Airwindows ADClip7 FIX: AU, Mac and PC VST

[ronaldsifontes]

Hi.
{Is this the known Reaper bug where the menu items are off by one?} So it seems.
{I can't change that without breaking it for all other DAWs.} Ups
{Is it that the first two settings are the same, then there's the gain match (?) and clip only is inaccessible?} Actually no way works the same both AU vs VST. VST works fine.

[theMuzzl3]

I am unsure of where the bugs are coming from, but I do know of a solution of some sort.

Use bluecat's patchwork or MB-7 with airwindows plugins, and parameter values will be displayed correctly. Additionally, the knobs in the graphical interface within bluecat's are very smooth and the graphic intensity is very small. I had issues in Ableton, in the past, where some of airwindows plugins displayed incorrect value parameters (I think ADClip 7 being one, where the 3 values of the bottom slider functioned correctly but displayed different names than the names that were in Chris's youtube video); and also, Ableton did provide sliders to change the parameters but when shift + click & slide were applied... it did "fine tune" the sliders but not as well as it did within bluecat's. Also, in bluecat's, you can double click a parameter value and enter a numerical value.

Bluecat's worked closely with Chris in order to make 100% sure that Patchwork worked perfectly with his airwindows plugins. So, IMHO, it is worth the investment. It also allows you to host AAX, AU, VST2, VST3 plugins in any DAW.

[heavymetalmixer]

Today I was trying this plugin on a snare bus (from a snare top and snare bottom) in Solo. The CPU-leak bug appeared again under these conditions:

Normal mode: Boost between 10.56 dB and 12.99 dB
Atten mode (I'm a W7 x64 user): Boost between 10.47 dB and 13.12 dB
Clips mode: Boost between 10.56 dB and 13.06 dB

If Soften or Enhance (or both) are at values of 0.400000 or lower once Boost enters in the specified range, then the bug doesn't appear. If any of the two sliders is moved to a higher value, no matter how small the increase, the bug appears.

I tried with the News-Update version I downloaded a few minutes ago.

MY PC specs:
Windows 7 Ultimate SP1 x64
Reaper 5.974
AMD FX-6100
8GB RAM
GPU R7 260X 1GB

[heavymetalmixer]

Today I was trying this plugin on a snare bus (from a snare top and snare bottom) in Solo. The CPU-leak bug appeared again under these conditions:

Normal mode: Boost between 10.56 dB and 12.99 dB
Atten mode (I'm a W7 x64 user): Boost between 10.47 dB and 13.12 dB
Clips mode: Boost between 10.56 dB and 13.06 dB

If Soften or Enhance (or both) are at values of 0.400000 or lower once Boost enters in the specified range, then the bug doesn't appear. If any of the two sliders is moved to a higher value, no matter how small the increase, the bug appears.

I tried with the News-Update version I downloaded a few minutes ago.

MY PC specs:
Windows 7 Ultimate SP1 x64
Reaper 5.974
AMD FX-6100
8GB RAM
GPU R7 260X 1GB

EDIT: Reaper says that the maximum peak on the snare without any clipping is -9.2 dB.

[jinxtigr]

I don't know why it's doing this, but there is some chance that work I'm doing now on revising the denormals code and combining it with the floating point dithering, will lead to a fix for you.

[heavymetalmixer]

I don't know why it's doing this, but there is some chance that work I'm doing now on revising the denormals code and combining it with the floating point dithering, will lead to a fix for you.

I'm crossing my fingers so it happens.

[goodsine]

Bit of a necro bump but it's boost that's causing the CPU bug with ADClip7 in Reaper and other daws. It's easy to reproduce by cranking up the boost slider then returning it back to zero or a low value. After a few seconds it'll start to bug out. I had LFOs swinging on the other parameters for 5 minutes and the plugin could have ran indefinitely. So it seems the problem is with boost specifically, whether anything else was touched made no difference.

[jinxtigr]

Well, that's interesting. Because this is the code, and in the AU the (A*18.0) part is also 0 to 18, floating point. And I'm able to use that in EveryTrim to do a decibel-to-multiply conversion where '0' gives you a *= 1.0 result. Though it seems weird to take something to the zeroth power.

double inputGain = pow(10.0,(A*18.0)/20.0);
if (inputGain != 1.0) {
inputSampleL *= inputGain;
inputSampleR *= inputGain;}

This is the code. Are we having problems with this dB to ratio conversion? The boost is otherwise just a multiply. EveryTrim uses the same method. Does that bug out too? With that you've got multiple sliders all of which can be made to do that behavior (taking something to the power of zero).