A video production company brought us a short but genuinely difficult job: a few tens of seconds of audio that took several rounds of rework to get right, because of how strict the delivery requirements were. Two voices, almost equally loud The footage was a reporter covering the atmosphere at an event venue. While she was talking, a venue announcer's PA announcement — guiding attendees elsewhere in the hall — was bleeding into her mic at nearly the same volume, roughly a 10:8 ratio against the reporter's own voice. Because the announcer's PA speaker was further from the mic, that sound also carried a bit of room echo. The goal was to keep the hall's live atmosphere while getting the announcer's voice down to nearly nothing — without touching the reporter's own dialogue. Why it couldn't be re-recorded The camera operator had been rolling audio but likely wasn't monitoring it closely, and this take had been approved as usable without anyone noticing how much the announcement was competing with the reporter. There was no clean ambience recording to swap in either — every quiet moment of the take was completely filled by the announcer's voice. It also wasn't something that could be reshot: a live event, gone. Where the real difficulty was Working primarily in iZotope RX Advanced, the actual noise-identification and separation work — distinguishing the announcer's voice from the reporter's — took the overwhelming majority of the time on this job. Both speakers were women, so their voices sat in a very similar frequency range, which made it easy for any aggressive processing to damage the exact voice we needed to keep. The first pass we delivered wasn't quite there for the client, but given how much they valued the technical result, they extended the budget and timeline, and a second pass was accepted. It's rare for a job this short to take this much time — but when the budget and schedule allow for it, we can take on jobs like this that most editors would have to turn away. If you're dealing with something that looks unsalvageable — overlapping dialogue, a background announcement that shouldn't be there, audio nobody can go back and re-record — the initial consultation is free. Get in touch about your project

We recently helped fix an audio problem on an e-learning video shoot — the kind of issue that's genuinely hard to catch in the moment. Noise that only shows up after the fact The client records instructors for training videos using a wireless lavalier mic. The audio had already been through standard noise cleanup, but a separate issue only became clear during final review: an intermittent crackling noise scattered through the recording. The instructor had medium-length hair, and as she moved, strands were brushing directly against the lavalier mic capsule. Because a lavalier is a highly sensitive condenser mic sitting right against the body, even light contact produces a surprisingly loud noise — something we also see with clothing rustling against a hidden mic on narrative shoots, where an actor's movement brushes the mic under their costume. Why a standard noise suppressor doesn't work here Noise that comes from direct contact with the mic capsule behaves differently from ambient background noise, and a typical noise suppressor built for room tone or hiss won't touch it. For this, we used iZotope RX's De-rustle module — a tool built specifically for clothing and mic-contact rustle, and one that gets regular use in film and TV audio post-production. De-rustle can remove a lot of this kind of noise, but rustle noise tends to spread across a wide frequency range, so pushing the removal too hard damages the voice underneath it and leaves its own kind of artifact. Getting this right meant judging how much to remove versus how much natural high end to restore afterward — which we handled with careful EQ work and some tape saturation to bring back a sense of naturalness once the noise was gone. One plugin is rarely the whole fix Problem audio like this almost never comes down to a single plugin solving everything. Even after years of doing this professionally, it's often hard to know exactly how much a piece of damaged audio can be improved until we actually get into the material — which is why we always review your file first, even at the quote stage, before committing to a price. If you're dealing with a similar issue on a lavalier or hidden mic — for e-learning, documentary, or narrative work — feel free to reach out. We take inquiries from outside Japan as well. Get in touch about your project

We recently handled the audio for a promotional video shot in Japan for an overseas brand launch — a YouTube piece built around interviews with a well-known actress, shot on multiple cameras by a single videographer. A denoiser plugin, pushed too hard Just before release, the actress's talent office rejected the video on audio quality grounds. The cause: it had been raining lightly but continuously during the interview, and rain noise had leaked into the recording. Based on the final result, it looked like whoever edited the audio had leaned hard on the denoiser plugin built into their editing software — a single denoiser can only clean up so much before dialogue starts sounding processed and unnatural, and that's exactly what had happened here. This isn't unusual. As more editors take on audio duties they weren't necessarily trained for, projects increasingly hit the limits of what a single built-in plugin can handle — especially continuous background noise like rain, which is much harder to remove cleanly than a single click or hum. No room for a reshoot With a well-known actress involved, a reshoot wasn't a realistic option for the client, and the deadline was close. We reworked the audio using a more careful, source-specific process rather than relying on one plugin's preset, and delivered dialogue that held up at broadcast/release quality without the processed, artificial character of the original edit. If you're facing a similar problem — audio that got worse, not better, after running it through a denoiser — we're happy to take a look, whether you're a professional production or an independent filmmaker. Get in touch about your project

I recently took on a request to restore a cassette tape recording — something we do a few times a year, much less often than when this service started. Note upfront: for cassette tapes, VHS, DAT, and other physical media, we can only work from an already-digitized file. We don't accept physical tapes by mail, especially from clients outside Japan — so if you have a box of old cassettes, the first step is digitizing them locally, and from there we can take over. The source: a singer's only surviving recording of her own voice The client was a woman who'd spent her twenties singing at small live venues around Roppongi and Shinjuku. She has a completely different career now, and the only proof that period ever happened is a cassette recording from about 30–35 years ago — a mix of cover songs recorded live. She wanted it cleaned up enough to put on a CD. Hiss noise, solved by decoding Dolby B correctly The tape itself was an ordinary cassette, and it had the hiss noise that's familiar to anyone who grew up with cassette decks. Most consumer decks of that era supported Dolby B noise reduction, which cuts hiss when engaged during recording — the trick is that you need to decode it the same way on playback. There was no label confirming Dolby had been used, so I tried a few settings on a deck with switchable Dolby B/C/S. Dolby B turned out to be the right call — the hiss cleaned up dramatically once decoded correctly, and from there the audio was captured digitally through an audio interface. The bigger problem: instrument balance, not noise Once the tape hiss was under control, the real issue was how the original recording had been mixed. It sounded like vocals, backing track, and a drum machine had all been summed together at the venue's PA mixer, and the line-out from that mixer was what ended up on tape — vocals and drum machine loud, piano barely audible (there was no mic on the piano itself, just some bleed picked up off-mic from the room), and light but noticeable audience noise underneath. This kind of recording — a raw PA line-out with no real mix — was extremely common for live-venue tapes from the 80s and 90s, and a lot of those recordings are effectively unlistenable as a result. With modern AI-assisted source separation, it's possible to rebalance the individual elements after the fact, even from a rough two-track recording like this. Since the budget allowed for more detailed work, we also cleaned up crowd noise and a few stray comments from the audience. After the detailed noise and balance work, the final step was mastering to bring the sound up closer to a modern release — delivered back to the client as a CD. Beyond music: archival audio for documentary and family history projects This kind of work isn't limited to musicians. Old interview tapes, oral history recordings, and archival footage audio all run into the same problems — tape hiss, uneven levels, and decades-old recording gear. If you're restoring archival audio for a documentary or a personal project, we can help once the source is digitized. Get in touch about your project

We've offered noise removal for music performance recordings since early on, alongside our film and video work. Shot from the audience, not the stage A parent asked us to clean up the audio from a video of their child's piano recital, shot from the audience on a personal camera rather than by a hired video crew — which is common at smaller recital venues. Recitals with a lot of young children tend to be noisy audiences, and the recording had picked up a fair amount of rustling and audience noise that made it hard to focus on the performance itself. A camera's built-in mic is physically closer to the audience around it than to the stage, so in a recording like this, it often captures more of the room than the performance — fine for a casual keepsake, but frustrating if you actually want to listen to the playing. Removing a human voice is harder than removing a click Modern restoration tools can remove short, instantaneous sounds — a camera shutter, a sudden clack — almost perfectly, because the software can isolate and resynthesize around them cleanly. A continuous human voice sitting in the same frequency range as the performance is a different problem entirely, and doing it without artifacts takes real editing judgment, not just a preset. This recording had a child's voice clearly audible in about three separate spots over a five-minute performance. Piano is particularly unforgiving here — unlike a full ensemble, there's nothing else to mask small resynthesis artifacts, so even a slightly heavy-handed fix can end up sounding worse than the original noise. Because of that, we're selective about which piano recordings we take on, and heavily damaged material may cost more to fix than it would for a louder ensemble recording. If you have a recital, performance, or rehearsal recording with audience or environmental noise you'd like cleaned up, we're happy to take a look. Get in touch about your project

We recently worked on an overseas documentary project: a long-running record of children's activities in New York, built around the topic of ocean environmental issues and how education shapes kids' awareness over time. Much of it was shot inside an actual elementary school classroom — and anyone who's spent time around a room full of kids on a break can imagine how loud that gets. One-camera interviews, a documentary staple Documentary audio is defined by its realism — the ambience of the room is part of the story. But when an interview subject is speaking in a noisy environment, dialogue often ends up harder to follow than expected, and once it's played back at cinema playback levels, background noise that was barely noticeable on set becomes distracting. That's where proper dialogue editing comes in. The director's first request was focused: clean up dialogue on roughly 30 of the film's most difficult cuts. The audio problems on those cuts included things like: Distracting rattle and rustling noise behind dialogue during a break Clipping caused by a child suddenly raising their voice A lecture recorded with a room mic in a large classroom, with heavy reverberation making the dialogue hard to follow Wind noise on the mic during comments recorded at the beach Background phone conversation bleeding into a recorded lecture near an entrance The director was happy with the results, and the project grew from there — eventually covering more than 70% of the film's cuts, plus cleanup on studio-recorded narration. Getting festival-caliber sound editing without New York studio rates The director is a Japanese filmmaker based in New York. Outsourcing full-scale audio post-production locally usually means going through an established NYC studio — and with New York's real estate costs baked into their rates, that's one of the more expensive markets in the world for this kind of work. Working with a tight production budget, the director found us while searching for a more cost-effective option. After delivery, the film went on to final mixing at a high-end New York studio — and the engineers there reportedly had good things to say about the dialogue editing work. If the film gets a release and we're credited, we'll follow up here with more details. We see similar situations from independent films and documentaries based in Japan, too — tight budgets are the norm, not the exception, and we've built our workflow around that reality. Beyond narrative and documentary projects, we also handle audio restoration on a wide range of damaged recordings day to day, from field interviews to consumer voice recorders, so heavily degraded audio is well within what we can help with. Feel free to reach out if you'd like to talk through a project. See pricing for independent film sound editing

Our sound design work on an independent Japanese comedy earned a Best Sound Design award at the Madrid IFF 2023 (Film Fest International), an indie film festival based in Madrid, Spain. The director attended the ceremony in person, and the win is listed on the festival's official winners page (linked below). This article walks through what the film needed on the audio side, and what that means for other independent directors — especially documentary filmmakers — considering audio post-production for a festival submission. The film: A Recommendation for Euthanasia A Recommendation for Euthanasia is a comedy feature and the director's debut film, following a man who joins a fictional "three-day euthanasia tour" in Okinawa. It was shot without a dedicated production sound mixer on set, which is a common reality for first-time independent directors working with a small crew — and exactly the kind of situation where audio post-production ends up carrying more weight than usual. What the sound work actually involved Most of the film's audio was recorded with a mix of lavalier mics and shotgun microphones, and nearly every piece of dialogue needed some degree of cleanup and rebalancing before it was usable. With a small-crew, run-and-gun shoot, that's expected — the goal in post is to make the inconsistencies disappear. The work broke down into three main areas: Dialogue editing and cleanup: Because the film relied on lavalier mics and shotgun microphones interchangeably depending on the scene, matching tone and presence across cuts took repeated passes, not a single blanket process. Ambience and location sound: Much of the film was shot on location outdoors, which meant every scene carried its own layer of ambience — traffic, wind, room tone — that had to be managed so it supported the scene instead of fighting the dialogue. Mixing: Balancing dialogue, ambience, and music across a feature-length comedy, with the pacing and tone of a comedy in mind rather than a drama. Why this matters for other independent filmmakers A lot of independent films — fiction and documentary alike — are shot the same way this one was: a small crew, limited or no dedicated sound recording on set, and audio that gets fixed in post rather than locked during production. That's not a flaw in the filmmaking; it's simply the reality of working with a limited budget and crew. What changes the outcome is how the audio post-production is handled afterward. Festival programmers and audiences notice unclear dialogue and inconsistent sound far more than most other technical issues — and it's one of the more common reasons submissions get passed over, even when the story and picture are strong. Sound editing and mixing for your own project We work with independent directors — including documentary filmmakers — preparing films for festival submission, handling dialogue editing, ambience cleanup, and final mixing for feature and short films. See pricing for independent film sound editing Source: Madrid IFF 2023 Winners, Film Fest International

We operate a specialized service for improving sound quality, not voice analysis. There are many requests for noise removal from the general public, but we have a consultation about "voice analysis" about once a month. This is not basically handled. The only most recent voice analysis The only case I received most recently was "I want you to check whether the conversation content of the voice data has been edited." While checking the audio material, I checked for any editing scars in the areas of concern. It is relatively easy to judge that the one edited by an amateur has a blatant joint. However, when edited by a professional voice specialist, the seams and background noise are completely composited, so even I, a professional, can hardly tell. I couldn't say that the sound source I received had a clear edit. The client is somewhat reluctant to say "I didn't have it" after receiving a guarantee for the need to "find an edit point". For that reason, we basically do not accept it. There are a certain number of consultations from people with hallucinations, but we do not accept them. I have hallucinations due to mental illness or similar symptoms, and I "always hear bad words", so I recorded it on the spot, but there are a certain number of requests to check if it is included, but I have received it. not. It seems that the faint sound in the room may sound like that. Of course, we do not deny the disease or those who are suspected of it. In the case of consultation using the above as an example, in most cases you cannot hear anything even if you actually check the data, increase the volume by several tens of times and cut the noise. (Although we basically handle data confirmation at the time of contact) We cannot accept work that has no room for improvement. Because we are launching a free trial as a business, if we cannot improve 100%, we will have a deficit for the work. It is the same as it is impossible to stretch low resolution photo data For voice restoration, iZotope's RX8 Advanced (2020 latest version) application is mainly used for business. I'm hitting. Since the sound component called the spectrogram meter (the image at the top of the blog) can be graphicized, you can roughly grasp what kind of component is mixed in the state of the sound. However, very small sounds that can be heard or cannot be heard are not converted into data in low-resolution sound sources. Also, because environmental sounds are a mixture of various sounds, in most cases it is not possible to find a specific component in the meter. Sound has similar characteristics to image data. Even if you stretch a low-resolution photographic image on your computer, it will look like a mosaic, and you may have some experience that you don't understand. Sound quality improvement is basically to make the sounds that are heard easier to hear. If you can't hear the target voice even faintly in the audio data, the improvement probability is zero. If you can't see a word in your own recorded voice that tells you what you're saying, even if it's fragmentary, you probably can't clarify the content. At best, a smartphone voice memo or an inexpensive IC recorder can record about 50 cm from the microphone clearly. Attempts to capture evidence of an affair in one corner of the room often fail. Of course, with a recorder (linear PCM recorder) with a high resolution and a built-in high-performance microphone, the improvement probability will increase a little more. I can hear it by myself, but if you want to find out if you can hear it, I recommend that you first ask your family and other close relatives to listen to the recorded data, saying, "I think I can hear something." Summary Voice analysis is basically not accepted If you can't hear the conversation at all when you ask yourself, the improvement probability is zero. Services that improve the sounds you hear Making a small sound louder is similar to stretching coarse photo data We are improving more than 100 hard-to-listen sound sources a year. Please contact us if you have any problems with the sound source.

Three principles to keep in mind when recording conversations with a voice recorder Recording with a voice recorder is very familiar as voice evidence for each business negotiation or dispute, as well as the minutes of a meeting. However, even though I recorded it, I often hear it when I play it back later, the noise is terrible, and there is a lot of reverberation in the room, so I can't hear the contents. The following three points are the main principles to avoid such troubles. 1. Avoid noisy places 2. The distance from the microphone to the mouth is as close as possible 3. Do not put it in your clothes pocket 1.Avoid noisy places A quieter place is recommended when recording conversations. We do not recommend restaurants, cafes, McDonald's, shopping malls, etc. Even in places where you don't mind the sound during normal times, the recordings are different. When I actually record it, there are a lot of crowds around me, such as other people's voices, in-store broadcasts, BGM, footsteps, etc. You can not. In a small cafe, the conversation in the next seat is clearly recorded, so even if you try to translate it later, you can't use it. In fact, many such improvement requests have been sent to this service. In addition, conversations in the car while driving are drowned out by strong engine noise and wind noise, and the probability of being unable to repair is high. ２.The distance from the microphone to the mouth is as close as possible Reduce the distance of the voice recorder microphone from the conversation you want to record. Avoid putting small voice recorders in your clothes pocket. The sound is also loud, and the sound of rubbing against the clothes is heard, and I can't hear the contents after all. This noise is difficult to completely repair because the microphone and clothes are at zero distance. The most trouble to avoid when recording with a voice recorder, which drowns out important conversations. If you can get the consent of the person you are talking to, place it on a table at a distance from each other. If that is difficult, put it in a second bag that you don't mind even on the table. Of course, if you cover it with a thick cloth, you will not be able to record well, so we recommend that you have a mesh pocket. Of course, if you want to use the other party's voice as evidence, don't forget to point the tip of the microphone toward the other party. Even if there is a physical distance to the performer who announces the content you want to record, such as a conference or lecture, install the voice recorder near the speaker or closer to the performer. In a large venue, the echoes in the venue are strong, and even if you can record beautifully, there are cases where you cannot hear it at all due to the echoes. ３.Prepare high-performance products IC recorders for conversation recording can be purchased for around $ 30 to 80, but if you check the functions carefully, there may be a noise cut function, or there may be a large spec difference even though there is not much price difference. In most cases, old-fashioned voice recorders are inconvenient because they cannot be connected to a computer to transfer data. Even if you own an old model, you should introduce a new model if you are planning to use it for trials or to transcribe it reliably. Summary How was that. I run a sound quality improvement service called Hybrid Sound Reform, but there are many inquiries about "I can't hear the recorded content of voice memos." In some cases, it can be improved, but their success rate is around 50%. There are many sound sources that even professional professionals cannot fix. I wrote this article for those who have the opportunity to record important conversations. #Voicerecorder #ICrecorder #Conversationrecording #Minutes #Voicememo #Recording #Noise #Trial #Divorcetalks

Cuts noisy noise (bird calls) mixed in video messages Nowadays, you can easily shoot videos with your smartphone, etc. Video messages are one of the familiar means. I think many people used it to celebrate their family and friends. The other day, the sound quality improvement requested by a professional creator who is editing videos for free is a video message sent at a company commemorative event. A big celebrity that everyone knows sent a congratulatory comment to the president of a company. "Hello Bob, congratulations this time" It's a comment cut that seems to be common, but most of the time I'm talking, the noisy bird's cry of "peepy" is mixed in so that the comment is covered. Even when recording outdoors in a quiet place, only the birdsong can be heard at a jarring level. For the general public, it is enough to convey the content of the message, but it is a level that is "worrisome" for public announcement such as screening on the company homepage or party venue. It would be a rude video for both the company and the people who commented. ・ Noisy noise mixed in MC can be significantly reduced by advanced technology. ・ Business apps are expensive and require experience to master them. ・ Background noise can also be resynthesized Noisy noise can be significantly reduced with advanced technology Even if the audio is given up 10 years ago and could not be repaired, the repair tools used in advanced post production (such as RX from iZotope) can repair troubled audio in such cases with high accuracy. Sounds that are not mixed with the surrounding sounds and that are clearly uncomfortable can be processed fairly cleanly. The above video removes the whistling of the live venue called "Pee". In the case of comments, it is not an easy task because it slightly covers the main sound, but it can be removed with considerable accuracy. ★ Business apps are expensive and require experience to master I use iZotope's RX8 Advanced (latest version as of 2021), but specialized audio repair tools are a bit expensive for the general public who don't use them often. The orange spectrum meter, which is also displayed in the above video, visualizes the sound, so you can perform various processing locally while watching the visual. A certain amount of experience is required to effectively use it, such as what kind of module should be used for improvement in a high-precision monitor (speaker or headphone) environment or various processes. ★Background noise can also be resynthesized Since the audio component is visualized, it can be processed like Photoshop, but the noise you want to remove (here, the voice of a bird) is identified on the meter and erased. We will try and error the most effective processing, such as locally reducing the volume of the part that can be judged as "bird's voice" or blending it with other sounds in a gradation manner, and search for a better processing method. However, if you erase the "clear sound" of the already recorded voice data, you will inevitably get a "twitching feeling". It's difficult to get a natural finish, but Ambience Match is a powerful help in alleviating them. It is a function to resynthesize the noise part that has been clearly removed with other background noise. If used properly, you can expect a "noise-free" finish. Please feel free to contact us if you have any video or audio content you are worried about. Hybrid Sound Reform helps hundreds of professional creators' troubles in such voice recording every year. This service is operated in Japan. Due to long-term deflation, prices are very low, and it is quite cost effective from orders from overseas. You can try the content you are worried about in a "free trial". Please feel free to contact us. Anyone can remove noise from video data, adjust sound, and restore audio. #Videophotographer #Voicetrouble #Noiseremovalmethod #Audiorepair #Videoaudioediting #Interviewvideo #Greenhouseimprovement

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.

Voice extraction of voice memos is difficult with compressed sound sources (mp3) with poor data density I'm Yamakawa, a sound reformer who may be dealing with poor quality voice data in Japan. Since the sound quality improvement site I run has a free trial system, more than 100 low-quality sound sources that say "that kind of sound is impossible !!" are collected annually. Among them, there are many worries about wanting to extract the voice of voice memos recorded with an IC recorder or smartphone, but to be clear, the probability of being adopted from a trial is about half. The probability of improvement without problems is quite low. What kind of recording is difficult to improve? In the first place, it is far from the sound source (I can only hear the sound of mosquitoes) Conversations other than the people themselves are involved in izakaya and noisy cafes. Make the faint voice of the other party that you can hear from the handset on the telephone. The microphone and the amount of data determine the recording quality Audio recording and video audio are digital, so they are digitized and recorded. That's why the microphone and recording resolution are important. Amateur recordings do not adhere to the basic principles. Only people who are recording without checking the recorded sound. Few people use earphones or headphones to check the sound on a monitor to see what kind of picture they are taking with a digital camera. The only way to extract voice from noisy conversation data is to amplify or resynthesize the remaining components of the digitized data. If you want to make a good sound later, the resolution of audio material with a large amount of data is the key. Imagine a digital camera image with a rough image. It's easy to imagine that a photo that looks blurry and looks like a mosaic is unlikely to be fixed. Voice data is basically the same. IC recorder long-time recording (MP3) MP3 files can be recorded for a long time because the amount of data can be saved. Many IC recorders can switch the recording method. (Wav and mp3) mp3 has a capacity of about 1/5 that of wav, so it is convenient for long-time recording. However, a large amount of data is being decimated. It's okay to compress professional quality sound sources such as artist music data and listen to them in MP3, but it's not suitable for amateurs who want to record with high sound quality. In noisy environments, such as where the recorder's microphone is far from conversation or noisy, it is difficult to technically repair it later. Hidden recording to suppress cheating scene In many cases, one of the spouses reminds us of an IC recorder (smartphone) in a room or the like in an attempt to suppress evidence of an affair. In fact, this usage is wrong. Although it differs depending on the model, the IC recorders that are sold at low prices have a proper recording range of about 1 meter over the desk, such as for interviews. Even if it is hidden in the corner of the room, it cannot record decent audio. If you need to record from a few meters away and listen to the details, you need a higher-grade recorder equipped with a high-performance microphone called a linear PCM recorder. Conclusion ・ Check with headphones what kind of sound can be recorded before recording if it is an important voice that you want to keep as evidence. ・ If you want to leave the possibility of repairing later, record with wav method. ・ If you want to record a conversation with a voice memo on an IC recorder or smartphone, it is within 1 meter from the microphone.