Noise filtering in sound files?

Question

Content

At school, I record lectures on a dictaphone. As a result, there are several mp3 files. Despite the fact that I am sitting on the first desk, it is not important to hear the teacher, when listening to a lecture, I have to turn the sound to the maximum. As a result, any rustle is very clearly audible, but if someone put a pen on a desk, it is generally akin to a grenade explosion.
Please advise a program that can filter out all these noises (preferably in automatic mode). Or somehow highlight a person's voice against the general background

Answer 1

Audacity, for example. In general, a practically more or less serious program can do it. But this only applies to "background" noise. The principle is the same everywhere. A fragment containing nothing but one homogeneous noise is taken as a sample, and on the basis of this fragment noise is removed throughout the entire recording.
Regarding the selection of a person's voice. I would probably try to use an equalizer (cut lows, boost mids). Well, play around with the compression. In short, the meaning of compression is as follows: Increase the volume of the sound, the amplitude of which lies below the specified level, and do not touch the part, the level of which is higher than the specified one.
In general, read about compressors, limiters and gates :)

Answer 2

Goldwave. The program is small and convenient, not free, but the serial number is easy to find. It can cut noise and do a lot of other things.

Answer 3

The only radical way to improve the quality of the recording is to try to negotiate with the lecturers and place the recorder as close to them as possible.
Otherwise, even if you clean out the background monotonous noise by sampling, limit the neighboring peaks, filter out the speech band, it turns out that the level of reflections from the walls and ceiling is comparable to the main one, and continuous jewelry playing with a gate in real time helps a little ... This is me very briefly, so as not to much ship. As a result, you still have to strain your thinking apparatus in order to realize the mess that has turned out.
This is me as a person who many times tried to solve a similar problem in a professional studio. Bringing an unsuccessfully recorded interview into a more or less digestible form, lasting several minutes, can take an hour or more.

Answer 4

Thanks everyone for the replies

Answer 5

It seems to me that a directional microphone (even a cheap one), normalization and equalization (perhaps a special preset for a particular lecturer) of the sound will help.