This composer was so obsessed with his image, he posed for his own deathbed photo while still alive
Demonstrating the Waves J37 analog tape emulation plug-in and comparison with a real tape recorder
Who should judge whether or not a mix is good?
Responding to a client's requests for changes - a vital skill in audio
Who should be responsible for the fade at the end of a song - the producer, mix engineer or mastering engineer?
How to find the best tempo (BPM) for your recording
M-Audio Studiophile AV 30 monitors - Could they be Auratones for the digital age?
Why haven't you had a Hot 100 hit yet?
How Isabel Fay got noticed on YouTube
Is it possible to *produce* classical music?
If you have a bass guitar recording that varies in level, it is likely that it will be too loud at some points in the song, too quiet in others. There is no one position of the fader that seems correct all the way through.
There are several ways to handle this problem (other than ignoring it and convincing yourself it is a product of the musicianship of the player)...
Fader automation would be possible, but when the problem could exist on a note-by-note basis, i.e. quiet-loud-quiet etc, it could get very fiddly. The same applies to clip-based gain, which might involve splitting the track into dozens of separate segments, many as short as just one note. Compression might seem like an obvious solution, but it is important to consider the alternatives.
Compression is often used because it just sounds nice, or to change the dynamic structure of individual notes (which is very applicable to bass in the appropriate context). But here we just want to get the level to be consistent, so that the track is easier to mix, and has a more solid foundation in the low-frequency region.
In this case therefore compression is only used to correct the varying levels of notes. Any other changes should be minimized.
A quick look through the waveform display, zoomed to make individual notes visible, will show up any major problems - notes that are massively too loud or too quiet. These are best fixed using clip-based gain, or whatever similar technique your DAW provides. By doing this, you can preserve the sonic texture of these notes. Otherwise, a) the very loud notes would be extremely compressed, and b) you would end up compressing all but the quietest notes, so the bulk of the bass guitar track would be compressed when it really isn't necessary to do so.
Since what we want to achieve is to smooth out the variations in level, the compressor should only kick in on louder notes. Most of the time it should be inactive. The gain reduction meter will show you this. Your aim should be that quite a lot of the time there is no compression at all. If the gain reduction meter shows compression all the way through, then you are compressing too much for the purpose of level correction.
If your compressor has a threshold control, then you will adjust the threshold and ratio to get a good control over the notes that are too loud in level. The gain make-up control can then be used to bring up the overall level, so that the quieter notes are raised up.
If your compressor doesn't have a threshold control, then you will balance the input control and ratio to achieve the same thing, always with an eye on the gain reduction meter.
Since the object of the exercise was to control the level of the bass, the end result should not sound obviously compressed. It should just sound more consistent in level. This will almost certainly be easier to mix than it was before, and the finished mix should sound more solid and more confident.
Takeaway - When compressing to control dynamic range, the gain reduction meter should go all the way down to zero in the quiet sections of your original recording. Otherwise you're compressing too much and changing the texture of the instrument.