I'm on a Mac, and after updating to ADAPTIVERB 1.1.0 the plugin isn't found when loading a session - what can I do?

In ADAPTIVERB versions prior to 1.1.0, there was a bug that caused the VST3 plugin ID to not match between Mac and Windows. This bug was fixed in ADAPTIVERB 1.1.0.

However, this fix causes the 1.1.0 Mac VST3 plugin to not be found when loading sessions made on a Mac using earlier versions of the plugin, breaking session recall. This applies only to VST3 on Mac - other platforms and formats, as well as plugin presets made with our own preset manager, are not affected.

As a solution, the ADAPTIVERB 1.1.0 Mac installer archive contains a separate installer for a special VST3 Legacy Support version of ADAPTIVERB - installing this allows opening and playing back problematic sessions. For more information, please refer to the relevant Read Me file, which can be found both inside the Mac archive as well as on the ADAPTIVERB downloads page.

Some presets seem to have a very low output level - why?

Most likely, you're feeding a signal like drums, speech, or other noisy, transient heavy  or unsteady sounds into the plugin - and loading a preset that is focused on the BIONIC SUSTAIN RESYNTHESIZER.

At the heart of ADAPTIVERB lies the BIONIC SUSTAIN RESYNTHESIZER (BSR). Part of what it does is filtering out noise, transients and other types of very short sounds. When you load a BSR-focused ADAPTIVERB preset while feeding such a signal - for example drums - into the plugin, output level will thus be low, as most of that signal will be filtered out.

You can either increase the output gain if you do want to process the pitched, steady parts of that signal with ADAPTiVERB, or you can lower the REVERB SOURCE parameter a little - this causes the input signal to enter the ALLPASS/RAYTRACE reverb module directly, (partially) bypassing the BSR.

ADAPTIVERB & Latency

ADAPTIVERB does some heavy mathematics. More important than the raw amount of calculations required is the fact that it needs to look at a certain amount of input signal to do what it does. This translates into a noticeable processing delay of up to 11264 samples.

The actual amount of latency depends on which modules are used, and is spread out over the components of ADAPTIVERB as follows:

• BIONIC SUSTAIN SYNTH with ALLPASS reverb: 4096 samples
• RAY-TRACING REVERB: 3072 samples
• HCF: 4096 samples

As the BIONIC SUSTAIN RESYNTH, being the heart of the plugin, is always active, the minimum latency achievable is 4096 samples - when using ALLPASS and with the HCF switched off. Using RAY-TRACING REVERB instead of ALLPASS increases the latency to 7168 samples. SUSTAIN & ALLPASS plus HCF weighs in at 8192 samples, while using RAY TRACE as well as the HCF results in a total latency of 11264 samples.

Please note that after changing parameters that affect the plugin's latency, some host applications may require restarting playback in order for the new latency amount to be correctly compensated.

As of version 1.1.0, ADAPTIVERB features a LIVE MODE, switchable using the little clock shaped button in the header bar, top left. When in LIVE mode, the plugin will report zero latency to the host, and disables its internal dry path delay. As a result, the dry path is latency free, and the effect signal will be late by the amount of samples that corresponds to the combination of modules used- and, most importantly, the latency compensation of your DAW does not need to move around all other tracks. Use this mode for example if you want to play virtual instruments live, or capture a live input - or if you are using a Pre-Delay setting in the same ballpark as the plugin latency anyway (make sure to adjust the Pre Delay accordingly). Obviously, the fact that the reverb tail is late makes this mode most useful with long tails and a REVERB MIX of 100%.

ADAPTIVERB CPU Load Optimization

ADAPTIVERB combines several state-of-the-art algorithms to enable functionality not found in any other reverb. As such, the maths involved are significantly more complex and thus computationally "expensive" than traditional reverberation approaches.

That said, there are multiple performance optimization strategies: 

 
• Increase the playback buffer size. We generally recommend you set the playback buffer size on your DAW to 1024 samples. Smaller values increase the CPU load as the plugin needs to "see" some amount of signal in order to "understand" it.
 
• Use PREVIEW mode (Version 1.1 and higher); Preview Mode uses a coarser time resolution and will reduce CPU load considerably if on, while only sacrificing a small amount of sound quality. On low powered systems and if you care more about performance than on sound quality we recommend you turn this on - and turn it off when bouncing tracks or the master.
• Use ALLPASS and switch off the HARMONIC CONTOUR FILTER (HCF). The RAY TRACE reverb models and the HARMONIC CONTOUR FILTER use additional CPU; switching REVERB MODEL to ALLPASS and (or switching the HCF to OFF conserves some CPU. Especially the HCF Keyboard Mode uses noticeable amounts of CPU.
• Disable SIMPLIFY. SIMPLIFY analyzes the signal recursively to determine its perceptually important features. A setting of 0.25 incurs the highest CPU load for this parameter. Switching it off by setting SIMPLIFY to 0.0 lowers CPU load.
• If you are using Logic Pro X, we have heard good things about using a medium processing buffer size and setting the processing thread count to its maximum value instead of "automatic".
• Disable Record Enable on instrument tracks that feed into ADAPTIVERB in Logic. Logic will process any plugins down-stream of a live input - like an audio input or a record-enabled instrument track - in a separate rendering context. As a result, all such down-stream plugins are rendered in the same thread, as well as with a lower buffer size. This can cause a higher CPU load and even CPU overs with one thread maxed out and the others idling. Disabling the record enabled button on instruments allows down stream CPU load to be spread out across threads.