Choosing the best format for compressing audio data: MP3, AAC or WavPack?

If not lossless, then a cat? MP3, AAC - what else? We have already studied music compression algorithms several times before; it’s time to compare the most worthy ones.

In a good way, we should renounce lossy codecs altogether, but it is always interesting to draw the line where quantity turns into quality. In addition, even a lossy codec can surprise you with something, you’ll see.

In this review, it was decided not to fool around with different VBR modes, but to immediately focus on the maximum bitrate with a constant value of 320 kb/s. Today, with the modern capacity of a portable device, should you beg for an extra 10 MB for the capacity of an album with the risk of losing quality? For what? In general, even with ancient codecs, a 320 kb/s stream ensures the absence of characteristic artifacts with nasty jingles. The first part of the review will be devoted to comparing the growth of artifacts using RMAA software, the second will outline the subjective experience of the listener using real phonograms.

Comparative frequency response of three lossy formats relative to the original WAV

If last time the iPad Mini was used as a sound source, now, to improve accuracy, we are taking any influence of hardware out of the equation, and further all distortion analysis will be carried out exclusively in the digital domain, without conversion to analog, since RMAA provides such an opportunity.

To do this, we generate a test sample in WAV in RMAA, then we alternately drive it into various lossy codecs. Next, we convert them back to WAV so that the program can “recognize” the file and evaluate deviations from the original template. Now we see how high frequencies are cut and distortion increases, which gives a nasty color to the sound. By the way, there will not be so many of them. In general, at a bitrate of 320 kb/s, it will not be so easy to catch something clearly harmful by ear. We are not even talking about artifacts, but perhaps some “dullness” of the sound compared to the original. The phonogram seems to be running out of steam, losing its mobility due to disruption of transient processes after psychoacoustic processing. But it is not always possible to clearly record this difference; it depends on the specific track.

Select one or more files on your computer

The service allows you to convert (convert) from the AAC format to the AC3
AAC format - this is an encoded audio file format that was developed by Fraunhofer IIS, AT&T, Dolby and Sony. Its name stands for Advanced Audio Coding. The AAC format is an industry standard and was created as an improved alternative to the most common audio file format, MP3. AAC files do offer slightly improved performance over MP3s. Let's say this format provides the possibility of improved processing of frequencies above 16 kHz. At the same time, the sound quality of AAC files is closer to the original sound sources.

AC3 is an audio file format. Its name is an abbreviation for Audio Codec 3. The format was developed by Dolby Laboratories, Inc as a modern and efficient audio compression technology. AC3 files feature multi-channel Dolby Digital surround sound. Because of this, they are widely used as a format for DVDs, cinemas and game consoles. However, all the benefits of such files can only be appreciated if you have the appropriate acoustics for surround sound. An audio file of the specified format can include up to six sound channels, which creates an immersive effect for the listener.

Advanced Audio Coding (AAC) is a proprietary (patented) audio file format with less quality loss during encoding than MP3 at the same size. [1]

AAC is also a wideband audio coding algorithm that uses two basic coding principles to greatly reduce the amount of data required to transmit high-quality digital audio. This format is one of the highest quality, using lossy compression, supported by most modern equipment, including portable ones.

Dolby Digital (AC-3, ATSC A/52) is a spatial audio system developed by Dolby Labs, led by Ray Dolby, a pioneer in the audio and video industry.

The format is standardized by the Advanced Television Systems Committee, coded A/52, and Dolby Digital (DD) is a trademark.

Shouldn't we broadcast it in mono?:

If your station is purely musical, then you can safely skip this paragraph

About mono from Evgeniy:

For example, I really respect the mp3 / 44100 / 64 kbps mono format. It has no obvious flaws in the ear, it is actually like a “VHF-sounding receiver”, that is, mono, but the sound is adequate. people often forget about it, and this is much better than turning on the stereo on it and an obvious gurgling will begin, or a frequency cutoff in the region of eight kilohertz, and nothing remains of the sound itself in detail. Sometimes the station is semi-conversational, for example. but they set it to stereo, and with a low bitrate it actually has to be divided by 2 (with nuances, but roughly speaking by two). And sometimes it is more profitable to use monophonic audio, in favor of optimizing the sound itself) At the same time, a significant part of users will not even notice that stereo, especially if they turn it on from the speaker of a mobile phone, or simply through the ass of connected acoustics (probably meant bluetooth approx. .author). But a cut off frequency response at this level is always noticeable, and the conclusion is drawn that “the sound is shit.” But it’s not shit, it’s not configured correctly and cuts it all off, again, a lowpass filter automatically? Alas, programs like Edcast don’t seem to know how to cut it off, unless there’s a custom line for console parameters (I don’t remember).

I’m glad that the same RadioBOSS has support for “custom” encoders, and the field for creativity is very huge. I think in the future we will definitely consider this as part of #makingradio

AC3 to AAC – Best AC3 to AAC Converter to Convert AC3 to AAC

I want this movie to be in a small size similar to the ones found on torrent sites. Make sure you are using the latest version of FFmpeg. For Windows and Linux, static builds are available on the main page. To control video quality, set the crf parameter, which defaults to Lower meaning better quality but higher file size. Try values ​​from 19 to 26 to see what works best. You can also set a specific bit rate depending on what file size you want.

If your audio stream uses multiple channels for example audio 5. If your input file has multiple video, audio and subtitle streams, ffmpeg will not convert them all by default.

Use -map 0 to instruct ffmpeg to take all streams from the input file. See the FFmpeg Wiki for more information. This is useful for preserving different languages ​​and subtitles that may be in the original. Please note that mkvdts2ac3. You will see them when you run the script. If you prefer a GUI instead of commands, I've described how to do it: Convert unsupported audio from video.

Why do we need Bluetooth audio codecs?

Bluetooth technology has made a huge career in recent years. As we said above, we can find it in more and more devices that we use every day. It has a lot of advantages. It is very common, inexpensive and easy to use. It can be used to transmit a wide variety of data, including, of course, audio. Defects? From an audiophile's point of view, it is important to emphasize that Bluetooth was not designed to transmit high-quality music.

The key word needed to understand the role of Bluetooth codecs is bandwidth. Technically speaking, this is the maximum speed and amount of data that can be transferred per unit of time. We can think of this as the width of the highway along which our sounds travel. The wider it is, the more information - details about sound, timbre, space - will pass from the transmitter to the receiver.

The simplest, most commonly used Bluetooth codecs do not provide much bandwidth. This means that the signal that ultimately reaches the loudspeaker, and therefore our ears, does not contain much of the information present in the original studio recording. As a result, the sound quality is lower.

As Bluetooth technology has matured and become more widely used in audio devices, many manufacturers have begun using—or developing—more advanced data encoding methods to transmit music.

Although AAC is not as popular as MP3, a large number of devices support this format. This applies to portable file players (e.g. Apple iPod), a group of stationary players/servers, a large group of cell phones, some game consoles, etc. A large number of computer programs for AAC playback are also available. On the other hand, the availability of AAC coding programs is poorer.

AAC includes solutions protected by patents, and the group of licensors consists of a dozen or so organizations. Licensing is handled by Via Licensing, an affiliate of Dolby Laboratories.

Which audio format should I choose?

Password Help Users Calendar All sections have been read. Reputation: For high-quality sound processing, use the tips in the topic Working with Sound. This topic provides some methods for converting audio files from one format to another, which do not claim to be the highest quality methods, but rather intended for personal use. Free Audio Converter aac, ac3, aiff, amr, ape, flac, m4a, m4b, mka, mp3, ogg, ra, tta, wav, wma ; Pazera Free Audio Extractor 1. Last edited by destruktor,

How is it different from other codecs?

Since Sony has its own codec, Apple needs to be as good as possible, so they created the Advanced Audio Codec (AAC). Apart from SBC, this is the only option for users of apple-bitten phones and tablets that - at least for now - do not support aptX or LDAC codecs (otherwise that applies to MacBook laptops). Thus, this is the only way to improve the quality of Bluetooth audio for those who use iPhone as their audio source.

Another advantage is that in the case of AAC, using headphones compatible with the codec and playing music from AAC files (the standard for the iTunes remote program) or streaming through Apple Music, we should receive uncompressed audio.

Is that all? There are many more Bluetooth codecs, but SBC, AAC, LDAC, and of course aptX are the ones worth knowing due to their frequent presence on wireless headphone specification charts. Those who want to achieve the best sound quality through Bluetooth headphones should definitely be interested in the codec compatibility between their devices. It is also worth investing wisely in equipment adapted to the files that we will play through it.

Considering the rapid development of wireless technologies, we can definitely say that the market is not yet saturated with codecs. Therefore, we look forward to updating the article with new proposals that will make listening to music without cable more enjoyable!

Which video format is better? MKV or MP4

Habr Geektimes Toaster My circle Freelansim. StrangeAttractor StrangeAttractor. Audio MP3 Ogg Audio processing mp4. I apologize for this issue that has already been discussed thousands of times, but today I would like to raise it again in the context of modern realities. As far as I understand, the only advantage of MP3 compared to its competitors that appeared later was its wide support - only this format made it possible to be sure that the file would not be a problem to listen to. Now this problem has practically disappeared: any modern smartphone, tablet, TV set-top box, browser, audio player applications, etc. Do I understand correctly that the main advantage of AAC is played on smartphones in hardware and, therefore, requires fewer resources? I also heard that MP3, unlike Vorbis, did not hear about AAC in this vein, it does not allow you to accurately address a certain point in time and accurately cut a piece from and to or something like that, I don’t remember what it’s called. The question is of interest in the context of high bitrates of lossy coding - 96 kbps for audiobooks and video soundtracks, for Internet radio, for offline music.

Instead of a thousand words:

I prepared a table with broadcast formats of popular FM radio stations. Please note that the table shows streams that can only be enabled on the official website. In fact, there may be much more of them. You can just listen to them and form your own opinion.

Direct comparison

Let's try to adjust the settings of both encoders in such a way as to obtain the same output bitrate and frequency cutoff. This will allow you to analyze and compare the quality of encoding algorithms under the same conditions.

Since the FhG AAC settings are limited, we will adjust the FDK AAC parameters. I managed to achieve the same result with the following parameters:

FhG AAC VBR 4 (130 kbps) FDK AAC -m 3 -w 16500 (132 kbps)

However, at this bitrate the encoders coped with the selected composition so well that I was unable to identify any differences even when comparing with the original. However, I noted a slightly denser spectrum in the HF region for FDK AAC.

Testing on the killer sample TyDi - Meet Me In Kyoto gave a subjectively better result for FhG, and this despite the fact that its bitrate on this sample was 18 kbps lower. I noted a similar result for the sample Tiesto - Do You Feel Me. The FhG AAC had noticeably fewer artifacts due to transients. I got the same results for the FDK AAC -m 4 mode; without manually setting the low-pass filter, this increased the bitrate by several kbps, but FhG still gave less distortion.

Packaging format

In fact, for media streaming application scenarios, you don't need to care about the specific encoding data, but rather understand the packaging format. The following table lists more clearly several packaging formats and application scenarios for AAC

The value of several basic fields (structure) in an encapsulation format

1. AOT

The AAC standard defines a number of specifications and also defines a way to identify them, which is the value of AOT (Audio Object Types), as shown in the following table:

The most common AOT values ​​are listed above. It is important to note that this value is a direct identification of the AAC encoding specification. This directly affects the package format design because different specifications require different parameters.

1. AudioSpecificConfig structure

For audio data, it is generally necessary to know some information about the audio, such as the acceptance rate, channel, number of sample points, and length of one frame of the audio data. This information must be recorded in this way, and the AudioSpecificConfig structure is used to do this.

AudioSpecificConfig is defined in MPEG-4. MPEG-2 has no definition. So not in ADIF, but in ADTS.

Below is the block diagram of AudioSpecificConfig

How to write a codec?

Codec

(English codec, from coder/decoder - encoder/decoder - encoder/decoder or compressor/decompressor) - a device or program capable of converting data or a signal.

Interesting materials:

How to download music to your mobile from your computer? How to download music to iPod from computer? How to download the play market to your computer? How to access Instagram from a computer? How to block Internet access on a computer? How to close mail on another computer? How to close your Instagram profile via computer? How to close the Viber program on a computer? How to password protect WhatsApp on a computer? How to record audio from YouTube to computer?

FhG AAC

FhG AAC has a choice of six target VBR quality modes. On average, these modes correspond to approximately the following resulting bitrates: 1 - 32 kbit/s, 2 - 64 kbit/s, 3 - 96 kbit/s, 4 - 128 kbit/s, 5 - 192 kbit/s, 6 - 256 kbit/s With.

The foobar2000 command line in our case looks like this:

—quiet —vbr n —ignorelength — %d

The quality here is affected only by the -vbr parameter, the values ​​of which (n) are from 1 to 6.

I tested the encoders on white noise (10 sec) with a level of -1 dBFS, the Pink Floyd track - Wish You Were Here and the RMAA 24/44.1 test signal. White noise is a completely random signal with power evenly distributed over the entire frequency range. Thus, this signal is the most difficult to encode and the output bitrate is theoretically the maximum possible resulting bitrate. The test signal has a fairly low complexity, so it can act almost as a lower limit on the bitrate. The music track will be something in between, reflecting real coding conditions.

Below I present the mode selected by the encoder, the resulting bitrate (FBR, white noise / track / RMAA signal) and the cutoff frequency of the low-pass filter.

1 - SBR+PS - 45 kbps / 34 / 15 kbps - 6.5 kHz (LC) + SBR = 16.5 kHz 2 - SBR - 88 / 63 / 16 kbps - 9.5 kHz (LC) + SBR = 20.5 kHz 3 - LC - 145 / 98 / 25 kbps - 16 kHz 4 - LC - 193 / 130 / 26 kbps - 16.5 kHz 5 - LC - 294 / 207 / 31 kbps - 19 kHz 6 - LC - 390 / 262 / 36 kbps - 22 kHz

conclusions

Based on the results of my experiments, I can conclude that FDK is not inferior to FhG AAC in terms of capabilities and configuration flexibility, and even surpasses it. But at the same time, it seems that the developers' claim about the advanced quality of the paid encoder is not unfounded, and the commercial FhG AAC from Winamp does use more advanced algorithms than the open source encoder. In addition, FhG AAC is noticeably better in terms of performance: 37x for FDK AAC -m 4 versus 60x for FhG AAC -vbr 4.