[AI transcription] 6 points of recording for high-precision transcription
April 28, 2023
In recent years, with the development of AI technology, the number of transcription services using speech recognition engines is increasing. Cheaper with AI! Efficiently! You can now transcribe.
In order to perform highly accurate AI transcription, it is important to incorporate several techniques when recording.
In this article, we will introduce the best recording points when transcribing with AI.
6 points for optimal recording when transcribing with AI
- high quality microphone
- the mic is in the right place
- Create a quiet recording environment
- Clarify the speaker's utterance
- speak one by one
- do a recording test
1. Use a good quality microphone
To improve the accuracy of the speech recognition engine, first use a high-quality microphone.
You can also record audio with the built-in microphone of your PC or smartphone, but it cannot be said to be of high quality.
A high-quality microphone keeps your voice clear and minimizes noise.
This makes it easier for AI to recognize speech accurately.
2. the mic is in the right place
Microphone placement is also important.
If the microphone is too far from the speaker, the voice will be small and the recognition accuracy will be low when transcribed by AI.
Conversely, if the microphone is too close to the speaker, the speech will become saturated, which will adversely affect recognition accuracy when transcribed by AI.
Keep your distance and aim your mic accurately.
3. Create a quiet recording environment
A quiet recording environment reduces noise and improves the accuracy of speech recognition when transcribed by AI.
Even so, if you use a square, narrow meeting room , your voice will echo like a bathroom. If you record in that environment, the recorded audio will be full of echoes and howling and not very audible.
In addition, the sound of air conditioners and fans also becomes noise. Try to keep it away from the mic.
Also, installing sound absorbing materials in the room can further improve the recording environment.
4. Clarify the speaker's utterance
Speaking slowly and clearly makes it easier for AI to recognize your voice.
The speaker speaks slowly and clearly, like a newscaster, and speaks in standard language for better transcription.
5. speak one by one
When multiple speakers speak at the same time, it becomes difficult for the speech recognition engine to recognize them accurately.
If you are thinking of transcribing with AI, try to speak one by one.
6. do a recording test
If you hit anything and do it in production, you may fail.
The same is true for recording. Before you start recording, be sure to do a test recording.
By performing a test recording, you can check the volume and sound quality, and make adjustments if there are any problems.
This reduces the risk of errors in your production recording.
Examples of audio that are actually difficult to transcribe
This is an example for Japanese.
I hope that the atmosphere such as the noise level and the way the microphone is inserted can be a reference for the difficulty of AI transcription of voice.
*As an example, there is no relationship between the distributor and the content of the video, etc., and this service.
Audio files that AI transcription is not good at
- The voice is low, the microphone is far away and it is difficult to hear
- Loud noise and drowned out audio
- Audio is blurry due to room reflections
- Items containing music, such as song lyrics
- without sound
- tight dialect
- something that is difficult for people to hear
When AI couldn't transcribe well
Appropriate recording techniques are essential for highly accurate transcription by AI (speech recognition engine).
Please keep the points introduced this time in check and use AI technology to efficiently transcribe.
However, for audio files that do not transcribe well with AI transcription, it is more reliable to ask for human transcription.
But isn't it expensive to ask someone to transcribe it?
If you are interested, please read this article.
AI transcription also has pros and cons.
For audio files that are good at AI transcription, use "AI transcription".
unsatisfactory audio files
Try transcription with AI transcription → If it doesn't work, we recommend transcription with "human transcription".
AI transcription is cheap, so even if it seems that AI transcription is not good at it, you may be able to transcribe it well if you try it with the intention of "It's not good, but it's lucky if possible."
Mr. Transcription of AI transcription can transcribe the first minute for free. You can check the accuracy of the transcription. Please try it once.
■ AI transcription service "Mr. Transscription"
"Mr. Transcription" is an online transcription tool that can be used from zero initial cost and 1,000 yen per month (* free version available).
- Supports more than 20 file formats such as audio, video, and images
- Can be used from both PC and smartphone
- Supports about 30 languages including Japanese, English, and Chinese
- Supports technical terms such as medical care, IT, and long-term care
- There is a customized dictionary function
To use it, just upload the audio file from the site. Transcription text is available in seconds to tens of minutes.
You can use it for free if you transcribe it for up to 10 minutes, so please try it once.
Transcription for audio / video / image transcription. It is a transcription service that anyone can use for free without installation.
- What is Mr. Transcription?
- Transcript images, sounds, and videos with Mr. Transcription
- Free registration
- Rate plan
- List of articles