What do people mean when they say modelers lack that "amp feel"?
In the early days of modeling the sound was more like they used a fuzz or distortion pedal with various filters to imitate the speakers. If you ever used a Fender Blender or a Fuzz Face, the feeling was closer to that. It didn’t respond to dynamics when picking or reducing the volume.
A good tube amp is analog and has almost infinite states between a clean sine wave and clipping that results in a square wave, where the old boxes really didn’t, they lacked the nuances and felt like they were only clean or only dirty. On stage, an electric rock guitar is a curious beast in that it requires an interaction with another device, the amplifier, coupled through physics, the movement of air from the speaker, to sound the way we like. Take away or change any of those and the feeling of the system changes and, because we’re touching the strings and body of the guitar we feel how the guitar behaves differently. The physics go much deeper than just the physics of the air moving though, because the electrons flowing through the circuitry, including the tubes, constantly react to each other, and the components pushing and pulling that flow cause the amp to react and change how it moves the speaker. Those changes are also part of the “amp feel” that was missing in early solid-state amplifiers and modelers. It’s not that we’re touching the amp to sense its feel, we’re sensing the guitar’s response to the air pressure waves coming from the amp, but we call it “the feel of the amp.”
Every parameter we can set in the Amp and Cab block is applied to the signal, then the resulting signal is processed again, multiple times, and eventually passed through to the Cab block which applies the IR as a filter. This is where Cliff’s design has succeeded so well because he’s studied those in-between states across the amp’s circuit, and the components, along with how the speaker’s movement affects the signal from the amp and pushes back to the transformer, and figured out how to define it all using mathematical algorithms.
Years ago a friend of mine was involved in the very early days of attempts to model amplifiers, and he said to pay attention to the field, so I have always kept my eyes open, watching for progress. I bounced off several attempts and was not impressed until I got my first Fractal, but it was the current generation of units that convinced me that they’ve arrived, especially after Cygnus was released.
Finally, I have two nice boutique amps that my friends enjoy using when I’m not using them, and they use my Fractals when I want to get their opinions of some preset I’ve been working on, and they have been very happy with the way the modelers sound and feel, especially the last couple releases of the firmware. Last Wednesday I was standing beside the rig listening as a friend was running through a particular preset, and it sounded just like a big Mesa stack as he gave it a workout. He was grinning from ear to ear when he stepped down, and I specifically asked him how it
felt and he said it was “just right”. I thought that was cool, because we’re past the “it doesn’t feel like an amp” B.S.