Multi-FX processors vary widely in terms of the features they offer - here are some of the most common:
By definition, all multi-FX units offer some or all of the most popular effects associated with compact pedals - chorus, flanging, phasing, delay, reverb, tremolo, octaving, pitch shifting and so on.
Most multi-FX processors begin by building a basic guitar sound. This generally takes the form of a number of amp models and distortion/overdrive programs, similar to those found in compact pedals. The use of distortion and amp simulation can have implications for the way the unit is integrated into a live setup (see 'Applications').
At the other end of the signal path, speaker simulation aims to imitate the response of a guitar speaker or 'cab', so that a guitar FX unit may be recorded direct, or amplified by connecting to a PA system. So that the same unit may be connected to either a PA system or a guitar amp without reprogramming every sound, speaker simulation is usually controlled globally via a physical switch, although some designs simply incorporate two sets of outputs.
Most effects programs can be pre-set and simply switched on or off. Many multi-FX units also include wah wah however, which requires the inclusion of a continuously variable controller pedal. It may also be possible to assign this to control volume, and there are some units that incorporate both wah and volume pedals.
Most multi-FX processors operate entirely in the digital domain - the signal is only converted back to analogue at the output stage. This helps to reduce some of the noise associated with a long chain of FX pedals, but any process that faithfully emulates analogue distortion will inevitably introduce an amount of gain, which by definition amplifies any noise present at the input. When cranking up the gain on a real Marshall stack in a live setting, noise is unlikely to be a problem as there is usually a fair amount of competition from other equally loud instruments and the audience! However many players use multi-FX processors to achieve a similar sound at a much lower volume - for example at home - and here any noise is much more likely to be heard. Processors therefore often incorporate a noise gate into high gain programs. If the level of noise is constant, it is relatively much more noticeable between notes than during notes. A noise gate shuts off the signal altogether once it has fallen below a given level. Sometimes, this can sound rather crude if notes are audibly 'chopped off' before they have decayed fully, though this effect can be masked by placing other effects such as delay and reverb after the noise gate.