I've noticed that plug-in chains require larger ASIO buffers when they are previewed via View->Plug-in Chain than via FX Favorites->Apply Plug-in Chain in order to avoid clicks, dropouts, or buzz. I can also see a difference in the CPU % monitor, although it's unclear to me why I get noise when the CPU % is only peaking at 2% or 3%. It seems to be more correlated with a rapid change in level than any absolute level.

I've had some success reducing the minimum buffer size necessary by following the Windows tuning advice from Native Instruments http://support.native-instruments.com/hc/en-us/articles/209571729-Windows-Tuning-Tips-for-Audio-Processing, measuring with LatencyMon, disabling SpeedStep, prioritizing background processes, and updating the USB driver associated with the USB port my audio interface is connected to (find which one by browsing through device manager under "view by connection"). This appears to have restored the performance of "Apply Plug-in Chain" to the level it was in my previous Win7/SF10 system, where I routinely previewed plug-in chains that included Acoustic Mirror blocks running at quality level 5 using an ASIO buffer of 64 samples.

Does anyone know why "View" is more CPU intensive than "Apply" and what system properties I should measure/tweak to help it along?

Sound Forge Suite Version 13, build 131, 64 bit, Windows 10 v1903, MOTU Ultralight mk4, 8 core i7-3280 @ 3.60GHz, ASUS Sabertooth x79, 16 GB RAM, 1TB SSD