Exponential Audio/iZotope R4 & Nimbus have a "Processor Threshold" control that can be adjusted between -48 and -120 dB (defaults to -108dB). I can hear a difference on the first few settings above -48 but not the higher ones. For years I've done extensive sampling/resampling in Live on all sorts of sounds (including those with looong decays down to inaudibility) and have never noticed the cutoff.pottering wrote: Wed Jun 08, 2022 8:06 am People should ask Valhalla Sean and other devs if their reverbs and plugins really sound different when their plugin's tail/release reaches under -120 dBFS compared to the SAME PLUGIN's audible tail/release (in fact very same tail/release too, only earlier in time).
I doubt it, I bet it sounds the same as earlier in the release phase, making the whole process of rising the gain of a inaudible tail worthless compared with just using the audible part.
EDIT: Maybe it does sound different and we will find out something interesting regarding how plugin devs lower CPU usage on their plugins? Gain-based CPU reduction (which is what Ableton does, ironically)? Yet another reason to ask.
But I started recording in analog tape days, so I am always conscious of ITB gain-staging. Plus I use a lot of real-world sources and files with all sorts of bit depths/sample rates, and occasionally add noise & hiss with plugins so my noise floor is probably always above the threshold. Good that there's at least a config file hack. Most likely a relic of the old denormal issue workarounds, and regardless I still find Ableton Live the most inspiring and satisfying tool for all sorts of audio mischief.