How to remove the lyrics from a song and leave only the accompaniment?

Some software can eliminate original songs

There are many software that can eliminate original songs. I personally know of Cool Edit Pro, T-RackS, Vogone, Dart Karaoke, WinRIP, etc. I won’t introduce them one by one here. In general, although the effects produced by each software are different, they all fall short of the ideal results. Cool Edit Pro is not an automated software. The most important thing is that you listen to it yourself and then give it commands so that it can help you achieve what your brain needs. If you are interested, you can also search for the following and try them one by one, depending on your needs. Here I will only introduce to you the noise reduction production of Cool Edit Pro and T-RackS.

3. The basic method of using Cool Edit Pro to eliminate the original singing

The method of using Cool Edit Pro to eliminate the original singing can actually be said to be very simple, but if you want to make it very Perfection still requires some effort.

The "Eliminate Original Singer" I am talking about here is just an independent function in the Cool Edit Pro menu. Select and use it and the results will be available immediately. But if you want to get the best results, just using the "Eliminate Original Singer" menu function of Cool Edit Pro is not enough! ! You still have to do a lot of detailed audio processing and settings.

In any case, let us try the simplest method of "eliminating the original song" first.

First enter the "Single Track Editing Mode" interface. Use "File" → "Open" to load an audio file. I have taken "Destined" by Yu Chengqing as an example, and the file name is "Destined.MP3". After tuning in, select "Effects" → "Amplitude" → "Channel Mixer...." (the Chinese version is "Effects" → "Waveform Amplitude" → "Channel Mixer"), and select "Presets" Vocal Cut" see (Figure 1), keep the default settings on the dialog box, and click the "OK" button. After processing, the accompaniment music of the song "Destined" was obtained. The principle of the "Vocal Cut" function is to eliminate all sounds (including vocals and part of the accompaniment) whose sound image position is in the center of the sound field. Therefore, the main thing to use this function is to look at the source of the accompaniment and whether there are musical instruments and vocals placed in the center of the sound field before mixing. If there are, using this function will eliminate them, resulting in attenuation of the sound quality. For example, generally the sound field placed in the center includes "main voice", BASS... and so on. If the audio source you need to mute is the source of the principles I mentioned, I still recommend not to use this function. In this way, the human voice will not be eliminated and the accompaniment music will be eliminated, which is really depressing

Okay, let’s try out your results! Listen very carefully and you will notice that the accompaniment sounds different from the original soundtrack. The original singer's voice in the accompaniment tape has become very "virtual", but the original singer's voice can still be vaguely heard. In fact, this is the so-called silencing effect (it is absolutely impossible to eliminate the original singer). This accompaniment effect can basically be used as karaoke background music. When you sing, your cow-like roar is enough to cover up the original singer's voice. If you are not a friend who requires very original accompaniment, this should still work, hehe. (I'm not finished here yet!)

In addition, the audio quality throughout the accompaniment tape has declined, and what we hear is a lack of three-dimensional accompaniment sound (similar to mono audio ), there may also be some "sizzling" sounds, and the low frequency part has also been weakened a lot. All in all, it sounds very unpleasant.

Perhaps you are very dissatisfied with this effect now, aren't you? Don’t worry, let’s learn some advanced processing techniques. Please pay attention to the idea of ????the “silencing principle” I mentioned earlier.

4. Advanced Cool Edit Pro elimination

First, let’s summarize the results that we are not satisfied with. First, the sound of the original singer seems not clean enough; second, there is a "sizzling" noise; third, the stereo effect is reduced; fourth, the low-frequency effect is weakened.

The above four problems will be corrected one by one below.

Start by opening the "Destined.MP3" soundtrack in Cool Edit Pro.

1. Eliminate the original vocal sound more cleanly

Select "Effects" → "Amplitude" → "Channel Mixer...." (the Chinese version is "Effects" → "Waveform Amplitude" → "Channel Mixer..." "Shrink"), select "Vocal Cut" in "Presets". Previously we dealt with the default values, now we try to adjust the various parameters on the left (Figure 2). During the adjustment process, you can click the "Bypassa" checkbox at any time to compare the original sound.

There are two "Invert" checkboxes on the left. By default, one of them has a "√" in front of it. Now we cancel the "√" and the sound becomes "dry", but The sound field becomes wider and the three-dimensional effect is enhanced, but the low frequencies are reduced more. Maybe this effect is more suitable for some songs, but it is not very effective for some songs. When processing, you still have to debug it slowly and carefully. If this method is not good for the song you want to process, naturally keep the default set up.

Under the default settings, this dialog box is already in a state to eliminate the original sound to the greatest extent. Therefore, if you want to eliminate the original sound more cleanly, there is basically no way here. I personally find that songs sung by different people have different effects on the original singer. Some are better, some are far from ideal. But we can optimize this processing through the four sliders on the left side of the "Channel Remix" dialog box (see Figure 2). Since the final effect is greatly affected by the original soundtrack, there is no fixed "set" for the adjustment here. You can only move the slider slightly to the center point. The stereo effect will be more obvious, but the vocals will also become more obvious. When it comes to stereo and vocal cancellation effects, you have to compromise on one side. Here, the processed result is first saved as "Destined (accompaniment).MP3".