Before describing our solution, let’s talk about what everyone else does: the commonly spread approach of the “single IR” to reproduce a cabinet is nothing more than a digital clean equalizer. Which means: non-linearities are not reproduced at all. Competitors generally model power amps to introduce harmonic distortion, but they totally omit the non-linearities coming from the speaker, the microphone and the mic preamp. While these can be considered somehow negligible for some types of high wattage speakers when dealing high gain sounds, it is essential when the speaker breaks up. Furthermore, sometimes solid state power amps are used to sample cabinet, speakers and mikes, but these power amps, despite being considered “neutral”, have their own “footprints” due to the different impedance, so the interaction with the speaker is noticeably different when compared to a classic tube power-amp.
Another limit of many well-known amp simulations is the complete absence of power amp clipping.
As DiBiQuadro we implemented our own version of Non-Linear Convolution (NLC) to reproduce all the final stages of a miked amplifier. We use to call Pwr/Cab a chain made of:
- a tube power amp,
- a cabinet/speaker,
- a microphone,
- and a high-end studio preamp.
This chain is sampled as a unique block to preserve the correct interaction between the specific power amp and the speakers. Our 64-bit (double precision) NLC engine allows up to 40 kernels (harmonics). In addition, a complete set of parameters gives the full control over the power amp headroom, the type of clipping and the interaction of the negative feedback with the Depth and Presence knobs.
The proprietary solutions just described are grouped under a common acronym: DNC+ (Double-Precision Non-Linear Computing and Convolution).