top of page
  • Writer's pictureSound Reformer / yamakaWA

Remove noise from video-recorded piano recitals

Remove noise from video-recorded piano recitals

I run a "audio restoration" service for videos and various recordings, but from the beginning of the service's launch, I have been providing noise removal for music performances.


The other day, I received an order to remove noise mixed in with the audio of a video of a children's piano recital. Recently, the classroom that hosts the recital sometimes asks a video company to record the video, but the video of the client who requested is the material that was shot by himself in the audience. The bleachers tend to be noisy at the recitals of piano classrooms, where the proportion of young children is high. Various noises such as noises and voices in the audience were mixed in the materials we received, and it was a video of the voice that felt stressful to listen carefully to the performance itself.

Shooting in the audience will pick up ambient noise rather than playing

If it's a video that is positioned like a snapshot, that's enough. However, it is difficult to enjoy the recorded music itself. Even if the performance is wonderful. If the voice of the audience is clearly mixed, the quality will almost always be "withered". The built-in microphone of the video camera tends to be in such a state because the sound of the audience seats is located closer to the stage where the performance is being performed. This is stressful for those who want to listen to music properly. When the general public shoots a movie, there is a high probability that such sound quality will be obtained.

Cutting human voice requires advanced skills

The current sound quality correction technology can cut various noises from the recorded data, but since the shutter sound and striking sound of the camera are instantaneous sound images, they are cut almost perfectly (accurately, the noise component is discriminated and others). It can be resynthesized by comparing it with the sound of (), but since the human voice that makes a continuous sound is mixed with various sounds that can be heard in the same range, it requires advanced editing skills to erase it cleanly. Will be.

During the performance of about 5 minutes, the client's sound source clearly contained the voice of the child in the audience seats in about 3 places, and he wanted to reduce that voice and loud noise and convert it to a beautiful performance sound source. .. In particular, piano performances are very delicate, unlike large-scale ensemble performances. If even a small amount of resynthesis marks remain, it often feels more uncomfortable than unprocessed audio, and it may not be possible to improve it well depending on the mixing position during performance.

It is difficult to deceive the performance of a non-ensemble instrument alone

In loud music such as brass band, orchestra, and rock band, small sounds are masked and small noises are drowned out. However, when playing the piano, the sound of turning over the pamphlet can be heard clearly, and the direct noise is more noticeable than the ensemble. Therefore, in many cases, we refuse to remove noise from piano performances. In addition, quotes tend to be soaring (if the noise is terrible) and may not be available from the free trial.

We also handle many sound sources for various instrumental music that are recorded only with a ceiling-mounted microphone. Please feel free to contact us if you have any problems with the noise of the precious performance sound source that you really want to improve. We provide Japanese quality polite voice processing with the best cost performance.


bottom of page