GB2229068A

GB2229068A - Playing back recorded speech at faster rate with pitch reduction

Info

Publication number: GB2229068A
Application number: GB8904471A
Authority: GB
Inventors: David Jones; Christopher Hartly
Original assignee: Open University
Current assignee: Open University
Priority date: 1989-02-28
Filing date: 1989-02-28
Publication date: 1990-09-12
Also published as: GB8904471D0

Abstract

Audio information, in particular speech, is enabled to be played back intelligibly at a faster rate than the rate at which it was recorded by reducing the pitch of the audio information. This involves speeding up the original signal, sampling it, converting it to digital values, storing the digital signal, removing components to achieve a reduction in pitch, and reconstituting an audio signal. This is particularly intended for use by the partially sighted. <IMAGE>

Description

AUDIO SYSTEMS SPECIFICATION This invention relates generally to audio systems, and is more particularly concerned with methods of and apparatus for playing back recorded audio information at a faster rate than the rate at which it was recorded.

The present invention is more particularly concerned with methods of and apparatus for enabling people to listen to recorded speech at a faster rate than the rate at which the speech was originally recorded. The present invention has particular application for people with sight deficiencies who cannot read normally and who have to listen to specially recorded tapes. The problem with listening to recorded words is not that the listener has a slower rate of understanding by ear, but that the speaker has a limited rate at which he or she can talk, hence limiting the speed at which information can be assimilated by the listener.

Broadly in accordance with the present invention there is provided a method of and apparatus for varying the pitch of recorded audio information, especially speech information, and in particular reducing the pitch of such audio information, so that the audio information can be played back at a speed faster than the rate at which the audio information was originally recorded.

It is well known that if recorded speech is played back at a faster rate than the original recording rate, then the words will become distorted and "squeaky". In order to overcome this practical problem the pitch of the recorded audio information is lowered.

Digital systems are known which can effect a frequency or pitch shift, but these are expensive and not appropriate for the application of listening to recorded speech. Such digital systems are particularly used for creating special sound effects in the music industry as part of studio type setups. The method and apparatus of the present invention on the other hand are particularly designed and appropriate for speech recordings and for assisting the listener.

With the present invention, in its preferred form, recorded speech is played back at a faster rate than the rate at which it was recorded. The audio information, such as speech, can be recorded on tape or in digital memory. The playback device may have a fixed or a variable rate of playback speed. The user will play the speech back at any speed at which the user can still comprehend the subject matter. In practice, this can be as much as three times that at which the recording took place, which means that a tape that would normally take an hour to listen to can be understood by the listener in only twenty minutes.

The apparatus of the present invention requires an analogue input, but will work with recordings made either with digital or analogue techniques.

One presently preferred embodiment of a speech pitch reducing apparatus is shown in the accompanying block schematic diagram. This is based upon a microcomputer system design, coupled with analogue interfaces. The heart of the system is a microprocessor, with analogue-to-digital and digitalto-analogue converters for interfacing, and with a memory for storage of the signals. As shown in the drawing, a speeded-up audio signal is sampled at a certain rate and is then converted to digital values, which are stored in a digital store. A selection is made from these stored digital values, i.e. some of the samples are removed to achieve the change in pitch, and an audio signal is then reconstituted. This reconstituted audio signal is then transmitted at the appropriate rate to an output to an audio amplifier.

The sound is also filtered to remove the unwanted high frequencies present due to the digital processing. The audio output can be built into the speech pitch reducer, or alternatively one can use the amplifier from the audio source device.

In the presently preferred embodiment of apparatus there are four settings of pitch reduction, 1 (no change), 2/3, 1/2 and 1/3, which correspond respectively to no change, 1. 5, 2 and 3 times increase from normal listening speeds. This gives a reduction in time equal to the pitch setting, i.e. time * 2/3.

More settings could be provided within this range, thus providing smaller incremental changes of pitch.

However, increased speed of listening outside this range will be of reduced value due to the natural limit of human comprehension.

Although the method and apparatus of the present invention are particularly suitable for use by the disabled, there is no restriction on who may benefit from use of the method and apparatus. It is appropriate for use by anyone who has a need to listen to recorded material in a shorter period of time than the original recording.

Also, although the present invention is particularly applicable to the recording of and listening to speech, it will also provide a low-cost way of producing variable pitch of voice or music for special sound effects on any audio output device.

Claims

CLAIMS:

1. A method of enabling recorded audio information, particularly speech, to be played back intelligibly at a faster rate than the rate at which it was recorded, which comprises processing an audio signal carrying said information including effecting a reduction in pitch of said audio information.

2. A method according to claim 1, which comprises the steps of speeding-up an audio signal carrying said information, sampling the speeded-up signal and converting the sample to a digital signal, storing the digital signal, selecting from the stored signal to effect a reduction in pitch, and reconstituting an audio signal from the selected information.

3. A method according to claim 2, in which the selection is effected by removing some components of the stored signal.

4. A method according to claim 2 or 3, which includes filtering the selected signal to remove unwanted high frequencies introduced by the digital processing.

5. A method according to claim 1, substantially as hereinbefore described with reference to the accompanying drawing.

6. Apparatus for enabling recorded audio information, particularly speech, to be played back intelligently at a faster rate than the rate at which it was recorded, comprising means to effect a reduction in the pitch of a signal carrying said audio information.

7. Apparatus according to claim 6, which includes means for speeding up a signal carrying said audio information, means for sampling said speeded-up signal and converting it to a digital signal, storage means for the digital signal, means to effect a selection from the stored signal to effect a reduction in pitch, and means to reconstitute an audio signal from the selected information.

8. Apparatus according to claim 7, which includes filter means to remove unwanted high frequencies introduced by the digital processing.

9. Apparatus according to any of claims 6 to 8, which has means to vary the rate of playback speed.

10. Apparatus according to claim 6, substantially as hereinbefore described with reference to the accompanying drawing.