Traditional Transcription versus Voice Recognition Software
Voice recognition software vs transcription services is something that many people want to know more about. With constant developments in technology and software for transcription and administrative work, business is changing with many companies moving on to consider saving money by using voice recognition software for dictating letters, writing reports and transcribing interviews. All of our work is carried out by fully qualified, experienced secretaries but we always strive to keep up to date with the latest developments in the transcription industry. So here’s the lowdown on why we believe traditional transcription is still preferential to voice recognition software, and why using voice recognition software can mean a compromised transcript in terms of quality and sometimes (surprisingly) even time spent.
Any experienced transcriptionist knows that for an average audio file, it takes between three and four minutes to type from the spoken word. So surely using voice recognition to type your interviews, and transcribe your focus groups will be quicker, but that’s not always the case. The speech of transcription when using voice recognition software is dependent on many things and these are the issues that newcomers to transcription services don’t always consider.
- If you are typing an interview, you will not be able to play that interview directly to the software and have it transcribed. You will need to listen and dictate what you hear as you go, as the software will be accustomed to your voice and will not be as well equipped as you to deal with poor sound, accents and overspeaking. For this reason, you will still need to have a good ear and understanding of the concept you are typing about, as well as a sound grasp of punctuation to produce a high quality transcript.
- Voice recognition is not always the fastest option for typing services. This is largely due to the fact that any mishears will need to be corrected, and the final document proofread, which takes time. Also, any formatting will also need to be taken into consideration. For example, where punctuation and formatting is concerned, you will need dictate to the speech recognition software which sometimes actually takes longer than pressing the keys. Bearing this in mind, it’s actually quicker to press the full stop key and then enter, rather than say the words ‘Full stop, new line.’ And wait for the software to make that action before you begin to type, as in voice recognition software there is nearly always a lag that does not exist when using a keyboard.
- Voice recognition software needs to be trained to your voice. For words that are initially mistyped due to your accent, you will need to follow a process to teach the software to type those words correctly in the future. An experienced typist may find this process a little trying as they may feel it slows down the rhythm of what they are trying to get onto the document. When we type, it’s easy to follow what we read as we go along.
- Proofreading of transcripts typed using voice recognition software is essential. Typing becomes an autonomic process, like driving and an experienced transcriptionist will think about what they write while they listen, as their fingers do the work. With voice recognition software, the rhythm of speech can be disturbed as effectively, when doing interview transcription, you will be repeating to the software each word as it is said, so it doesn’t follow the order. For instance, you will be thinking just as much about repeating the sentence sections as you are about what the sentence means. For instance, where you would type ‘I would like to order high quality focus group transcription, that would be lovely’, as you heard the words, with voice recognition software, you would hear the words, then repeat them, which is confusing as there is more of a break between the sentence as you speak. For instance, if the words in brackets are counted as your speech, you would hear when using voice recognition software, ‘I would like (I would like) to order high quality focus group (high quality focus group) transcription, (transcription comma), that would be lovely (that would be lovely full stop). This process can be really hard to get used to, and the disjointedness that occurs in the sentences as you hear and speak them can sometimes lead to errors.
These are just some of the reasons that voice recognition software is not all it’s cracked up to be. At the moment, as there is no software sophisticated enough to deal with multiple speakers, punctuation, poor quality sound and different accents, we still need to rely on an experienced typist to carry out transcription services to a high quality. Our guess is that will remain the case for a long while, at least until artificial intelligence is developed and introduced. After all, a large part of transcription is in the understanding of what is said, not just the typing of it. We would only recommend voice recognition software for individual dictation work, but not anything more complex, and even then, the proofreading time can sometimes balance out and negate the benefits.