DC offset is the simplest and can be easily eliminated by general audio software. For example, in Sound Forge, just select DC Offset in the Processing menu. This is the first step after recording a piece of music.

Background noise is the biggest problem of personal computer recording, because the room has poor sound insulation ability and the environment is not quiet, resulting in various background noises. For example, the noise of sound cards, speakers, household appliances and fans of computers and hard disks. . . ,

Sampling noise reduction is a scientific method to eliminate noise at present. It first obtains a frequency characteristic of pure noise, and then removes the noise that conforms to the frequency characteristic from the sound in the music waveform mixed with noise.

Cool editing Pro and sampling are experts in sampling and noise reduction.

The principle is understood, and it is also very convenient to implement.

1. Before recording, you can record an environmental noise separately, which is exactly the same as your environment when recording formally. Or before you sing and play the guitar, record pure environmental noise for dozens of seconds. Then record your voice or guitar or something. At this time, this environmental noise should always exist in your recording.

2. After recording, select a piece of pure noise that has just been recorded, and then "sample" the noise (when sampling, it is the Get Noise Sample command). The noise of the sampling section is generally long, otherwise it will not contain enough noise samples. But make sure it is "clean" noise, which means it should be pure environmental noise.

3. Select the waveform range that needs noise reduction and enter the noise reduction setting window (both Cool Edit and Samplitde are noise reduction commands). Adjust the parameters appropriately, and then press OK.

Need to be reminded: eliminating noise will have different degrees of loss to the original sound, listen more and try more, and choose the appropriate compromise. Not only the unbearable background noise is removed, but also the sound will not be excessively deformed.

Common filters for graphic equalization and parameter equalization of graphic equalizer

Graphic EQ graphic equalization can strengthen or weaken the selected frequency band to correct or modify the signal spectrum of this frequency band.

1。 Band attenuation/gain attenuators Each attenuator is responsible for increasing or decreasing the signal strength of the corresponding frequency band. 0 is a change.

2。 Accuracy Accuracy (high, medium and low) Accuracy is used to decide the trade-off between processing speed and filtering accuracy. Low-precision mode is not suitable for dealing with very obvious and strong equalization actions, or waveforms with large low frequency and high sampling rate.

Parametric graphic equalization of segmented equalizer It is a kind of parametric equalizer displayed graphically. It consists of four parameter peak filters, a high-order filter and a low-order filter.

1, dry out: output or intensity of unprocessed signal.

2.Wet Out: the output or intensity of the processed signal.

3.EQ graph: EQ graph) It shows the curve of the current amplitude versus frequency.

4. bandpass filter bandpass filtering: the gain determines the increase or decrease of the band within a certain range (SoundForge4. In 0, double-click the mouse on the fader, and the gain can immediately return to 0. 0dB) width: 8 degrees, located in the center of the selected frequency band (i.e. center frequency). ) symmetrically spread to gain or attenuation of two frequencies. Therefore, the higher the gain value, the wider the width range. (The width range of voice forgery 4 .0 is (0. 3-2。 5 8 degrees)

5. Center Frequency Center frequency is used to fine-tune the center frequency required by the user, so as to facilitate very detailed equalization.

6. The low frequency of the low shelf is equivalent to the cut-off frequency. A form of). Those below the low shelf frequency can be processed.

7. The upper limit frequency of the elevated frame is the same as that of the low frame, and those higher than the elevated frequency can be processed.

(Voice Forgery 4 .0, high shelves and low shelves can be turned on or off. On the, you can adjust and specify whether the frequency is low or high. )

Common Filters for Parameter Equalization

Qualcomm, Qualcomm

High-pass filter is used to remove low frequency and noise caused by low frequency, such as wind, electrical noise or vehicle noise. Its optimum operating frequency band is above 150 Hz.

Peak filter bandpass (peak)

Used to limit or highlight frequencies within a specified range. For example, highlighting the characteristics of sound. Its optimal operating frequency band is in a very narrow range. If it is too wide, it will easily produce "fragile" noise.

Band-stop filter (notch)

This kind of filter can weaken the frequency band in the selected range, and is usually used to remove some narrow bandwidth noise. For example: speaker/microphone feedback, or electrical noise of 60Hz circuit. The optimum operating frequency band is above 150Hz.

If you really want to use balance in production, you need to control the frequency and bandwidth (or Q value).

Frequency control refers to which components of the spectrum you increase or decrease, while bandwidth refers to the increase or decrease of the frequency band. A very basic and useful principle is that when balancing operation, the effect of lifting operation is not as good as that of lowering operation.

If you feel that the low frequency component is insufficient, you can try to reduce the medium frequency or high frequency component instead of blindly increasing the low frequency. If you have to balance the lifting operation, then you should use a smaller lifting amount and a wider bandwidth to get as smooth as possible (at this time, the Q value is set below 2).

The frequencies mentioned here are defined by their effects. The thumping sound is about 70Hz, the warm sound is about 250Hz, and the mud sound is generated from 400Hz to 800Hz. The nasal sound is generally 1 to 2KHz, the urgent sound is about 3-4KHz, the dragging sound is 5KHz, the 6-8KHz is "poof" sound, and the bright sound is 10 to 13.

For example, in order to add some warm elements to harsh sound, we can try to reduce 1 to 2 dB at 3KHz and increase it by 0.5 dB at 275Hz.

If the frequency on the equalizer is high enough, you can try to increase it by 0.5 dB at 18KHz. Doing so can get a very open mixing effect, although you may have to do some attenuation and slightly lower the frequency of 12KHz to prevent too many high-frequency components.

Many remixes benefit from a low q value (0. 2-0. 7), and appropriately reduce the frequency of about 400Hz (-1.5 dB when the q value is 2 to 3). If you use a digital equalizer, you will find that you can't get this rate directly, and they will give you a wider range.

Reverberation reverberation

1, all kinds of reverberation only provide default values including FB and time, and in most cases, the combined sound field has no depth and clarity.

2. The time, FB and delay of reverberation are different.

3. The application of reverberation balance is very important. Its function is to add it at the end of reverberation adjustment, so that the sound field and its timbre have a sense of contraction and tension.

4. Only when the effect of other parameters is not strong enough or extreme, it is suggested to use Delay Dry/Wet, otherwise the sound field adjustment will be very chaotic. What is more important is the harmony of time, FB and balance. Diffusion is dispersion, which plays an indispensable role in the diffusion of timbre on the sound stage. It shows the affinity between timbre and sound stage. Diffusion is the most effective way to make a timbre have a clear outline in the sound field, or to completely fit other timbres. You can even make the timbre a little noisy, so that the timbre and sound field are full of live effects, especially in bars or small parties.

5. In XG format, the advantage of White Room, Tunel and Basement is that they can design different walls, heights, widths and depths for sound field. In other words, it can simulate the sound absorption and reflection ability of surrounding wall materials and materials in real situations. But the bad result is that it will conflict with some effects such as density.

6. In Yamaha CS 1x, when the density is 00, there is no timbre continuity between each sound and the reflected sound. It is the largest at 03, and there is obvious noise feedback with the reflected sound.

7. In order to make the sound stage have a certain sense of space, in addition to the above-mentioned different walls, the room size, even ER 1 and ER2 can also be used. In most cases, we will need reverberation with deep timbre, instead of playing karaoke at home. Wall Vary suggested that 10 should be filled in Cakewalk. Not ten, but one and zero. )

8. reverberation sound phase effect reverberates to the change of sound phase in the feedback part of timbre, which can control the propagation direction of timbre and its reverberation sound effect.

Dynamics and compressors

Problem: When recording vocal music, I always feel out of place. It sounds a little too hard and soft, and it will never get enough impact.

Solution: First, control the track dynamics and attenuate the part between the maximum peak and the average level. The purpose of this is to control the sound channel and obtain full and impactful timbre. At the same time, after the maximum wave peak value is reduced, the overall volume has room for improvement. This is why the compressor is used. In audio software, some peaks can be attenuated, because there are many maximum peaks that can be reduced appropriately, but there is no effect on sound quality.

Defect: Usually, after the intensity of the audio track is controlled, the sound will be strained due to the contraction of the intensity, but at the same time, the high frequency will be lost, and the sound will appear dull and dull. What you get, you lose! Therefore, the adjustment of the compressor must be repeatedly tested before it is officially recorded. This is really time-consuming, but with the gradual enrichment of experience, it can be determined by intuition. Haha, I'll be more creative then. Don't underestimate the processing power of the compressor, the master can definitely play the timbre creativity on the compressor! )

For friends who have just come into contact with compression, here is a brief introduction to the use of the compressor, please correct me!

1, set the compressor: ratio = 2. 5: 1 attack = 10ms (ms) release = 150ms (ms) gain = 0dB If soft/hard files can be selected, select soft.

2. Listen carefully during the audition and slowly adjust the threshold until the sound begins to change. It all depends on your hearing. It is very beneficial to have a pair of good ears. But most compressors have pointer display, and the threshold is usually between -5dB and -20dB, depending on the actual situation.

3. Adjust the compressor to bypass the file and carefully compare the difference between the original signal and the compressed signal. Here, you can increase the gain appropriately (but be very careful) to balance the volume difference. Then continue to adjust the threshold until you are satisfied.

4. To get fuller timbre, you can try to gradually increase the proportion. If you are not experienced enough, I suggest that you constantly compare the difference between the original signal and the compressed signal when debugging. (Bypass is enough) The ratio is usually between 2: 1 and 8: 1, and generally does not exceed 10: 1 unless it is used for special purposes. Excessive suppression will only lead to dry and lifeless voice. (Remember this! )

5. If the squelch generated by the compressor is heard between sound paragraphs, the release time can be appropriately increased, and of course, the threshold can be raised a little. If you find that some sounds become "slurred speech" and have no natural edges and corners, you can adjust the starting time or appropriately attenuate some proportions.

The use of compressor depends largely on your experience and hearing, and the above methods are not absolute! If there is an opportunity, try to compare the working conditions of some compressors as much as possible, and the real experience comes from this. A good producer should first have a very strong professional sensitivity. Almost every excellent sound engineer and producer has his own set of working methods. By observing their work steps, we can learn more from them than from books.

There are many software compressors now, and some of them are really great. However, remember that software compressors are usually used in intermediate and post-production. After the recording. For people like me who can't afford a compressor at home, the software compressor is really a savior! Because in my experience, you have to buy a high-end compressor, or you might as well not buy it. The compressor is another thing with a higher price than the sky, alas. . . )

From the way of use, it is not much different from the hardware compressor. After all, the principle is the same. And now many audio processing software (such as SoundForge, Cool Edit pro, WaveLab) have audio statistical analysis function, which can help you deal with some timbre features that are difficult for human ears to judge. In addition, you can compress different paragraphs differently. If you are used to seeing waveforms, so much the better. By looking at the waveform display results, we can judge which peaks can be appropriately attenuated, which is much more "comfortable" than adjusting the hardware compressor. But the disadvantage is that the pre-recording equipment should be as good as possible. (Distortion, various gains and attenuations caused by the machine itself, signal-to-noise ratio, pickup dynamics, these indicators should be as good as possible. Otherwise, a small step will become a permanent regret! There are many places that are difficult to change on the computer later. Unless you are very proficient in audio theory, rules and audio software, you can really "steal the column".

So like traditional recording, you must carefully consider what sound quality you need before recording. How much room is there for post-production? (In post-production, because the overall effect is emphasized first, sometimes the timbre of human voice or musical instrument has to be trimmed again. If the early recording is not fully considered, there will often be trouble in the later stage! )

The limiter is a special type of compressor. Compression ratio of 10 to 1 or higher is generally defined as limiting action because the output level wave is effectively clamped at the threshold level. For good restrictions, the most basic requirement is to act quickly. The recovery time is generally 0. 1 sec to 1 sec.

Multistage compressor

Typical multilevel compression has 3 to 5 adjustable frequency bands. Three-step type is the most common: low frequency, intermediate frequency and high frequency. Generally speaking, the three-stage formula is enough. If more adjustable frequency bands are needed, the parameter equalization effector is usually more useful. Anyway, it is important to save production time, and try not to complicate things. Because of this, it is also important to establish a complete master tool scheme for yourself.

The final result will be different with different processor connection order and mode. This processor-to-processor connection is called a "chain" in English. The most common connection methods are as follows (both software and hardware are the same):

1 pre-multiband compressor EQ pre-multi-terminal compression equalizer

2 multi-band compressor multi-stage compressor

Post-3 Multi-band Compressor Equalizer Post-stage Multi-stage Compression Equalizer

Loudness amplifier

The use of compression can only start with the minimum compression rate that is not absolutely sure, and don't try to "do it in one step".

Even some experienced engineers will only choose the compression amount by experience, and then draw a conclusion after listening. In the process of mixing, there will never be any "absolute", and any operation is only responsible for the final result.

Compression is generally considered from the bottom. This lowest level means that compression only processes sounds above this level. There is also a common method to compress the lower limit, not the level, but the frequency; That is to say, the compressor will only affect the frequency band above this frequency. But now more and more compressors (especially digital compressors) will provide these two parameters for adjustment. The reason why compression starts from a lower level is because the compressor can act as an "exciter" at this time, that is, it can present high frequency at a lower level. However, this compression method is sometimes "too effective" for low-order high-frequency bands. Therefore, before compression, the high-frequency lifting (through the equalizer) should not be too ostentatious, and must be enough or even reserved. This depends largely on experience, because the actual situation is often very different and cannot be generalized. Otherwise, no one needs to do it, just design a program. )

Using the compressor can also expand the force distribution at high frequency, thus diluting the high frequency. However, it should be noted that this technique is more suitable for short high-frequency timbre, such as stepping on a cymbal. High frequency is very fragile, and a small change in frequency intensity will change the characteristics of this timbre, but for short timbre, it is at least relatively relieved in hearing, so it is not suitable for continuous high frequency.

Another important feature of high frequency is that if the level is not very high, it is more likely to be "swallowed up" by other frequency bands. Therefore, such timbre usually needs to strengthen its peak value to ensure that it has a chance to "show its face" before being swallowed up.

Using professional compressor or compression software, the frequency band of each instrument can be effectively controlled, and the sound will be extremely full when recording in stages and tracks. (However, it should be noted that the use of the compressor will also cause you to lose some high frequencies, so you must grasp it yourself. ) It is worth mentioning the vacuum tube compressor, equalizer and sound box. If the system adopts the signal processing chain of "analog-electron tube-compression" mode, it is too easy to get full and gorgeous sound quality. Unfortunately, it is too expensive and inconvenient.

In any case, the maximum compression should not exceed 5db.

The attack time and release time are generally between a few milliseconds and hundreds of milliseconds.

The sharper the peak at the top of the waveform, the greater the ratio, such as floor heave. The wider the peak, the smaller the proportion, such as strings and voices.

Phase butterfly

Stereo music works need phase modulation, which is a very important part of mixing. In this way, the mixed music works can be clearly layered and truly reflect the "stereo" effect.

The adjustment of phase can be divided into left and right, front and back. The adjustment of the left and right channels is relatively simple, which can be easily adjusted by general audio software. As long as you wear headphones, you can easily adjust and hear the adjusted effect.

The adjustment of the front and rear levels is more complicated. Generally speaking, it is difficult for simple hardware devices to record multi-track music with a strong sense of hierarchy. Later, it can only be adjusted slowly by software. There are also some special effects plug-ins to adjust the sound phase, such as Pan Handle and Ultrafunk's surround sound plug-ins are all good phase adjusters.

In fact, their principles are all based on the auditory characteristics of human ears, using reverberation effect to create so-called "far" and "near". So you can make a feeling before and after. Of course, this effect is certainly not as good as the "3D" sound effect in the game, but it is enough for the samples of music works.

The relationship between before and after is actually the relationship between reverberation time. For the sound to be in the rear, if the reverberation time is long, its sound will be vague and have a sense of distance; For the sound to be in front, if the reverberation time is shorter, its sound will be clearer and "approximate".

In a song, sometimes we need the sound to go first and the instrument to follow. Sometimes we need bass in front and drums in the back. Sometimes we need a guitar solo in front and an accompaniment behind. These effects can be adjusted by reverberation. The key is to listen and try more.

Overall downmix output

In fact, there is nothing to mention in this part. Because we use computer software to operate, the production process is naturally different from that of professional masters.

After editing in Samplitude 2496, Cool Edit Pro and other multi-channel recording software, select the "export" command (Mix Down command is in Cool Edit Pro), and multi-channel audio can be synthesized into 2-channel stereo output. You can choose the output format, for example. wav、. mp3、。 Rm, wait. Of course. Wav files have the best sound quality and take up much more space.

In this way, the hard work of recording, editing and mixing finally paid off. Sometimes, it is deduced in this way. Need to process Wav files. For example, T-racks, the main processing software, is used to compress it a little at last, which makes the sound softer and warmer and more in line with people's hearing habits.

Finally, find a CD-R and put it on. Carve wav files into CDs and show them to others.

Or ... hang the .mp3 file you output on the Internet and share it with everyone! ! !