Ever since voice recognition transcription software became viable, this topic comes up from time to time. Which is better? Human or software? Sarah Connor or T-1000? Getting your recordings transcribed by a professional transcription service or buying a voice recognition software? We have had our fair share of experience with voice recognition software. Why wouldn’t we? If we can save money on hiring transcribers, good for us! But our experience has been less than satisfactory. Unless the speaker talks really slow, enunciates each word clearly, and there is no background disturbance whatsoever, all voice recognition software struggle to discern sentences — forget sentences, they even get words wrong.
How far has AI VR come?
Although software like Dragon have come a long way since they started, they’re still nowhere near the accuracy of a human transcriber. To put it in simple terms, unless they come up with artificial intelligence which functions, at least on equal terms to a human brain, a voice recognition software will never, ever come close to delivering the accuracy and quality that a professional transcription service can. You can try it yourself. Get the best voice recognition software in the market today, and have it transcribe your interviews. The time you will spend editing and proofreading the file will be proof enough. First off, you have to train the software A LOT for it to adapt to your accent and writing style. Secondly, most of the voice recognition software available in the market today are not comfortable with a fast speaker, or if there’s even slight background noise. If your speaker is sitting away from the mic, or you’re in a restaurant, good luck!
Any voice recognition software cannot handle background noise, overlapping conversation, or any other disturbance, and that is a fact. It also cannot understand context, which is a massive factor when it comes to discerning what’s being said. An experienced human transcriber has information beforehand from the client. They can understand the context of the interview. They can tap into their already vast general knowledge, or do research to find complex terms, which a voice recognition software simply cannot.
We don’t want to advertise our transcription services. The aim of this post is to help clients save time and money. While technology is good, sometimes it is not as advertised, or the limits of that tech is put in fine print at the very bottom. A lot of people buy these software, and end up realizing that it takes a lot of their effort and time proofreading the document, than they originally thought. That is on top of buying the software, which is not cheap.
Our rates for human transcription
Upload your audio files for transcription
Email for transcription quotes and discounts