18 digital audio technologies

    The new digest of the “ Hi-Fi World ” is dedicated to the influence of technology on how we listen to music. From the articles you will learn about new ways to use machine learning, the market for streaming services, as well as a variety of digital audio formats.


    Photo by Jonathan Kriz / CC



    Streaming services




    • The present and the future of musical streaming. Streaming services dominate the music market. In this article we understand their stories. Let us tell you how Spotify became the market leader and what competitors are doing in an attempt to take its place. You will also learn about technologies that allow Spotify recommender algorithms to select music exactly to the user's taste.

    • “For every taste”: streaming services for listening to music. If the first article dealt with the market of music streaming in general, then in this material we understand the features of specific services. From the article you will learn in which application you can listen to playlists from famous musicians, and which one is the easiest to upload your collection of tracks.

    • Explore Hi-Fi audio in Tidal. Tidal service promises subscribers studio quality sound in streaming format. To check whether this is in fact the case, the author of the article compared the tracks on Tidal with their lossless masters. It turned out that not everything is so simple. Read more about the experiment in the material.


    • Unsinkable SoundCloud: from the idea to the 10th anniversary of the project. For the first ten years of its existence - from 2007 to 2017 - SoundCloud failed to become a profitable company. In July 2016, the service was even on the verge of bankruptcy - the company had only money for 80 days of work. At the same time, the number of SoundCloud users was ahead of other music services. We understand the history of the company and talk about how SoundCloud returns to its original task - to help independent musicians.

    • What is it: the Russian market of streaming services. The number of paid subscribers at large Western services is increasing, and new platforms are entering the market, for example YouTube Music. The Russian market is a little behind the foreign market, but in our country streaming is becoming more and more popular. Experts predict that by 2020 the number of Russians who are paid subscribers to streaming services will exceed 10 million. We talk about the innovations of domestic streaming services and Western platforms in the Russian market.



    Machine learning




    • Why retailers "overhear" their employees. Studies show that staff productivity is reduced under supervision. But this does not prevent the retail giants from eavesdropping on their employees. This summer, Walmart, an international hypermarket chain, patented a device that records and analyzes cashiers ’conversations with customers. We will tell you how and why retailers use audio technology at their outlets.

    • “Machine Sound”: synthesizers based on neural networks. Artificial intelligence technologies are used not only in work, but also in creativity. For example, Google has already released several software synthesizers based on AI systems. At the same time, companies such as Flow Machines, Amper and Jukedeck are developing services that generate full-fledged musical compositions. We tell you what other projects use neural networks to help musicians.

    • Audio Search, Audio SEO, and Podcast Help - Google experts are already working on this. Artificial intelligence experts from Google are not only interested in music, but also in podcasts. The company plans to create an audio search function. For example, users will be able to enter the name of the book in the search box and get a podcast with the discussion of this work as a result. In addition, Google will be working on audio SEO technologies and a podcast recommendation service.

    • Word for human speech: what he can do. Before the advent of speech recognition technology, transcribing and editing audio recordings was possible only manually. Now this task is simplified by “audio editors” that allow you to swap speech fragments, add pauses or intonations and synthesize new phrases in the voice of the speaker. This is how Descript and VoCo services work. We talk about the goals of the creators of editors for audio recordings and the ethical side of speech synthesis.

    • Recorded negotiations of the Apollo missions with the NASA Mission Control Center are digitized. Often, neural networks are taught to recognize speech on the arch of specially prepared audio recordings without extraneous noise. When the algorithm is trained on such “simplified” data, it recognizes speech worse in combat “conditions”, for example, in a car or in a public place. Therefore, the researchers decided to use for recording neural networks audio recordings from the NASA archive with thousands of hours of astronaut talks with the control center. However, these records were not in digital form. About how they were digitized - this stuff.


    Photo mlange_b / CC



    Digital formats




    • MP3 VS FLAC. How to hear the difference? Although FLAC music is far superior in quality MP3-audio, this difference is not always felt. You can hear the differences between lossy and lossless formats only on high-quality audio systems. We will tell you how to get the best sound from home equipment and from which portable devices you should listen to music in FLAC format.


    • We measure the number of hidden errors in the CD. Errors in reading the signal from a CD-disk occur even on high-quality media in perfect condition. This may be due to the low sensitivity of the CD player's laser or other factors, such as disc curvature. We describe how to evaluate the quality of records with seemingly perfect CDs.

    • Why do all modern audio recordings sound the same? Modern popular music often sounds the same, and not casual. Labels seek to push their performers on the radio and for this they adjust their recordings to well-established radio standards. We retell the main thing from the material of engineer and journalist Tom Whitwell about how music is recorded in our days.

    • Comparative test of Internet radio stations with Lossless-stream. The increase in Internet connection speed has led to the fact that Internet radio stations can now broadcast lossless audio. This material contains links to online stations broadcasting in FLAC and AAC formats. Among them are both streams with “mainstream” electronic or rock music, as well as more exotic stations: a Norwegian resource with Japanese songs or a Dutch radio with soundtracks from games.

    • How to encode audio files to MP3. MP3 files with the same bit rate will not sound the same - a lot depends on the codec used. In this material, we compare the capabilities of the Lame and Fraunhofer codecs and share the most successful settings that we recommend using when compressing audio files.

    • Why CDs can sound better than vinyl records. Although some music lovers, such as singer and musician Neil Young, consider vinyl to be the best audio carrier, CDs often sound as good as records. Vinyl does not always reproduce high frequencies correctly, and therefore recording on a CD may sound cleaner and closer to the original. Much depends on the taste of the listener, and therefore neither vinyl nor discs can definitely be called the best music format. Let's talk about this in the material.




    Telegram-channel "Audiomanii" - about music and formats acoustics microformat: Our buyer's guide: Bookshelf speakers vs floor Our rating headphone guide for the beginner: it is important to know about the ear cushions headphone How to start a home range of acoustics? How do you use old audio equipment?









    On November 10, in the showroom on Elektrozavodskaya, we will hold a special event, at which Artyom Faermark, the head of Audiommania, Yuri Fomin, design engineer, and invited experts will speak. It will be about modern approaches to the design and production of acoustics, recommendations for the selection of audio systems. You are waiting for reports, demonstrations, answers to questions. There will also be an online broadcast. Start at 12:00 - an additional announcement and other details we will report in our upcoming materials and Telegram-channel Audiomanii .



    Also popular now: