Pin microphone recording trouble repair example ① Professional edition
I would like to introduce the sound quality improvement requested by a video production company the other day. The needs of corporate video content are increasing. While a lot of content is being produced, there is no end to audio troubles during recording.
An interview video of a customer who used the service of a certain company. It introduces episodes from the interviewer to the conclusion of the contract, the response from the start of the service, the satisfaction level after delivery, etc. as "cases".
■ Situation at the time of recording when voice trouble occurred
The setting is fixed camera x 1, pin microphone x 2
One interviewer (pin microphone), customer couple (only husband wears pin microphone)
Imagine a material where the audio recording leaks from the customer's microphone for some reason and there is no off-shot audio from a professional camera. The interviewer's voice is the best sound, and the off-mic (voice recorded at a position far from the microphone) voice picked up by the interviewer's microphone is recorded on the same track.
The customer's voice is a distant voice that picks up a lot of echoes in the room. The sound quality is too different for the voice to be spoken on the same screen, which makes me feel uncomfortable. I was asked to repair this audio.
How did you improve
Editing to bring your voice closer to the on-mic is not that difficult. After careful noise processing, you can reduce the reverberation with the audio repair tool, and fine equalizing and compression processing can bring it closer to the on-mic atmosphere.
Off-mic improvement case
The problem that the conversation between three people crosses
As in the improvement example above, processing is easy with a single MC, but conversations involving interviewers always have many voice intersections.
In particular, there were a lot of interviewees saying "I was at a loss about purchasing ~" (yes, yeah, that's right ~), and it was a difficult edit.
Published the procedure of how to edit this difficult problem by voice
Step ① Copy the original audio file for the time being
Judging that it is impossible to trim the material that contains conversations of multiple people with different sound quality. For the time being, I started working in the direction of copying the original audio and mixing the two files.
① File 1 = original voice (use the interviewer's voice)
② File 2 = additional voice (customer voice is boosted)
Step (2) Processing to weaken the interviewer's voice from the additional audio file
To music instructors
Recover recording mistakes from video vendors
Customer's voice Load the file 2 you want to boost into the audio repair software RX-5 and enlarge the spectrogram meter in detail. Since the part where the sound of the audio material is strong is displayed brightly, search for the interviewer's voice phrase in the meter, specialize in that part and reduce the volume with a pinpoint of about -20db.
After -20dB processing, it becomes the meter on the left, and you can see that the color is lighter than the image above. When you watch this, the interviewer's voice becomes weaker than the original voice.
It is a processing enlarged screen of about 200Hz to 1.2KHz where the interviewer's voice is most strongly output, but in reality it is difficult to completely erase it unless the processing is performed over a wider area due to the overtones of the voice. Here, in order to mix with the original voice, we will process it so as not to affect the mixed customer's voice as much as possible. Even with a 5-minute sound source, there is a certain amount of conversation, so we will add detailed processing for each phrase, such as aizuchi. Of course, we also handle noise for the entire material here.
Step ③ Boost processing of customer voice of additional audio file
In order to bring the customer's voice recorded with the off-mic closer to the interviewer's voice recorded with the on-mic, we will apply plug-in processing to the audio data processed in (2). I wanted to use a matching equalizer , so I use a mastering software called Ozone7 to process it.
The sound source that makes it easy to hear the voice is FM radio. Matching EQ processing based on the recording source of the extra edition released on YouTube by a popular DJ. Add sound pressure and sound image processing with a compressor and TRANSIENT MASTER . By compressing the whole voice, the density of the voice increased and it became thicker.
But there is a problem with this process
It was possible to boost the customer's voice, which is a distant sound image, thickly, but due to the deep compression, the reverberation component also rises considerably. This requires cutting the reverberation component of the original material to some extent.
Return to the RX-5 used in (2) and use a plug-in called De-Reverve that reduces the reverberation component. This has a strong effect on the original sound quality, so repeat the process of shaving thinly so as not to apply too much. After finishing this process at the last minute compromise, execute process ③ again.
Step ④ Mix two files with DAW and write volume automation
The original audio is also processed with noise to some extent, and the two files are mixed with a DAW. The original data containing the voice of a beautiful interviewer and the boosted customer's voice are put together in a track, and the volume is controlled according to the conversation. While grasping the flow of conversation, we will record the fader in real time and fine-tune the curve later.
I managed to deliver it with the total EQ and limiter for final adjustment. This type of project is often requested in a situation where it is difficult to meet the delivery date due to the situation where it is not possible to retake the picture. To make it completely natural, we had to do some confirmation work to find a compromise, but we delivered it in one shot.
The video content client will be the final judge by comparing the degree of perfection with the uncorrected data.
Even if it is delivered in a limited time, it may not be adopted, but I want to finish it as a professional. After a while, I will try a different approach for the non-adopted materials, and research on sound reformers will continue (laughs).
For audio problems with professional video production , please contact HybridSoundReform.com . HybridSoundReform.com is a service that specializes in improving sound quality. We are happy to accept recovery projects such as recording mistakes that cannot be retaken. Emergency response is possible. Since the trial is free, you can try the improvement projects that seem impossible without risk.