macro and signal file delivered with Multi-Speech. Window A shows the IPA symbols time-linked to the speech waveform with glottal impulse markers. Other windows show a spectrogram with formant tracking, a pitch contour, an amplitude contour, and a palatogram, all linked together so that cursor movement shows the different palatometric patterns with the associated acoustic analysis.

Easy to Use

Although powerful, Multi-Speech is easy to use. Extensive online Help and tutorials guide the user. Commonly performed functions are on the toolbar or are accessed by defined keys (e.g., the F3 key plays the signal). All functions are available on the pull-down menus. A feature called macros stores a list of functions which can be recalled. Multi-Speech is delivered with many pre-written macros, and users can create their own as well. Macros make it easy to automate teaching lessons or clinical protocols. For example, Multi-Speech includes a macro that loads a speech file, filters the signal to sound like a phone connection, and illustrates the spectrum changes typical of a phone connection. Other macros automate clinical protocols so that evaluation and feedback with clients is quick and easy. Reports are easy because all graphic displays include complete numerical analysis and automatic statistical analysis.

Powerful

Multi-Speech software’s ease of use is complemented by the powerful feature set of the program. Every function can be adjusted to suit even advanced research requirements. For example, the spectrographic analysis has adjustments for nine filter bandwidths, four window weightings, full scaling, color, pre-emphasis, and dynamic range.

Comprehensive Graphical and Numerical Analyses

Comprehensive graphic and numerical analyses are available
for the rich array of easy-to-use Multi-Speech features. The
LPC-derived formant frequencies are shown above.

Example of multi-speech screen with waveform (A), colored spectrogram (B), gray scale spectrogram with formant history (C), and LPC formant analysis (D). All graphic views have cursors and numerical analysis. All views are sizable and movable.

Features

Capture and play with any sampling rate
Up to eight channels of data
Read and write to CSL or .wav file format
Complete IPA Symbols
Editing includes cutting, pasting, digital filtering, and appending
Complete online Help and tutorial
Analysis includes multiple pitch extraction methods, pitch-synchronous LPC, glottal event markers, FFT, formant history, spectrogram, energy, etc.
Logging of cursor values
Macros to repeat operational steps
Learn mode to memorize steps
Tags, storable with signal
Complete suite of features for each analysis and editing command
All 32-bit, MDI-compliant for fast operation

The Video Phonetics Program shown above is one of several optional programs for Multi-Speech for targeted applications. Programs are targeted at clinical, teaching, and research applications.

Summary

Multi-Speech is a low-cost, powerful, and easy-to-use speech analysis program with many optional programs. These are ideal for research, teaching, and clinical applications. Multi-Speech, because it is based on the CSL (the leading speech analysis system), includes a wide array of functions and features for speech professionals.

Multi-Speech requires a host PC computer (Pentium >266MHz, CD-ROM, 16 MB RAM) and sound device.

Current Multi-Speech, Model 3700, software and database options include:

Analysis-Synthesis Laboratory (ASL), Model 5104

Applied Speech Science for Dysarthrias, Model 5153

Auditory Feedback Tools, Model 3506

Disordered Voice Database, Model 4337

Games, Model 5167

Motor Speech Profile, Model 5141

Multi-Dimensional Voice Program, Model 5105

Palatometer Database, Model 4333

Phonetic & Perception Simulation Programs, Model 5151

Phonetic Database, Model 4332

Real-Time EGG Analysis, Model 5138

Real-Time Pitch, Model 5121

Real-Time Spectrogram, Model 5129

Respiration, Phonation and Prosody Simulation, Model 5152

Signal Enhancement Program, Model 5142

Sona-Match, Model 5127

Speech Articulation: Animation of Muscle Vectors, Model 5154

Video Phonetics Program and Database, Model 5150

Voice Range Profile, Model 4326

Multi-Speech is compatible with Windows 95, 98, ME, NT, 2000 and XP.

Click here to view this document in PDF format