Short version: as others have said, the DSP (and limitations of 48/30/20) are the same for both the GLD80 and GLD112. The extra faders on the 112 can be used to control whatever you want: input channels, buses, fx send and returns, DCAs, LRM and matrices. It’s just a higher assignable fader count, not a higher channel/bus count; a fader doesn’t necessarily have to control an input channel.
Long version: the GLD80 is called such because it has 80 “strips:” (4x12)+(4x8)=48+32=80.
These 80x strips can be used to control a combination of inputs (local or remote preamps, local line level, local USB 2-track, expansion card (Dante, ADAT, MADI etc,) to a maximum of 48,) bus masters, fx sends and returns, DCAs and matrices. You can assign anything to any fader you want. As with a lot of digital systems with remote stageboxes, high channel counts and assignable faders, it’s best to think of “inputs,” “channels,” and “faders,” as three separate things.
Your “inputs” are the remote preamps on the AR2412 and AR84, the local preamps on the console itself, the local RCA inputs, the local USB 2-track, and whatever you’ve got coming in via expansion card (Dante, MADI, ADAT etc.) In theory, you could have 112 “inputs” (40x remote, 8x local, 64x Dante) plugged into the desk at the same time.
If you want to process or mix those inputs in any way, you have to assign those “inputs” to “channels.” Both the GLD80 and GLD112 essentially have the same “brains” (DSP,) and can “process” up to a maximum of 44mono and 2stereo channels at a time. You might source all 48 channels from remote preamps and local IO, as is the default. You might have some channels source from the remote stage boxes, and some from the expansion card. You might do virtual y-splits so that two channels are share the same input preamp, but can have different processing applied to each (useful for sending one channel with appropriate dyn and eq to FOH, and the other with different dyn and eq to monitors.) You might not use any stageboxes at all and have all your channels sourced from a Dante card for a virtual sound check. You might be doing a small corporate gig and only be using the local IO for some wireless mics and some stereo playback from an ipod. No matter which way you organise it, you’ve got a maximum of 48 channels to process and mix at the time. The 8x stereo fx returns are counted separately from the 48, however they don’t have full processing; from memory they only have peq. If you want dynamics control on an fx return you can either route it to a bus and apply dynamics to the bus, or you can route it to a (pair of) “proper” channels, (which have full processing,) in which case it will count toward the 48 channel limit.
Now that you’ve assigned which “inputs” (48 or less) are going to be the “active channels” you’re processing, you then need to assign those channels to faders. By default ch1-48 appears on the 4x layers of the bank of 12, and the bank of 8 gets DCAs, bus masters, fx sends and returns, matrices etc, essentially dividing the board into an “inputs-on-the-left, outputs-on-the-right” sort of setup. You can re-arrange them however you want though, and assign the same channel to multiple faders (eg you could assign your money channel to fader(X) on all 4 layers, so no matter which layer you’re on, you’ve always got access to the money channel without having to switch layers,) or have some channels not assigned to any faders (eg using a DCA as a mute group and assigning it to a softkey, or having an fx return at unity and just control it using the fx send master, or vice-versa.) You can set it up whichever way is most comfortable for you to mix.
You can use different show\scene recall with the appropriate safings to quickly change the desk’s configuration to jump between different input/channel/fader routing. Also note that there’s not actually enough faders on the GDL80 to control everything all an once (44mono+2stereo+8fxsend+8fxrtn+16DCA = 78 strips, and that with no buses, mains or mtxs. You’d need another 18 faders (the GLD112 provides an extra 32, yay) if you wanted fader control of everything. Thankfully, users of the GLD80 seem to get by without having fader control of everything, probably due to anything stereo only occupying one fader, using some DCAs as mute groups and assigning them to a softkey instead of a fader, leaving either the fx send or rtn at unity and un-assigning it from the surface, ganging faders on pairs of channels/buses and un-assigning one side, not using all 48 ins, all 30x buses and all 16x DCAs etc.
The GLD112 is called such because it has 112 “strips:” (4x12)+(4x8)+(4x8) = 48+32+32=112.
As with the GLD80, these 112 strips can be used to control inputs, bus masters, fx sends and returns, DCAs, matracies and the input/channel/fader setup is the same as for the GLD80. As the GLD112’s DSP is the same as the GLD80’s, you’ve still got a maximum of 112 inputs, assigned to 48x channels, into 30x buses into 20outs.
The difference is that with the increased fader count, you now have 8 additional faders on 4 layers, (so really and additional 32 faders) to which you can assign those 48 channels/buses/DCAs etc. This means more channels/buses/DCAs etc are immediately available to you; you don’t have to jump through layers as much. The GLD112 also has an additional 4x softkeys (total of 14, as opposed to the 10 softkeys found on the GLD80,) and a sort of mini-ledge on which you could sit an ipad, or a setlist etc.
The disadvantages include initial cost, weight, space required in a truck and at the mix position etc.
AFAIK they’re the only differences: it’s only an increased assignable fader count, NOT an increase in DSP capability.
If simultaneously processing 48 channels into 30x buses into 20x outs is not enough for your application, you need to look at other desks. The iLive will do 64(+fx rtns)/32, other systems from other manufacturers will do more still.
You can assign an input/bus, (and it doesn’t have to be one of the 48 “active channels”) directly to an output (either physical or expansion card,) essentially using the IO as a routing matrix. Apart from a polarity flip on the output you wont have any sort of control over the signal if it isn’t part of the 48/30, but it can be useful in some situations, eg sending playback from a PC to another room. As long as you control the level from the PC itself (not the GLD,) and/or the other room has a level control, and you have the spare output, you can do this without burning up any of your 48/30.)
HTH
