Just thought I'd add a little insight into the idea of windowing and how it relates to my Minnesota presets. Specifically the "North" version, and its "Digital" mode multiband)...
The way it works in a "normal" way... not only accomplishes what windowed hold does, but also accomplishes what windowed multiple time constants does. Not just accomplishing, but blending completely fluidly between the benefits of both of those, to the exact degree needed to perfectly adjust to the dynamics. The benefits are that it doesn't gain-ride the sound, and it changes gain quicker when the difference in input loudness is larger so relatively small changes in input loudness still cause relatively slow changes in gain.
That is done by setting "Multiband" (in "Digital" mode) > "Behavior" > "Attack Flatness" and "Release Flatness" to 0.5, and then turning up the "Speeds" > "Attack Time" and "Release Time" a LOT (WAY LOTS TONS more). Those times then become the slowest that the attack & release will operate at. Minnesota uses a 3.2 second "Attack Time"! But with the "Attack Flatness" at 0.5, the attack can end up happening VERY fast if there's a large increase in the input.
You'll also want the "Behavior" > "Release Inertia" and "Continuous Release" to be turned all the down/off. This is where I'm doing things differently in Minnesota.
Minnesota's release uses "Release Flatness" of 1.5, which makes larger releases happen
slower than small ones. It also uses "Release Inertia" of 2.0, which causes the release to speed up the longer that it's happening. The result is that tiny changes (between 0-2dB) in release are allowed to happen quickly (80ms per 1 dB after release hold, and 1dB of change is already slower). Medium & large releases happen at progressively slower initial speeds, and progressively higher final speeds. So it's kind of a dual effect where the release won't ride medium to large reductions in the input loudness, but it also will eventually "kick in" fast enough for the ear/brain to not notice that the loudness changed for a split second.
Cheers
If you're not a processing freak, or freak in training, you can stop reading here if you want to.
It's very important, for either method of dynamic release speeds, to get the "Levels" > "Gate slowdown"/"Gate" set correctly. I recommend a starting point that you need a little basic math to calculate, unless you're using ~Infinity:1 "Speeds" > "Ratio" (then you can just set the "Gate" levels to be the same as the "Levels" > "Threshold level"/"Thres"). You need to do this for each band of course.
"Threshold level" + (Median Gain Reduction / "Ratio" Numerator)
So for instance... -24dB "Threshold", -10dB median gain reduction, 4:1 "Ratio" numerator is 4.
-24 - (-10 / 4)
-24 - -2.5
-24 + 2.5
Recommended starting point for "Gate slowdown"/"Gate" =
-21.5 dB
Lower ratios will cause the recommendation to be higher. More gain reduction will also cause the recommendation to be higher.
Register
Login
FAQ
Search
BETA003 is going to be around 1300 MB, as I said I need to look into this...