I’ve got a preset using input 1 on row 1 and row 3, then 2 different amps/cabs with delays,reverbs,etc.
I think I understand that the qc splits it’s processors between 1/2 & 3/4, but what I don’t understand is this. I’m at 60% overall, but now can’t really add more devices. If I add to row 1/2, cpu goes to 61% but if I try and copy and paste to 3/4 it says not enough cpu to paste, or if I just try to add another block to 3/4 the options are severely limited.
Reverbs, Delays and time based FX use the most processing, that being said you probably need to find the most efficient way of building your signal chain. I have presets that have multiple captures and FX built into several scenes with no CPU limitations. In fact, since 1.3.0, the overall DSP efficiency has improved greatly. Make sure if you are using a gate block at the beginning, maybe remove and use the built in gate (long press on the signal chain input) and take advantage of splitters / mixers to get the job done. If you continue to have issues, email support@neuraldsp.com and they will get you sorted quickly.
If the QC’s architecture is similar to the Helix in this respect, then the problem is that each pair of rows uses approximately half the DSP. The DSP is NOT being dynamically assigned out of one large single pool to all four rows. That means if you are using close to 50% of the available DSP on the second pair of rows you will need to assign any additional blocks to the first two rows. Even if the CPU monitor indicates less than 100% usage, and there is space left in the second pair of rows to place a block.
The other question would be, how accurately the CPU monitor is reflecting available DSP for blocks. Have not really done any testing of this yet so I can’t answer this question.