As digitization transforms industries, the demand for business transcription services is growing. Companies need to transcribe different types of audio and video interactions such as meetings, conference calls, podcasts, interviews and so on, so that the content can be stored, shared or used for reference purposes.
One of the most important challenges that transcription outsourcing companies face is poor audio clarity. Common audio problems transcriptionists have to deal with include:
- Garbled voices
- Speed of speech
- Mispronounced words
- Poor audio due to wrong positioning of the microphone
- Inaudible audio in some parts of the recording
- Overlapping and lagged conversation
- Noisy background due to exterior noises such as traffic, weather, conversation, animals, lawnmowers, keyboard clicks, etc.
- “Plosives” – the explosive sound consonants make when spoken into a microphone
- Unnatural sound effects created by audio
- Echoes from the room where the recording is made
- Slang and heavy accents
- Different languages
- Volume variations
Though transcription is a time-consuming, laborious task, audio which is clear and audible and in the same language is easier to handle, even if there are multiple people in the recording. However, an audio recording may often have multiple issues, making it extremely challenging for the transcriptionist to handle. This could affect the quality and cost of the project.
To deliver quality transcripts, teams in digital transcription agencies use certain strategies to deal with poor quality recordings:
- Using a good quality headset: Transcriptionists use headphones that connect to a computer or audio-playing device via a USB port or a 3.5 mm headphone jack to listen to audio content and convert it into text. Using high-quality headphones is essential for audio clarity. Features of that determine headphone quality include:
-
- the diameter of the headphone driver (the larger the driver’s diameter, the better the sound quality);
- diameter of speaker – for e.g., if it is an over-ear headset, a speaker of at least 40 mm is considered the best choice; quality of the electrical conduction;
- sound sensitivity or the ability of the headphones to detect sound, even at the smallest change/volume;
- electrical resistance
- frequency response
- noise characteristics such as noise isolation and/or noise cancellation
Using top quality headsets go a long way in helping professional transcriptionists handle files with poor audio.
- Slowing down of increase the speed of the dialogue: Transcription software programs allow you to slow down the speed of the playback of the audio. Transcriptionists adjust the speed of the audio when words are difficult to understand. Slowing down the speed of the audio:
-
- can eliminate background noise and help complete the files faster
- avoid the need to continuously rewind the file and listen to it multiple time to understand the difficult audio
- improve turnaround time (TAT), an important goal for most projects
- Help with accents of non-native English speakers or a speaker who has some form of speech impairment
However, in some cases, slowing down the audio playback can lead to spending more time on the file, but reliable transcriptionists work on their typing speed to avoid this problem.
- Using a sound editor to enhance audio: Another strategy to improve sound quality for transcription is to use a sound editor. Professional transcriptionists use a sound editor to recover hard-to-hear dictation. A sound editor can reduce the background noise, eliminate the echo, adjust the audio’s pitch, modify the volume level of the frequencies, and overall, make voices sound better.
- Using noise cancelling software: There are many free software tools that can clean up sound and remove noise. Audacity is a popular, easy-to-use option that is specifically designed to address noise in an audio file. Krisp, another top tool, uses machine learning to identify the voice of the person speaking, attaches to it and removes all other sounds that are not the speaker’s voice.
- Marking inaudible words: Professionals will not try to guess inaudible words in the audio recording. Instead, they will mark words that are hard to comprehend to indicate that that part of the recording was not clear. Usually, they will add timestamps to indicate the part of the recording that lacked clarity for transcription using a specific format, for e.g., “inaudible at 00:15:35”.
- Translators for multiple languages: It can happen that an audio recording has people speaking in different languages. In this case, an online transcription company would employ translators to help their team do their job.
Many factors can combine to make the transcription task complex and add to the time taken to complete the work. If using the strategies listed above cannot help, transcription outsourcing companies will inform the client about it. It’s important that clients know that poor audio and extra transcription time can add to the cost of the project. They can also provide their clients with guidance on how to ensure good audio recordings.
