In this task you will listen to unprocessed and processed speech from a female speaker and try to identify which sounds have been changed the most by the processing. In each recording there are 14 different consonant sounds between two /i/ vowels. These vowel-consonant-vowel (vcv) utterances have been widely used in a range of studies.

There are 6 different conditions to listen to. These are:

• Unprocessed speech
• *Low-pass filtered speech with a simulated audiogram of:

Frequency (Hz)	125	250	500	1000	2000	4000	8000
Threshold (dB HL)	0	0	20	55	70	90	105

• Frequency -compression hearing aid simulation - compresses frequencies above 2.6kHz with a 4:1 compression (i.e. squashes components above 2.6kHz downwards)
• Frequency -compression hearing aid simulation - compresses frequencies above 1.5kHz with a 4:1 compression (i.e. squashes components above 1.5kHz downwards)
• *8 channel cochlear implant simulation
• *32 channel cochlear implant simulation

* these were produced using the AngelSim software.

6 sets of vowel-consonant-vowel utterances have been recorded in unprocessed and various processed forms. The links to the sound files are given below. A pdf file with details of the sounds and with space for response to be written can be found here . Listen to each of these files and indicate in the table what you here. Which consonants are changed most and why? Do the vowels change, why?

The speech sound files can be found below:

---

Unprocessed Speech