Causes of vocals being depressed by a compressor｜Settings that bring them forward but don't make them hurt

The moment you turn on the compressor, even though the volume is the same, the vocals may fall deeper into the accompaniment. The cause is not the ``compression'' itself, but the fact that the components that create the sense of front and back of the voice are also removed.

1｜Attack too fast

The rise of the voice contains information that determines the outline and distance of the words. If you set the attack to near its fastest speed, the contours of the sound will be suppressed, resulting in a voice that has volume but doesn't come forward. First, slow down the attack a little and find a place where the beginning of the consonant will naturally pass.

2｜Release does not match the movement of the voice

If the release is too long, the compression will remain until the next word and the entire phrase will sink. On the other hand, if it is too short, the gain will fluctuate depending on the end of the word or breath, making it difficult to settle down. It will be easier to judge if the starting point is how much the meter returns to just before the next word.

3｜Gain reduction is too deep

Use clip gain and volume automation to smooth out only the big peaks before continuing to hold down by an average of 6 to 10 dB. Then, try using a compressor of 1 to 3 dB at all times, and around 4 dB for strong parts, to make it easier to preserve the expression of the voice.

4｜Reacts too much to high frequencies

Sibilants and plosives can be a trigger, causing the entire voice to drop lower than necessary. Lightly use the side chain high pass and de-esser in the front stage to create a condition where the compressor does not go wild with just the low range and sibilance.

5｜Incorrect comparison based on makeup gain

Only after processing, if the sound is loud, you can hear it better, and if it is low, it feels like it has gone deeper. After adjusting to the same perceived volume as when bypassing, compare the naturalness of contours, breath, and endings.

The order of practice is to smooth out large differences with clip gain, create density with light compression, add another layer of support with a slower attack if needed, and finally bring the song forward with automation. If you don't try to solve everything with one device, your voice will be softer and more listenable.

The goal is not to keep the meter constant. If you can feel that the same person is singing both the soft and loud voices, and the consonants do not sting even when you turn up the volume, then the settings are working.