| Well maybe as a professional transcriptionist I | | | | at a rate that reflected the 2.5 hours per hour. I |
| could be considered biased, but this article | | | | agreed to purchase the software as I'd been |
| indicates a number of reasons why transcription | | | | toying with the idea of doing so anyway, and do |
| will not be replaced by voice recognition software, | | | | a trial tape for him. I said I wasn't prepared to |
| at least for the foreseeable future. | | | | agree to 2.5 hours per hour until I'd proved it to |
| What is voice recognition software? | | | | myself, and he accepted this. |
| Voice recognition software (the best known | | | | Funnily enough, several months on, I'm still waiting |
| example is probably Dragon Naturally Speaking by | | | | for that trial tape. Having experimented with the |
| Nuance) is software that allows the operator to | | | | software I find that re-speaking actually takes me |
| speak in the computer in order to give the | | | | ...guess how long - once the checking has all been |
| computer commands (e.g. to save a file) or to | | | | done, about four hours per hour of recording! |
| dictate speech into the computer. Over the past | | | | I do use the software now and then, because on |
| ten years it has been hailed a number of times as | | | | a busy week it saves the fingers and wrists, but |
| the answer to fast and accurate dictation - but | | | | what it doesn't save is time! When I'm transcribing |
| we're still waiting. The latest version of Dragon | | | | I proofread as I type. Then just a quick skim |
| claims to be 99% accurate and three times | | | | through is required afterwards to make sure no |
| faster than typing, so I recently purchased a | | | | errors have sneaked through. Using the software |
| copy to see. | | | | to dictate one can probably do this, and there |
| | | | may be a little time saved, but one cannot do this |
| Dictation only? | | | | while re-speaking because one is concentrating on |
| It is important to remember, especially if you are | | | | listening to the words being said and repeating |
| producing interviews rather than one-person | | | | them a few words after, while trying to |
| dictation, that the software will only recognise the | | | | continuously play the recording. |
| one voice it's trained in. If you try to record an | | | | Also because the speech recognition software is |
| interview into it, it will pick up your voice but not | | | | not intelligent it often doesn't know which |
| the interviewee's. It is possible to re-speak an | | | | homonyms to use. This is something that has |
| interview - listen to the interview recording and | | | | been vastly improved over the years, and in |
| speak the words of both interviewer and | | | | simple sentences it is often capable of working |
| interviewee a few words after they're said. | | | | out whether to, two or too is required; or |
| However, that's an art in itself, and even a trained | | | | whether it's here or hear. However in a longer and |
| court stenographer would struggle to listen to an | | | | more complex sentence it frequently struggles |
| interview straight through, with no pausing, and do | | | | with this and very careful proofreading is required |
| this, so how much time does it really save? | | | | afterwards. |
| | | | Also, when you're dictating you don't really want |
| Why is this software not about to take over the | | | | to also have to use the keyboard, so the |
| world? | | | | software allows you to include commands for |
| A recent seminar given by a seller of digital | | | | making words bold, underling them etc. However, |
| dictation systems and associated voice recognition | | | | you need to leave a small pause between the |
| the presenter indicated that voice recognition | | | | command and general dictation or the software |
| would probably only work well for three in ten | | | | just things the command is part of the words |
| people. This is because the software has to be | | | | being dictated. That also slows things up, as |
| first trained to recognise your voice (a process | | | | compared to professional transcribe using |
| which really does only take a few minutes) and | | | | keyboard shortcut keys for these commands. |
| then continuously corrected and trained to | | | | And re-speaking an interview is even worse |
| recognise words that it's either not familiar with or | | | | because every time there's a change of speaker |
| simply doesn't recognise from your speech | | | | that needs to be indicated either with the name |
| patterns, which is an ongoing process and does | | | | or initials and a tab, or the name and a new |
| take time. | | | | paragraph, so there are a number of commands |
| I used this software first about five years ago | | | | involved around that. |
| and I have to say it's come on in leaps and | | | | To save time the latest version of Naturally |
| bounds since then, but even the latest version | | | | Speaking will actually try to punctuate for you ... |
| can be a struggle! I purchased the latest version | | | | so you dictate and it puts the commas and full |
| on the advice of a potential client who claimed | | | | stops in the right places. That's the theory |
| that even re-speaking his interviews (listening to a | | | | anyway. Suffice it to say that I wouldn't |
| recording and then speaking the words into the | | | | recommend it! |
| software as they were said) he could get an hour | | | | |
| of recording down in 2.5 hours rather than the | | | | Conclusion |
| four hours per hour it usually takes to transcribe. | | | | All in all I'd say Dragon 9, the latest version, is a |
| He asked if I would be prepared to use a voice | | | | very useful tool, but it will be some time before |
| recognition program and re-speak his files for him | | | | it's really ready to replace transcriptionists. |