Allophone Plotter


Service “Allophone Plotter” allows to display the sound in graphic form. Using this service one has an opportunity to use allophones’ basis of online version text-to-speech synthesizer, select a specific allophone from the database and get its oscillogram – audio amplitude vs. time plot or spectrogram – plot of audio amplitude vs. time and frequency. Going to the service page (see Figure 1), the user can enter in the appropriate field the name of allophone, which graphical display he/she wants to get, or choose allophone from phonetic database in the drop-down menu.

AllophonePlotter_GUI_2015-03-23

Figure 1 – Interface of Allophone Plotter

When the user tries to get a spectrogram, he/she can choose its black-white or colourful view. After an allophone is chosen the user must press the button and wait until the service processes the information and generates the answer in html-page with the png-picture.

On Figure 2 you can see an oscillogram of the allophone “M002″. By the x-axis time changes are indicated, by the y-axis the amplitude of audio signal is shown.

AllophonePlotter_M002_Osc_2015-03-23Figure 2 – Oscillogram of the allophone “M002″

On Figure 3 a colorful spectrogram of the allophone “M002″ is presented. The plot looks like a two-dimensional colourful view of the signal, where time is indicated by the x-axis, frequency – by the y-axis, and the power of the signal is indicated by the colour intensity.

AllophonePlotter_M002_Spectr_2015-03-23Figure 3 – Spectrogram of the allophone “M002”

The service can be used by phoneticians, linguists, students of philological and pedagogical specializations or just by interested people while learning and studying the Belarusian language.

 

Access to the service via the API

To access the service «Allophone Plotter» via the API, you should send a AJAX-request (type: POST) to the address https://corpus.by/AllophonePlotter/api.php. With an input array data the following parameters are passed:

  • text — allophone. For example, “A001”.
  • voice — recording voice. The available voices:
    • The Belarusian speech: AlesiaBel, AlesiaBel (dictation mode), BorisBel, BorisBel (dictation mode), BorisBelHigh.
    • The Russian speech: AlesiaRus (dictation mode), BorisRus, BorisRus (dictation mode), BorisRusHigh.
  • img_type — type of graphical display. Following types available:
    • oscl — oscillogram.
    • spec — spectrogram.
  • img_color_map — color scale. Following types available:
    • color — colorful color scale.
    • grey — grey color scale.
  • img_contrast — contrast. Following types available:
    • high — high contrast.
    • normal — normal contrast.
  • fft_win_type — weight function. Following types available:
    • hamming — Hamming weight function.
    • hann — Hann weight function.
    • kaiser — Kaiser weight function.

Example of AJAX-request:

$.ajax({
   type: “POST”,
   url: “https://corpus.by/AllophonePlotter/api.php”,
   data:{
      “text”: “A001”,

      “voice”: “BorisBel”,
      “img_type”: “SpecButton”,
      “img_color_map”: “color”,
      “img_contrast”: “high”,
      “fft_win_type “: “hamming”
   },
   success: function(msg){ },
   error: function() { }
});

The server returns a JSON-array with the following parameters:

  • status — request status.
  • img_path — the address where the generated image is located.

For example, the following reply will be formed on the above listed AJAX-request:

[
   {
      “status”: “success”,
      “img_path”: “cache/img/2018-11-08_13-11-29_80-94-171-2_517_spec_color.png”
   }
]

 

Page service: http://corpus.by/AllophonePlotter/?lang=en

Cross references

  1. Zahariev, Vadim Grapheme-to-Phoneme and Phoneme-to-Grapheme Conversion in Belarusian with NooJ for TTS and STT Systems / Vadim Zahariev, Stanislau Lysy, Alena Hiuntar, Yury Hetsevich // Automatic Processing of Natural-Language Electronic Texts with NooJ: 9th International Conference, NooJ 2015, Minsk, Belarus, June 11-13, 2015, Revised Selected Papers / ed. T. Okrut, Y. Hetsevich, M. Silberztein, H. Stanislavenka. — Springer International Publishing, 2016. — P. 137-150.

If you have found a spelling error, please, notify us by selecting that text and pressing Ctrl+Enter.