The description for peak hold is as follows:
Peak hold defines the time that signal level detector holds its maximum. You can imagine that when an attack stage ends, before the release stage starts, there can be an additional peak hold stage and the level is not falling yet. This is true only when true peak mode is enabled (check advanced detector settings if available).
It is often used in gates to avoid the level falling below the threshold too quickly, while having short release times. If you want the gate to close quickly, you need a short release time. But in that case the ending may be too abrupt and even cause some distortion. So you use the peak hold to delay the release stage.
It is also used along with look-ahead to avoid distortion in limiters and compressors. If you need a very short attack, the attack stage may be too quick and cause distortions. In limiters this attack is often 0ms, in which case it becomes a clipper. Setting look-ahead and peak hold to the same value will make the detector move ahead in time, so it can react to attack stages before they actually occur and yet hold the levels for the actual signal to come.


