A Bandpass Filter (BPF) describes the filtering process. A BPF refers to a type of filter circuit that passes frequencies within a specific range and weakens signals that fall out of that range. A bandpass signal is a signal that involves a band of frequencies that are not joined to a zero frequency. This signal usually comes out of a bandpass filter. A perfect bandpass filter would have a totally flat passband that would mitigate all frequencies that fall out of the passband. The transition would be completely solid.
BPF have standard settings regarding frequency band, quality factors, gain and slope. Let’s start from the beginning. The frequency band is capable of setting a band that is not affected by the filter. A frequency band is able to mitigate the frequencies that fall outside of a specific spectrum, and once it reaches its limit, the amplitude will tend to decrease until the signal turns very weak and/or disappears. The Q-factor is reciprocal to the fractional bandwidth, so a high-quality filter will have a narrow passband and a low-quality filter will have a wide passband.
Another standard-setting regarding the bandpass filter is gain. Gain sets the amplification degree applied to the audio signal and has a direct impact over the frequencies of both the pass and the attenuation bands. Also, it tends to rise or reduce the sound volume. Slope is another standard setting of the bandpass filter and it sets the frequency attenuation rate. By doing this, the transition between the affected and the untouched frequencies varies. A higher value is traduced into a more precise transition.
It’s impossible to find a perfect BPF. It’s impossible to mitigate all frequencies that fall outside the desired range because there is an area outside the intended passband that is capable of weakening frequencies but can never reject them. This is a filter roll-off expressed in dB of attenuation per octave or decade of frequency. The BPF is designed to make the roll-off as tight as possible.