A new audio engine powering Neural DSP plugins

Our analog modeling technology has been shifting towards completely automated machine-learning-based methods.

We have developed a new audio engine that uses a proprietary anti-derivative trigonometric interpolation (AATI) algorithm to convert any sample rate used in the system or DAW to an optimum internal processing sample rate. The AATI algorithm allows us to use our machine-learning-based black-box modeling techniques for a wide range of effects in special amplifiers and distortion pedals.

The benefit of using our AATI algorithm is vastly improved stability and CPU efficiency, allowing us to achieve optimal CPU performance and audio quality for any DAW sample rate. This change removes the need for an oversampling switch in our plugins - a change many users have noticed in Archetype: Gojira, with which many users are reporting dramatically improved CPU performance.

We intend to roll this change out to all our other plugins over the following months.



Congratulations on the great sound of Archetype: Gojira. I wrote in another thread that this improved CPU usage has come at the cost of unprecedented latency for an amp sim (129 samples for amp 2 of Gojira). Some people are more sensitive to latency than others, and whilst for some this latency will be inconsequential, for others it may cause difficulties, particularly if they have an interface that struggles at low buffer settings. Since Neural’s plan is now to roll this algorithm out across all its range, I would ask that they please consider following the example of many other developers of high-quality plug-ins in offering a lower-quality zero-latency mode that will automatically render at a higher quality.


I agree, zero-latency is so good for amp sims. Using the Plini archetype, it has gone from zero latency to a small latency, it would be so bad if it increases even more…

But apart that it’s a great news!
Anti-derivative methods prevent aliasing to appear before it even exists, it doesn’t just remove it after it has been created like oversampling does :slight_smile:


The latency in Gojira is a deal breaker for me. When it’s added to the sound card latency and other effect plugins in the chain, it is too noticeable for real-time playing. I really hope that you find some sort of solution for people like me who don’t care much about CPU usage and prefer minimum latency, maybe some switch that toggles between latency/CPU preference. Please don’t ruin the other plugins with this engine before there’s a solution for latency concern.


I thought this was great news, but now you guys have got me worried. Lower CPU load is excellent, but not at the expense of latency. At the end of the day, we’re all guitar players and that immediate responsive feel when playing is essential.
PS - I do hope that the Abasi update deals with the crappy noise gate (I’ve moaning about this for almost a year!)

1 Like

Hey! I was earlier against this but then… I run Neural Cali with full 4x oversampling and 88200Hz sampling rate in the Reaper DAW. It makes a big difference. If the new engine could do this without the high sampling rate this would be great.

Other option: allow 8x oversampling in the old engine like Mercuriall does!

any news about updates?