Just for the books, no sluggish behaviour here, either crashing or running fine (on outdated CPU and GPU).

 

Might I add then that the sluggishness I get only occurs inside the plugin... No heavy CPU-load, nothing. Doubt it's the drivers, I never encountered this before (and I have quite an array of plugins, all working perfectly.)

 

Interesting facts:
- dll/vst3 naming is different (tbtech 1.5.1 vs PA 1.6.0), so updating to 1.6.0 may caise a problem for opening old projects, settings. Don't know if renaming a newer version dll/vst3 to the older plugin name as it was before may help, so I keep both, 1.5.1 and 1.6.0
- 117 bit.... Don't know what it is for.... Dynamic range or quantization noise..... But if you create simple 1kHz tone with 64 bit float precision (can you create more precision than that?) or plugindoctor test and add tens of high shelves boosts, it will lift noise from the ground up. Switching from 64bit to 117bit shows that 117bit is slightly (few dBs) less noisier..... But how it will be better for your 16/24/32 bit samples, or after your 32float/64float processor where quantization shit is higher? Or what it is for? What can this 117bit improve? If it was 65-66 bit, I would believe in that small precision boost :D I know that Pro-Q3 is 32 bit float, and people call it megasuper quality eq you ever need, but 64 bit is more than enough? Or should I wait for 234 bit (117 bit twice) to kill cpu just for noise at -50000000dB instead of -49999999.8dB?

 

Even the manual doesn't say what it's good for exactly:

Spoiler

The internal processing of Kirchhoff-EQ uses 64-bits double
precision, but you can choose to have precision higher than
64-bits. When this function is on, inner processing would
use 117-bit precision, and doubles CPU workload.
16-bits, 24-bits, 32-bits, 64-bits, and now we have 117-bits. Why the higher precision? A 16-bit CD-
quality is good enough? Unfortunately, not enough for equalizers. Let me describe the underlying: in
computational science, the crucial thing is not error itself, but accumulation of error. Imagine a flying
spaceship with a navigation system that accumulates 1 cm of error per second. Though 1 cm seems
small, it would be 2.5 kilometers after 3 days! Similarly, the IIR-based equalizers have recursion
calculations, in which errors are accumulated to future results.
The essential way to solve this is to use higher precision. When we start to design Kirchhoff-EQ, we
choose 64-bits and it's impossible to use a lower precision. You may ask that since there are 64-bits,
can we simply combine into 128 bits? It’s not that simple: the 64-bits calculation is supported by CPU
hardware which is very fast, while high precision is software simulated that is much slower. We have
searched for a long time to find approaches that achieve higher precision with acceptable CPU
consumption. We have found "Double-Double" that uses hardware 64-bits and some mathematics
techniques to achieve ultra-high 117-bits precision.
So, what does it sounds like? Most times 64-bits are enough, and the improvements of 117-bits are
subtle for to people. In our comparative tests, sometimes the effect of 117-bits is significant. Sometimes
not perceivable, so it relates to the processed audio content. It's hard to clarify in which specific
scenario this mode is required. However, since it is simple to turn it on and off, you can simply make
your own comparisons.

 

Yeah
Better than ideal ever
Higher than physically possible maximum
We did it but don't know why and what for, u decide

People still don't know the total purpose of 24 bit dither before sound from floating point DAW hit the DAC.... People still don't know about oversampling, how much to use it, will it help and what's the difference (except sines clipping, ofc).... Many don't know about difference: cramping vs decramping vs natural/analog phase.....true peak vs pcm limiting/clipping... Or fir vs iir..... And blah blah more serious stuff.... And now get the 117 bit and its enigmatic power nobody knows, even developers don't know...... Good, ok.... But will help if this difference 64/117 will be below audio,below recorded noise, below 24bit recorded maximum, or being below aliasing, and 16 bit dithered (or not dithered) audio or even most usable mp3 final file......
Where my f*cking cassette tapes? O_o

Marketing....
Maybe their 117 bit works fine.... But..... It is better than PSP MasterQ2 with their 80bit extended precision..... So, MasterQ2 is better than all those 64bit/32bit float EQ you all know? Ahahahahah....... Who will release EQ with more bits?

Digital audio problems solved :D
64 bit, 80 bit, 117 bit, Reaper with up to 16x oversampling for plugins with its own oversampling which are put in 64x oversampling in DDMF Supermetaplugin for any sample rate chosen in Reaper.... Burn cpu for analog audio)

Happy New Year
Merry Xmas all
Wish you more bits and less aliasing