AIJA PRESENTATION

Technology for Justice 23-25 March 1998

DIGITAL AUDIO AND TRANSCRIPTION

 

David Harrison

General Manager Asia / Pacific

Future Technology Resources Pty Ltd

Enterprise Unit 1 / Suite 2

11 Brodie Hall Drive

Bentley Perth W.A.

Ph: 08-94704366

Fax: 08-94704310

E-Mail: david@ftr.microweb.com.au

 

 

 

Introduction

 

 

Verbatim transcription methods have developed out of the need to provide a long term accurate court record for use by the all parties involved within current and appeal based litigation and to also provide public access to an accurate record of the facts of the case. Transcription technology has grown as an extension of the practice of note taking by the Judge which in many judicial environments outside Australia still remains the official court record.

 

There are currently a range of methodologies to obtain a court record. These include the traditional pen based Stenographic reporting techniques through to the verbatim technologies represented by Steno typing, video and audio recording.

 

I will endeavor to outline how these transcription methods are benefiting from the introduction of digital audio using commercial installations within Australia and overseas.

 

What is Digital Audio

From an operational view point digital audio is a technology by which it is possible to mathematically describe an analogue signal, the sound a human ear can hear. As can be seen from the diagram this is achieved by sampling the audio at discrete points then assigning a numeric value to the point. Sample sufficient points ( twice the frequency of the original analogue waveform ) and enough information is gathered to allow recreation of the analogue waveform from the numeric data at any time.

 

The key advantage of the digital audio is that once digitised the numeric data can be manipulated mathematically.

This ability has seen it permeate our society from the telephone calls we make through to the music we listen to in our cars.

 

 

 

The Advantages Afforded by Digital Audio.

 

To be more specific the advantages of digital audio are as follows;

 

  1. A Format Readily Adopted by Computer Technology

A format which can be stored indefinitely as a file on computer based systems, manipulated and retrieved at will.

 

2. Fast Search of Audio Files

This allows audio stored to be retrieved instantly and played back through standard headphones/speakers from any point in the audio. Thus unlike tape based recorders which require you to arrive at a point in the audio based on the fast forward or rewind functions, sometimes taking minutes, digital audio enables this operation to be achieved in sub-second times.

 

3. Random Access of Audio Files

Because the digital audio file is random access it means that the same file (which may be 2 hours long ) can be accessed by multiple operators simultaneously. This allows for instance 10 transcription operators to access a 2 hour recording simultaneously each taking 12 minutes of audio.

 

With tape based recording the 2 hour tape could only be accessed by one person at a time. To achieve the parallel efficiency of the digital system it would be required to take the 2 hour tape and re-record this into 10 twelve minute tapes then deliver them to each transcription operator. This is a far more labour intensive operation than that represented by the digital audio solution.

 

4. Audio Quality

In an analogue based system the tapes will degrade in quality through age. Also each time an original tape is copied the successive recording will exhibit a correspondingly lower level of quality. In the worst cases this can render the audio indecipherable and thus reduce the accuracy of the transcription below the required 100%.

 

In comparison digital data is more robust and will allow the same file to be copied thousands of times with no reduction in the original quality. Storage media such as compact disks ensure the long term stability of the court audio record.

 

  1. Audio Copy Speed
  2. Copying of analogue tapes is measured in minutes while similar lengths of digital audio can be copied in seconds.

     

  3. Text and Data Base Integration.
  4. It is possible to embed text information within the audio file which will ensure that text based information entered during the court case will be available at any time the audio is accessed. Through database links it is possible to search the audio based on time and date or through the embedded text within the audio. Using these facilities the ability to locate the audio is rapidly enhanced over previous tape methods.

     

  5. The Digital Domain

Existing in the digital domain digital audio can take advantage of the proliferation of local and wide area digital networks. This infrastructure provides the vehicle by which the audio can move electronically from one geographic location to the next. This may be from building to building within a city, state, nationally or internationally. This also enhances the work at home opportunities for the transcription industry.

 

The advantage of this is that it for the first time raises the possibility of the virtual transcription office gaining economies of scale by transmitting digital audio between these offices. The peaks and troughs of transcription orders can then be shared between the offices ensuring transcription staff have a consistent work load and overheads are kept to a minimum. This methodology has the potential to be used for all specified transcription delivery times.

 

The idea of load sharing is not new within the transcription environment, it is being carried out to some extent through the use of manual transportation of audio tapes. This process is only feasible for transcription that has a specified turnaround time measured in days.

 

Application to Voice Recognition

Digital audio enables the technology of Voice Recognition to exist. This technology supports direct voice to text conversion for speaker dependant large vocabulary environments. In this situation it is possible to reach near 100% accuracy. The ability to apply this to the high vocabulary speaker independent court room environment and achieve the required 100% accuracy has yet to be successfully implemented.

 

The Current Transcription Process as it Relates to Analogue Tape Based Electronic Recording.

 

 

By far the most common method of electronic recording within Australia is the in-court cassette tape recorder. This is normally controlled by an in-court monitor who ensures changes of cassettes are facilitated and provides hand written notes of a kind which will assist the transcription operator in generating the verbatim record. Information such as barrister change, witness names etc. This information can be vital in the operation of audio based transcription as they assist in overcoming the problem of out of court transcription.

 

The number of tapes generated within a court can vary from 1 to in excess of 30 depending on the circumstance. In the worst case the court will require the transcript same day which will necessitate the tapes being limited to approximately 10 minutes. This allows the tapes to be removed from the court on a regular basis during proceedings then delivered to the transcription office. The short tapes allow multiple operators to type in parallel. Typically to type 1 hour of audio will take 3 to 4 audio typists working continuously approximately 1 hour.

The playback is facilitated via the use of a cassette player and the typing is performed on computer based word processor. The paper logs from the court are normally delivered wrapped around the tape to help to ensure their match to the tape. Loss of the written log notes will seriously hamper the generation of the verbatim record.

 

On completion of the transcription the turn of transcript is manually merged with the numerous other cuts being generated by the other transcription operators. The final merged copy represents the official court record.

 

 

 

 

 

Digital Audio Recording and how it is Applied to the Current Transcription Process.

 

 

In its most basic format a digital audio recorder would replace the current analogue recorder in court. In the absence of a computer network ( which would allow electronic delivery to a remote location ) a high capacity archive media can be used to manually transfer the audio record to the transcription office.

 

Within the court a monitor person is present to facilitate start and stop of recording. There is no longer the need to change the recording media as the audio is buffered on the computers hard disk. The Monitor now enters the court log notes directly into the computer. They no longer need to register the times against these notes because the note is automatically inserted directly and permanently into the audio file. Private notes can also be inserted by other court participants including the Judge.

 

On receipt of the digital audio, manually via archive disk or directly via a network, the transcription staff can immediately begin transcription.

 

As per the photo it can be seen that for the typical transcription workstation the tapes, tape player and paper log from the court has now been removed. To replace these we now have an on-screen audio playback panel representing what was the external cassette player and a log notes window showing all the monitor notes that are embedded in the audio. These would have originally existed on paper. These facilities all reside simultaneously with the word processor. This allows playback of audio, a view of the log notes and transcription all to be handled simultaneously by the computer. There is also parallel footpedal control allowing hands free operation of the audio playback.

 

 

 

 

 

 

 

 

 

 

 

 

 

Case Studies of Current Usage of Digital Audio for the Generation of Transcript.

 

The following information will be referencing operational sites both within Australia and Internationally that operate using digital audio as their court record. The sample visited will represent a cross section of the techniques currently in practice. It should be stressed that these are all commercial operations. The distinction between the sites will be based on what method is used for the delivery of the digital audio record to the transcription group. This provides a distinctive method of describing the different processes under which digital audio is used.

 

Advanced Integration

The Judiciary within Hong Kong has implemented over the past 4 years the most advanced Digital Audio Recording and Transcription environment in the world. This technology extends throughout the court system and numbers in excess of 150 courts.

 

 

 

The site shown here is the Wanchai District Court on Hong Kong Island and the number of courts here totals 45. All 45 courts are recorded on computer and monitored at this central position within the building. Monitoring is performed via closed circuit video of the courts allowing the operators to remotely start and stop recording. In addition text information concerning the proceedings can be entered directly into the computer, automatically embedding itself into the audio. All recording operation is transparent to the courts, allowing a free flow of proceedings.

 

The audio is recorded, stored and managed within this central location. In this case the transcription operators are also located within the same room and are connected via computer network to the computer servers storing the digital audio.

 

On recording the system automatically cuts the digital audio into user defined lengths ( typically 10 minutes ) . These are then ready for delivery via the computer network to any transcription operator who may request it. Once requested the audio is played back from the computer and typed. Given sufficient transcription staff assigned to the one court it is possible to maintain a running transcript for the court which at any one time would be only 15 minutes behind the proceedings. The digital audio system is being employed for both the generation of English and Chinese transcript.

 

One of the great advantages of this site is that all audio is held in a central pool which allows immediate access by any of the transcription operators to any of the current or past recordings from all the 45 courts. This is all achieved without the transcription operator having to leave their computer or employment of any intermediary.

 

In comparison the use of analogue tapes would require the use of thousands of tapes annually, initiating all the problems of a labour intensive activity in relation to management, storage and retrieval.

 

As a by product of this operation all Judges ( in chambers ) and courts have been provided with the facilities to access the audio at any time for review via supplied computer workstation.

 

 

Goulburn Street Family Law Courts in Sydney

 

A total of twelve courts are recorded centrally using digital audio techniques in the Goulburn Street Family Law Courts building in Sydney.

 

Under the reel to reel tape based system that existed previously once the court proceedings were complete on all the courts, for that day, the reel to reel tape was available for transfer off site to the transcription centre located at the Auscript offices. Because of the low transcript requirements the manual transfer of the tape provided the most economical means of delivery.

 

Where problems would occur is when it was required to identify a segment of a case that had been requested for transcription. The analogue tape held hours of court recordings and it was a very labour intensive operation to identify required audio. When found it was then necessary to manually dub the audio to a cassette tape for delivery to the transcription operator.

In an effort to streamline this operation digital audio was introduced into the operation. Transfer from the courts to the transcription office is still manual employing digital archive media. The benefits become apparent when it necessary to locate audio that has been requested for transcription. Via its embedded database information digital audio allows rapid location of court, case and time. This can be carried out by the transcription operator who then has the option of accessing the audio via the installed network for playback and transcription.

 

It has been estimated that digital audio has provided a 20% increase in efficiency over the previous tape based method of transcription generation.

 

Geographically Separated Transcription Operations.

 

Within the Queensland State Reporting there is an initiative to investigate the feasibility of sharing the transcription load in Brisbane with transcription operations within their offices in country centres.

 

This site has been installed using digital audio technology with delivery of the audio from Brisbane to the remote office via the public digital communications network. The system will be used as a peak load buffer in that it will be used if and when it is decided that completion of the transcript in the Brisbane office for that day will not be possible due to lack of resources.

The aim is to achieve a situation in which all transcription staff are provided with a consistent work load no matter what their geographic location and availability of work is not solely dependent on the courts being in session in a particular location.

 

 

In Court Digital Audio Recorders

 

This is representative of the area in which digital audio recorders would provide direct replacement for current analogue equipment. The digital recorders reside in court and are controlled by an in court monitor.

 

This type of system is currently the most prolific of installations due to the following reasons;

 

  1. It minimises any change required to the current familiar process of audio recording and subsequent transcription.
  2. It provides a low cost introduction to the application of digital audio.
  3. It offers a base platform for the application of the benefits available through digital audio.

 

The transcription office will take the digital audio either electronically or through standard removable computer disk technology.

 

 

Thailand

 

Installations within Thailand have applied digital audio technology to extend the use of the current practice of the court record being based on the presiding Judges bench notes.

 

In this system the Judge and clerk have computer workstations within court connected by a computer network. The Judge wares a lapel microphone and speaks all his notes as the trial progresses. These notes are recorded as digital audio and simultaneously played back from the computer by the clerk. The clerk generates a typed copy of the Judge notes from the audio recording and this transcript is seen in real time on the Judge computer. The Judge can make alterations to the transcript as required in real time. At the end of the proceedings it is simply necessary for the clerk to print the transcript ( within the court ) and the Judge leaves with a hard copy of the days notes. This is the official court record.

 

This degree of integration and transcript delivery would not be possible without the features provided by digital audio.

 

Real Time transcript

 

Digital audio is being used on a trial basis integrated with Real Time transcript production using stenographic methods. This is being carried out within the Supreme Court in Melbourne.

 

The court reporter is in court supplying Real Time transcript on-screen to the court participants. An additional operator is editing the initial transcript in real time to ensure 100% accuracy. In the pursuit of ensuring that the transcript is verbatim, digital audio recording has been integrated into the system.

 

The scopist’s computer workstation is used to simultaneously record digital audio while the initial transcript is appearing in real time. The digital audio on the scoping machine can be played back at any point while recording is still under way. Thus as the editing function is taking place the audio can be delayed to run in synchronisation with the appearance of the initial version of the transcript. The scoping function thus allows the appearance of the text to be checked against the original audio ensuring that the Real Time transcript appearing in court has the desired 100% accuracy.

 

This procedure is only possible through the use of the features provided by digital audio.

 

 

Conclusion

 

There is not one overriding advantage of digital audio which makes it a "must have" commodity for all areas of the Judicial environment. Rather it represents a suite of persuasive advantages over its analogue predecessor which can be applied to the Judicial process to attain specific cost / benefit goals.

 

Support for digital audio will also become more compelling as we see the spread of information technology throughout the Judicial environment. This, supported by the falling electronic communication and computer hardware costs and increased computer literacy, will see a number of the current barriers to digital audio fall.

 

The Judicial electronic recording and transcription environment due to its specific requirements, has in the most part managed to deflect the infiltration digital audio has made into all other electronic recording industries.

 

This will inevitably change in the future.