Voice analysis using software products of DiagNova Technologies:

  • DiagnoScope (previously named DiagnoVoice) – product for full vocal folds analysis and video recording from endoscope, ultrasonograph or any medical imaging device
  • DiagnoStat – software for voice analysis parameters determination of a single sample and a group of files; analysis resoults can be previewed in tabular form, what enable further statistic analysis
  • Low-level voice analysis modules, used in third-party applications since 2002
  • VoicePrints – voice samples comparative analysis environment for comparative analysis of two voice samples containing full spectrographic and parametric analysis

DiagnoScope – voice analysis and video recording environment

DiagnoScope (the previous version, for phoniatry only, was called DiagnoVoice) is our original software for full voice analysis with new advanced and enhanced modules. The version DiagnoScope Specialist (previously named DiagnoVoice Basic), designed for individual practitioners, is available since the beginning of 2010. The version DiagnoScope Clinic (previously named DiagnoVoice Lite) is available. This version is designed for polyclinics.

Characteristics of DiagnoScope Specialist:

  • “a” vowel analysis,
  • phrase analysis,
  • long phrase analysis,
  • voice capacity analysis,
  • vocal loading test,
  • singers’ voice analysis,
  • speech disfluency analysis (stuttering),
  • voice recording,
  • voice processing with recording quality control,
  • recording in advance,
  • recording based on model curve,
  • fundamental frequency (F0) analysis,
  • parametric analysis, formant analysis,
  • long-term analysis,
  • inverse filtering,
  • scale measurement,
  • LTAS long-term spectrum,
  • creating oscillograms and spectrograms,
  • automatic determining the length of phonation,
  • automatic, configurable reporting,
  • saving on any data-storage medium informations about report with all data,
  • templates for specialist examinations,
  • templates for referrals and medical certificates,
  • installation and full configuration of program,
  • modular construction allows individual configuration,
  • optional modules on clients demand,
  • attendance of chosen HD cameras,
  • working in MS Windows XP, Vista, 7 or 8 environment,
  • additional video recording module.

DiagnoScope Specialist modules:

  • “A” vowel analysis – voice sample analysis based on vowel “a” extended phonation with possibility to evaluate voice acoustic parameters using standards, allows full voice folds diagnostics, also for detection of laryngeal cancer,
  • Phrase analysis – voice sample analysis based on any phrase or sentence, allows to evaluate words intonation, and also allows sample comparison to examine any changes,
  • Long phrase analysis – allows to record long phrase or whole long speech, to analyze a fragment of it later using “a” vowel analysis or phrase analysis,
  • Voice organ capacity analysis –voice capacity analysis based on time measurement and long-term parameters of the longest phonation, allows to evaluate voice capacity,
  • Vocal loading test – vocal loading test allows for voice quality assessment as a consequence of increasing vocal effort; defines acoustic parameters value changes during reading a long text,
  • Singers’ voice analysis – voice field analysis, scale measurement with evaluation of the vocal folds near the limits of the scale and near the register transitions, support voice training, voice emission exercises, and also allows to diagnose voice disorders like dysphonia.
  • Speech disfluency analysis – voice sample analysis in segmental form, which is a vowel, phrase or sentence repeated several times; allows for speech fluency quality assessment and can be used in the assessment of stuttering.

Features:

  • Voice recording recording using a microphone
    • simple sound editor functions,
    • record quality checking,
    • setting the length of recording;
    • recording a very long speech (long phrase analysis);
    • recording in advance (voice capacity analysis);
    • recording “a a” extended phonation every 60 seconds during increasing vocal loading (vocal loaded test);
    • recording a vowel, phrase or sentence repeated 3 to 30 times (speech disfluency analysis);

    Voice recording window
  • Singers’ voice recording:
    • smooth frequency change recorded with time,
    • model frequency given on screen in the form of a curve, or as a reference signal,
    • fundamental voice frequency relative to frequency of the model curve,
    • model frequency matched to singers’ vocal scale;

    Singers’ voice sample recording window with model curve
  • Fundamental frequency (F0) analysis:
    • mean,
    • maximum and minimum values,
    • stability,
    • enhanced statistics,
    • pitch chart;

    Example output of a fundamental frequency analysis
  • Inverse filtering:
    • glottal source signal estimation,
    • glottal closure moments detection;
  • Oscillograms with zoom function;
  • Spectrograms: multiple types (wide- and narrow-band, resonant, F0-synchronized, enhanced contrast with formants removed);
  • Parametric analysis
    • advanced statistics,
    • all classical parameters (Jitter, Shimmer, NHR, Yanagihara, Unharmonic-to-Harmonic, and much more);
    • long-term parameters analysis (voice organ capacity analysis),
    • parameter analysis against fundamental frequency (singers’ voice analysis);
    • specific parameters for speech disfluency assessment (speech disfluency analysis);
  • Formant analysis;
  • Voice profile;
  • Scale measurement based on phonation frequency range;
  • Long-Term Average Spectrum (LTAS) type of a long-term spectrum:
    • information, what portion of the total energy is transmitted in a specific frequency band,
    • signal power mean value as a function of frequency,
    • spectrum determined in every acoustic analysis,
    • determine also LTAS parameters;
  • Voice capacity:
    • automatic determination of phonation length without manual measurement,
    • results automatically putted in the report,
    • special long-term parameters;
  • Vocal loading:
    • voice acoustic parameters assessment during vocal effort,
    • recording test phrase (in example “a a” extended phonation) every 60 seconds,
    • vocal effort by reading very long text,
    • automatically displayed text to read,
    • automatically displayed message asking to say phrase;
  • Visual chart editor
  • Advanced semi-automatic, configurable reporting: visit report forms are:
    • automatically created from templates,
    • automatically filled with selected images,
    • can be edited in a visual editor;

    Automatically created and filled visit report
  • Built-in expert system for computer-aided diagnosing;
  • Additional documents:
    • optional templates created on clients demand for referrals and medical certificates,
    • fast access from main window,
    • easy template edition if needed;
  • Medical history:
    • storing of patient data and visit reports,
    • instant access,
    • easy browsing,
    • data comparison functions,
    • complement of the existing report.

    Main program window with patient data and visit reports

DiagnoScope allows full acoustic diagnosis and diagnosis of laryngeal cancer, singer’s nodule, laryngeal diseases or occupational diseases to vote. In addition to voice analysis, DiagnoScope is designed for endoscopic video recording and allows complex phoniatric and laryngological examination. The program includes full video imaging support for laryngovideostroboscopy.

Hardware requirements:

Minimum requirements (for video signal recording with maximum resolution 352×288)

  • Processor: Intel Pentium III or later with clock frequency min. 800 MHz, AMD Athlon 64 or later with clock frequency min. 800 MHz;
  • Sound card with DirectX 9 compatible drivers;
  • Operating system: MS Windows XP 32 bit, MS Windows Vista 32 bit or later 32 bit;
  • RAM at least 512 MB (for Windows XP) or at least 768 MB (Vista);
  • Integrated graphics card compatible with DirectX 9;
  • Hardware to convert analog video signal (frame grabber) compatible with DirectShow (DirectX 9), external (USB) or in the form of expansion card;
  • 24 (32) bit color depth and screen resolution at least 1024×768 pixels.

Minimum requirements (for video recording signal with resolution above 352×288) as above, except:

  • Processor: at least dual-core: Intel Core 2 Duo or later with clock frequency min. 1800 MHz, AMD Athlon 64 X2 or later with clock frequency min. 2500 MHz;
  • RAM at least 1024 MB (for Windows XP) or at least 2048 MB (Vista, Windows 7);

Recommended

  • Processor: at least dual-core: Intel Core 2 Duo or later with clock frequency min. 1800 MHz, AMD Athlon 64 X2 or later with clock frequency min. 2500 MHz;
  • Sound card with DirectX 9 compatible drivers;
  • Operating system: Windows XP 32 bit, MS Windows Vista 32 bit; Windows 7 32 bit or 64 bit;
  • RAM at least 1024 MB (for Windows XP) or at least 2048 MB (Vista, Windows 7);
  • Graphics card compatible with DirectX 9 with dedicated memory at least 64MB (preferred cards NVidia from 8400 family or later);
  • Hardware to convert analog video signal (frame grabber) compatible with DirectShow (DirectX 9) in the form of expansion cards (not USB);
  • 32 bit color depth and screen resolution at least 1440×900 pixels, optimal 1920×1200.

In parallel, we are working on a simple and cheap glottograph in order to provide comprehensive solutions for phoniatry.

DiagnoStat – acoustic analysis parameters determination environment

DiagnoStat Specialist allows to determine acoustic analysis parameters of a group of samples. Analysis resoults can be presented in a tabular form and used for further statistic analysis. The program is based on samples generated by acoustic analysis software DiagnoScope Specialist. However software basic function is to determine single numerical values describing voice sample.

DiagnoStat Specialist features:

  • Supported files of almost any type and origin:
    • condition – files contain fundamental frequency (F0) analysis results and analysis intervals,
    • recommended files *.ost are saved using DiagnoScope Specialist (files contain all necessary information);
  • Easy to operate over a large number of files – currently the file number limit is 200
  • DiagnoStat main window
  • F0 and analysis intervals determination control:
    • allows to control determined F0 and analysis intervals,
    • registered version – allows file edition (e.g. in doubtful cases fundamental frequency re-determined);
  • Determined analysis intervals control and edition window

    Voice analysis modules

    Low-level voice analysis modules developed by the founders of DiagNova Technologies, currently owned by the Company, are used in third-party applications since 2002.

    Example output of the old voice analysis modules: color spectrogram

    VoicePrints – voice samples comparative analysis environment

    VoicePrints is our original software for comparative analysis of two voice samples aimed at identification of speakers. New improved version of VoicePrints contains new features that improves work much. It is used internally in our company for on-demand specialist comparative analysis of voice samples.

    Features:

    • Analysis divided into two parts:
      • Sample editor – first part of analysis allows to divide recording to single samples easy to compare and to classify them to groups,
      • Sample analyzer – second part of analysis that allows to analyze samples to obtain acoustic parameters that describes samples and then compare with each other;
    • Parallel operation on two voice samples;
    • Direct access to all samples from analyzed class;
    • Several steps (increasingly precise) in cutting samples from analyzed recording;

      Cutting samples from analyzed recording
    • Division of samples into classes according to the spoken phrase or word;
    • All analyses found in DiagnoScope;
    • Sample preprocessing with background removal and voice enhancement tools;

      Example output of a sample preprocessing
    • Advanced comparison using standard acoustic parameters;

      Example output of the VoicePrints: comparison by acoustic parameters
    • Reference sample – sample selected by the user, which is compared with the rest of the samples;
    • Simple copying diagrams from program to any popular text editor.