News - The Ultimate DSP Guide: How to Improve Audio Performance？

What's DSP?

In this ultimate guide, we introduced DSPs for audio equipment in detail.

They include concepts, types, working principles, and applications.

You can learn from this guide all the basic concepts of DSP.

Okay, let's start.

What's DSP?

DSP (Digital Signal Processor) is digital signal processing technology.

It works by taking an analog signal and converting it to a digital signal of 0 or 1.

The digital signal is then modified, deleted, enhanced, and the digital data is interpreted back to analog data or the actual environment format.

In audio equipment, DSP is an important component, and it is an important symbol of digital audio.

In fact, it does not change the timbre, but improves the sound effect through tuning, making the music pleasant to listen to.

It's not a product, but a technology that exists to process digital audio signals.

DSP Improves the Sound Quality

A DSP chip is a chip that can realize digital signal processing technology, which is equivalent to a computer CPU. The transmission of sound files is generally compressed, and part of it will be lost.

For example, if a song is recorded in a recording studio, it may be a file of several gigabytes, and it may only be GM to tens of megabytes when it is finally transmitted to the mobile phone.

DSP performs calculations on audio numbers through various algorithms such as transformation, modulation, and programming, and quickly uses algorithms to improve sound quality.

After touching these adjustments, the speaker power of the stereo is higher, the sound is clear, and the listening is more comfortable.

Classification of DSP Chips

1. According to the basic characteristics

a. Static DSP Chip

On any clock frequency within a certain clock frequency range, the DSP chip can work normally.

b. Consistent DSP Chip

There are two or more DSP chips, and their instruction sets and corresponding machine code machine pin structures are compatible with each other.

2. Data format

The DSP chip whose data works in a fixed-point format is called a fixed-point DSP chip. The floating-point format used by different floating-point DSP chips is not exactly the same. Some DSP chips use a custom floating-point format, while some DSP chips use IEEE. Standard floating point format.

3. Use

According to the use of DSP, it can be divided into general-purpose DSP chips and special-purpose DSP chips.

What Are the Advantages of DSP System?

Compared with the analog signal processing system, the digital signal processing system based on the general-purpose DSP chip has the following advantages:

1. High precision, strong anti-interference ability and good stability. The accuracy is only affected by the quantization error, that is, the limited word length, the signal-to-noise ratio is high, and the device performance is small. It is less affected by external factors such as temperature and environment.

2. It is easy to program and implement complex algorithms (including adaptive algorithms). DSP chips provide a high-speed computing platform that enables complex signal processing.

3. Programmable, when the function and performance of the system change, there is no need to redesign, assemble and debug. Such as realizing different digital filters (low pass, high pass, band pass); radio communication in different working modes in software radio; filters and spectrum analyzers in virtual instruments, etc.

4. The interface is simple, the electrical characteristics of the system are simple, and the data flow adopts a standard protocol.

5. Easy integration.

6. It can realize functions that cannot be realized by analog processing: linear phase, multi-sample rate processing, cascading, easy storage, etc.;

7. It can be used for very low frequency signals.

The emergence of digital signal processors makes digital signal processing technology come into being and develop rapidly.

In terms of audio quality, the closer digital audio is to analog sound quality after analog-to-digital/digital-to-analog conversion, the better. Digital technology has advantages in audio editing, synthesis, effect processing, storage, transmission and networking, and in terms of price. great advantage.

Since digital signals do not suffer loss and interference like analog signals during transmission, more and more audio products use digital input interfaces, such as CD players and DVDs.

What are the applications of DSP on audio signals?

Application 1: Active Noise Control

The traditional passive sound insulation method simply uses sound insulation materials to block noise, and has almost no ability to block the noise generated by medium and low frequency noise sources.

It must use thick sound insulation materials to produce effects. Active noise control uses an electronic closed-loop control method to produce a sound that is out of phase with the original noise to offset the original noise.

It is extremely effective in suppressing low-frequency noise. The disadvantage is that it cannot control the noise in the middle and high frequency bands (above 1.5K Hz) )

Application 2: Speech Signal Processing

Speech technology can be divided into the following four categories:

A. Speech Enhancement

B. Speech Recognition

C. Speech Coding/Decoding

B. Echo Suppression

A. Speech Enhancement

In terms of the means of acquiring voice signals, various pickups (microphones) have different frequency responses, directivity, stability, and pickup mechanisms.

A combination of multiple microphone arrays with different characteristics can meet the various requirements of users for signals in various frequency bands.

The effective grasp of the electro-acoustic system achieved under the task of noise control enables us to meet the requirements of various user systems for signal pickup.

In terms of signal processing, different schemes can be formulated according to the application, background noise characteristics, and the relative requirements of speech intelligibility to allowable speech distortion.

For example, the requirements of speech recognition software for speech signals are different from the requirements of human ears for speech signals.

Therefore, different procedures are used when performing communication tasks than when performing speech recognition tasks.

If R&D institutions for different tasks cannot have a comprehensive understanding and grasp of voice characteristics, they will not be able to achieve truly optimized results on this.

In addition, DSP technology can perform single-channel signal detection and multi-channel signal comparison at high speed.

Its speed can make users feel that there is no delay in time, and it feels completely real-time.

B. Speech Recognition

The core of the speech recognition system should have the characteristics of less hardware requirements, self-contained time correction, and energy correction.

At present, the independent speech recognition of the small vocabulary (200 characters) system has been practically applied.

In the development direction of automatic speech recognition, it will focus on the development of voice control technology rather than voice input technology.

The focus is on the accuracy of the first recognition, not the auxiliary recognition of mixed semantics.

C. Speech Encoding and Decoding

DSP has powerful functions in speech processing, and can use "code-excited linear prediction" (CELP) type algorithm with high compression ratio.

The currently used open standard is G.723.1 of the ITU.

This algorithm is widely used in IP codec, with two transmission rates of 6.3Kbps and 5.3Kbps, high voice quality, moderate anti-noise ability and calculation load.

It can be used by users on various platforms.

At the same time, the exclusive 2.4Kbps voice coding algorithm is also under development, which is expected to achieve a better balance in voice quality, anti-noise ability, voice compression ratio, computing load, and computing delay.

D. Echo Suppression

In long-distance communication and mobile communication, it is often troubled by echo.

Whether it is linear echo or audio echo, when the delay exceeds 0.5 seconds, it will be clearly received at the receiving end.

For these two phenomena, there are applicable echo suppression algorithms.

The algorithm based on DSP is stable and simple, not only suppresses the response speed quickly, but also maintains the noise reduction performance for Double Talk, Near-End-Speech and mute state.

At the same time, because the linear echo time delay can vary in a wide range from 1 millisecond to 900 milliseconds.

Therefore, there is an algorithm based on DSP to overcome the extra load brought by this variability to the system.

In a traditional echo suppression system, a delay of 300 milliseconds means a sharp deterioration of the system's performance-cost ratio.

The source codes of these algorithms can also be applied to various communication platforms to solve problems arising in various links of long-distance communication.

Application 3: Music Signal Processing

Since digital music standards became popular, due to the flexibility attached to digital signal processing, many open standards and exclusive standards have been produced in the storage, transmission and playback of audio and video signals.

For users, the effects they bring include not only more durable and cheaper storage media, more diversified receiving channels, but also more brilliant audio-visual effects.

However, it is still expensive and not necessarily effective to obtain an audio-visual effect equivalent to that of the original audio-visual signal source at the terminal.

In order to achieve the so-called "surround sound effect", there are various open specifications such as Dolby Surround, Dolby ProLogic, AC-3, THX, etc., and there are also commercialized decoding chips.

But the weakest link in the whole link is the section from the speaker system to the human ear.

The transfer function of this section changes randomly due to different listeners and different listening environments, and even varies greatly.

The hard work of the original recording engineer is often wasted in this section.

And the same as the traditional audio system, this part of performance is the most difficult to grasp, and it is often a part of expensive investment.

For this link, the solution proposed by DSP. It is independent of the above-mentioned open specifications to establish an approximate surround-field sound effect system.

The end

DSP processing technology is a very important application in the audio industry, it makes the playback sound better.

To be precise, it is the same as the camera dynamically adjusts the beautification. So DSP is very useful in the processing of sound and digital signals.