<d1b2>
<louis> related question to the glitch filter: is there a reason the threshold filter outputs uniform digital waveforms at all?
<d1b2>
<louis> currently it's sparse->sparse and dense->dense, but in both cases it just does it pointwise with no coalescing
<d1b2>
<louis> i see the parallelization advantage, but it produces like 99% waste of space output waveforms when the input dosen't transition often, no?
<azonenberg>
Yes, it does waste some space. I wanted to try and develop an efficient parallel dense -> sparse conversion block we could use for this kind of thing
<azonenberg>
My typical use for thresholding is high speed serial waveforms where you don't have long runs without toggles
<azonenberg>
so there's not a huge amount of waste. maybe 5-10x oversampling at most