The majority of people use a computer on a daily basis. Though, just because you spend hours in front of the computer screen does not necessarily mean that you can type at lightning speed and with a high degree of accuracy.
If you are considering a career as a transcriptionist, the ability to type fast, without compromising the accuracy, is fundamental. You will not be as competitive as your fellow transcribers if you take longer than the industry standard to transcribe an hour of audio.
At the end of the day, if you can transcribe faster, you can also make more money by taking on multiple transcriptions.
To cut a long answer short, according to the industry standard, it takes more or less four hours to transcribe just one hour of audio. In other words, to transcribe 15 minutes of audio, you will need about an hour (even if you work with foot pedals that help to free your hands so that you can type faster).
Though, if the audio recording is not that clear and not loud enough, it can take quite significantly longer. As a matter of fact, it can even take double as long! So, spending an entire standard working day just to transcribe one hour of audio is not unheard of.
That being said, the reverse is also true. If the audio file has been recorded professionally and the speakers speak very slowly and clearly, it can take faster than four hours. A 60-minute high-quality audio file could possibly be transcribed in three hours or even two! This works out to about one hour needed to transcribe 20-30 minutes for an hour of audio.
Taking all of this into consideration, a transcriptionist will then need anything between two hours to about nine hours to transcribe an hour of audio. As transcribing an audio file can be quite a time-consuming task, it will make a lot more sense to opt for the automatic approach as opposed to completing it manually.
Fortunately, there is a couple of transcription software programs available that can cut the work (and time spent) in half. For example, a software solution like Audext makes use of artificial intelligence (AI) and then transcribes the audio recording into text automatically.
Whereas transcribing it manually would take a couple of hours, Audext can complete it in just a couple of minutes. You do not even need to add the names of the speakers after the transcription, as Audext offers you to complete that beforehand and add automatically to your transcription.
In essence, the only thing that is left for a transcriber to do when using Audext is to format and edit the transcription that it has generated by checking the text against the audio recording.
As mentioned earlier, the quality of the audio recording plays a major role in the transcription speed. If there are any background noises, it can distract the transcriptionist from that what needs to be transcribed – the actual dialogue.
The number of speakers can also affect the average transcription speed. If there are more speakers, it becomes more challenging as you first need to identify which speaker is currently speaking.
That is why if you know that there are going to be many speakers, it is good practice to ask the speakers just to say their name out loud before they speak further. Also, if there are more speakers, there is also a greater risk that they will talk over each other which makes the job of the transcription a lot harder.
While still on the topic of speakers, accents can also affect transcription time. If one of the speakers has a very strong regional accent, it will take longer to transcribe. And, if they speak rather quickly too, the transcription task at hand has become even more difficult.
Not only do the speakers affect the transcription time, but also the actual content. If the interview uses a lot of specialized terms and proper nouns, like the names of people and places, it will call for more research which translates to more time.
Lastly, the type of transcription required also plays a role. Specific instructions like true verbatim (when everything, in other words, every laugh, sneeze or pause, needs to be transcribed exactly) and timestamps call for more work which also means more time.
In short, if there are more of these factors that the transcriber needs to consider, the transcription will take longer to complete.
Depending on the quality of the recording, the topic and the number of speakers, it can cost anything between $2-10 per minute of audio recording. In other words, transcribers usually do not charge per hour worked, but instead per audio hour that they need to transcribe.
The majority of professional transcribers will charge less for something such as a podcast. That kind of content usually does not have as many speakers and is recorded with professional recording equipment. Meanwhile, transcribers will charge more for a conference call where speakers tend to interrupt the other participants and often do not finish their sentences before starting a new train of thought.
When it comes to transcription, an hour of audio does not translate to an hour of work as there are a couple of factors that will affect how much time is needed to transcribe an hour of audio.
Factors like the typing speed in words per minute, the transcriptionist’s experience, specific requirements for the transcription and the quality of the audio file can all play a significant role.
Though, even when you have a high-quality 60-minute audio recording with only two speakers discussing an easy-to-understand topic without industry-specific terms, it can take about two to four hours to complete.
Therefore, it is best to make use of automation by getting a transcription software program that can complete the bulk of the work for you automatically in just a couple of minutes.
Katin runs her own podcast show, and interested in best ways to transcribe audio recordings into text.