Speech analyser is a fantastic tool that I use when learning languages. It lets us graphically see the way we speak and when used properly, is a great tool in helping us sound more ‘natural’ when speaking a foreign language.
You can install it from the CD ROM that you have or download it from [http://www.sil.org/computing/speechtools/speechanalyzer.htm](http://www.sil.org/computing/speechtools/speechanalyzer.htm)
If you’re installing it from the CD ROM, go to the folder ‘Speech’ analyser and click on the file Setup-ST15min.exe. It is at path:D:\ Linguistics\Speech Analyzer\Setup-ST15min.exe. .This will install Speech Manager and Speech analyser.
We will be concerned with the Speech Analyzer programme. The Speech Manager is a programme that is used to coordinate sound files used in conjunction with Speech Manager – so its useful, but not necessary.
When you run the programme, you will get a screen like this:
If you would like to record your voice straight off, click ‘record’, otherwise just click ‘close’. To record your voice, if you aren’t already in record mode, click the microphone icon in the tool bar. The settings for recording are quite straight forward – just like using a cassette machine. I would suggest recording a Thai saying some words and vowels, and then you say them straight afterward. You can then compare them. You can also go to ‘open’ and open any of the audio files from the CD ROM that are saved as wav files in the ‘Learn Thai’ section.
Once you’ve recorded your sounds, you’ll get a screen like this. I recorded the words from Session 3 that we used for sample tones in English – Why (High Tone) Why (Rising Tone) Why (Falling Tone) – Yeah (Common Tone) Yeah (Low Tone).
Here you can see the sound wave file and also the pitch / tone in graphic form. You can use the bottom graph to check whether you’re actually producing the correct tone. Ask a Thai to say the same words / tones as you and see whether or not your curves come out the same.
Another great tool is the vowel analyser – in this program it is the ‘F2 – F1’ window. To get to this window, click on the icon in the toolbar that looks like a ‘graph’ (just next to the pencil icon). Select option 5.
In this section, you have to set in 2 bars – red and green to mark out a section of the recorded file that is a single vowel that you want to analyse.
Once you do it, the vowel will be automatically plotted on the graph in the bottom right hand side.

If you haven’t seen this kind of grid before, it is basically a map of the mouth. THe ‘i’ is at the top left hand corner, represents the top front of the mouth – where ‘i’ is produced. The ‘u’ is the back top of the mouth cavity (toward the back of the throat). When learning a language, pronouncing the vowels as natives of that language pronounce them is invaluable. It will help fluency and understanding to no end.

Other functions you can try out – Create a spectogram by clicking on ‘recalc’ in the bottom left hand window once you’ve selected a region with your green and red bars. You can also go and choose from a whole lot of different graph types by clicking on the graph icon up top. Some of the graphs include plotting the musical note of your voice input and other pitch graphs.
I hope this quick guide has been useful!
Happy voice analysing!