I've seen it done both ways. Most of the Pro crews I have seen it's in the
FOH rack and there is a different
snake (called a drive
line) going directly to the amp racks.
In my experience, most pros put the
DSP at the amp rack and remotely control it. One of the major limitations of the DriveRack PA, and why many people recommend moving to the 260 or better if possible, is that it does not allow such remote control.
I have often heard that nothing on a
DSP needs adjusted after it is set and I disagree. If you are using the delays to align drivers in the boxes, they stay set, but if you are using the delay to align the cabs to the
stage, that will change on a regular basis.
PEQ per output is is determined by the drivers and cabs, any EQ
system wide would vary per
venue.
It is indeed important to realize that
system tuning is potentially more than just EQ. Although it can be broken down even further, I like the concept of looking at
system tuning as potentially consisting of four general components; box,
array, environment and artistic. The box tuning is usually done to get a nominal response for each
speaker and once set should not have to be adjusted.
Array tuning addresses how the multiple boxes in the
system work together. If you have a fixed
system configuration then this may be
preset and left alone, if the configuration changes then some elements of the tuning may also have to change. Environment tuning addresses both the room in general and specific issues such as the
effect of surfaces near speakers, thus it may be general to all of the
system or specific to individual devices and may be fixed or variable depending upon whether the installation is fixed or changing. Finally, the artistic tuning is the subjective adjustments that may change from artist to artist or even during a single performance, it is getting the desired sound for a specific use much more than optimizing the
system itself.
In the above approach there are several layers of both fixed and variable tuning and this is where I think more advanced processors have a major advantage as they can be configured exactly that way, with multiple 'blocks' of related processing, some of which may never need to be touched once set and others which can then be adjusted without directly affecting the fixed processing. Unfortunately, more basic
DSP devices do not support this approach which is why I typically try to separate the more fixed
system processing that an operator should not have to normally adjust from the more routinely adjusted processing that is located at
FOH.
If this separation is not possible then one consideration may be who is accessing the processing. If it is someone you can trust to not negatively impact the processing already applied, then it may make sense to have it located at
FOH. If you are concerned about someone getting in an messing with the basic
system processing, say adjusting output limiters or playing with crossovers, then it may be best to have it located in the amp rack and secure from such users.
I know the DRPA has a function that uses a
pink noise generator to "auto set" the
FOH graph and that would be better operated from
FOH.
Or even better, not operated at all! Seriously, from only being able to
address one listener position to being limited to the graphic EQ to trying to use EQ to correct for
phase and modal issues, the AutoEQ function in the DRPA is very limited and easily misapplied. It can be useful if you are aware of the limitations but most people relying on AutoEQ probably are not that knowledgeable.