I'm Yamakawa, who runs a sound quality improvement service. For the purpose of submitting voice as evidence in proceedings or troubles, transcription of meetings and seminars, etc. There are many consultations from people who record with voice memos and voice recorders on smartphones. Most of them were recorded, but I was worried that I couldn't hear the conversation.
The ambient noise is terrible
The distance to the person who is speaking is far
Difficult to hear due to many echoes
Recording equipment trouble
The above are the main causes of voice problems that make it difficult to judge the words in the recorded data.
Now, why are you playing these recorded audio? If you can't hear the conversation after playing it normally on your smartphone, voice recorder, or computer speaker, we strongly recommend purchasing headphones or earphones (preferably in the $ 80 price range) before using this service. The current situation is that.
Most smartphones, voice recorders, and PC speakers are low-cost, cheap speakers.
Although it depends on the model, most smartphones, voice recorders, and PC speakers that are not designed for sound quality are low-cost cheap speakers. It is a spec that "you can check the sound for the time being".
The playback resolution is very low, and you can listen to music and voice content produced as sales content to some extent, but it is not suitable for playing back noisy amateur recording data.
You can listen to it without any modification
There are many cases where even a sound source in poor condition can be grasped considerably by using headphones and earphones. We have a professional monitor environment, but there are many cases where you can hear it at all without any modification.
Conversations that cannot be heard even with high-quality music earphones or headphones
This is where the sound quality improvement service comes into play. We handle a large number of low-quality sound sources, and at Hybrid Sound Reform, we always keep the sound sources that you are having trouble with in the form of a free trial.
However, even professional professionals do not have a very high success rate in improving voice recorder sound sources.
Especially in conversations where the content is difficult to hear, the adoption rate is around 50%. (Of course, there are cases where it doesn't match the client's budget.) No matter how long it takes and complicated repairs, only a small improvement can be made.
(Reference) Main factors when giving up sound quality improvement (difficulty ranking)
5th place: Completely drowned out by other loud sounds The accuracy of noise removal has improved in recent years. If it is a momentary sound like "Bos, don, Paan, click", it can be removed with considerable accuracy. Apart from them, there is a possibility that you can retrieve the conversation in between. However, the sound of "gaaa --------------" and "Zaaa ---------" that keeps ringing can be reduced.
4th place: The volume difference between the interlocutors such as telephone conversations is too large
Telephone conversations are not recorded with a dedicated app, but if you put the recorder in your mouth, you can of course record only the sound of the other party with a small sound like a mosquito. The volume difference from your own voice is also large, and it takes a tremendous amount of work to improve.
3rd place: Low sound source resolution
Cheap IC recorders take voice memos with your mouth, so low resolution settings may be standard. Even if you record a long distance with such a recorder, the data density is low, and there is a low probability that it can be improved by retrofitting.
2nd place: Various voices are mixed in a complicated way It is not possible to focus on a specific person when other conversations are mixed at a close distance, such as in a bar or a noisy cafe. Since the frequency bands of human voices are approximately the same, there is currently no voice editing technology that can be individually judged to reduce or eliminate pinpoints.
1st place: The distance to the sound source is so long that you can hardly hear it (in short, it is not converted to digital data)
Naturally, in digital recording, the sound picked up from the microphone is recorded as a signal. There is not enough data to improve small sounds that cannot be picked up from the microphone.
It is difficult to record a conversational IC recorder at a level that can be heard without recording with the mouth.
The voice recorder (iPhone voice memo) is only for meetings and memos of your own voice, so if you are more than 1 meter away from the sound source, the probability of clear recording will drop considerably. If you want to record at a certain distance and record the clear contents, it is wise to record with a high-resolution linear PCM recorder with high microphone sensitivity.