We recently did a blind A/B test:
The cabinet was a stereo 2x12.
Into one side was a Deluxe Reverb. Into the other side was the Axe-Fx into a Matrix amp. The side which each was plugged into was randomized between tests.
The Axe-Fx's speaker impedance curve was set to match the cabinet (it's one of the stock curves). The knobs on the model were set to match the knobs on the amp.
The volume levels were matched.
A footswitch was connected to switch back-and-forth between the amp and the Axe-Fx. The tester was unaware which side of the cabinet was the amp or the Axe-Fx.
In every case the tester picked the device plugged into the right side of the cabinet as sounding "better", regardless of whether it was the Axe-Fx or the amp.
Form you own conclusions.
This matches a lot of my own testing, where even though I knew which is which, by the time you stomp a switch enough times to swap, you get confused which was which if they are dialed close enough.
My setups were these:
- "Into real cab": Fryette PS-100 -> Bluetone 4x10 with 10" Greenbacks or TubeTown 1x12 with Alnico Gold
- Tube amp -> Fryette PS-100 amp in (reactive loadbox), Fryette poweramp out -> cab.
- Axe-Fx 3 Mk2 -> Fryette PS-100 line in, Fryette poweramp out -> cab.
- "Direct into studio monitors":
- Tube amp -> Fryette PS-100 amp in, Fryette line-out -> Axe-Fx 3 -> cab sim IRs created of the Bluetone or Tubetown cabs -> Genelec M40 studio monitors.
- Axe-Fx 3 -> cab sims -> Genelecs.
I used a Sonarworks XREF20 "almost flat" reference mic to make IRs of my cabs using the Axe-Fx 3 IR capture feature.
Since the "Fender-meets-Marshall" Bogner Goldfinger 45 SL and "modernized Vox AC30" Victory VC35 heads are not modeled on the Axe-Fx 3 directly, I used the following:
Real amp | Real amp channel | Axe-Fx 3 model | Axe-Fx 3 model is based on |
Bogner Goldfinger 45 Superlead | Clean | Super Verb | Fender Super Reverb |
| OD '69 mode | Plexi 50W | Marshall Superlead 50W |
| OD '77 mode | Plexi 2204 | Marshall JMP 2204 |
| OD '81 mode | Shiver Lead | Bogner Shiva |
Victory VC35 "The Copper" | Only one channel | AC-20 12AX7 Treble | Morgan AC-20 Deluxe |
This might be a bit off because it's just off the top of my head and it's been about a year so I don't remember the Plexi and Super verb variants totally. JCM800 50W would work just fine for the Bogner GF '81 mode most likely.
The amp models were each dialed by ear to what was my preferred settings on the real amps. No advanced settings were used besides setting the powertubes to 6V6 for the Fender/Marshall models since that's what the Bogner uses.
Volumes were always matched with a decibel meter at 1m distance from the speakers. Testing was done at about 90 dB @ 1m volume levels which is IMO the "tube amps start to sound real good" point. For studio monitors I used maybe about 80-85 dB @ 1m because they are listened to close up, at ear level and in stereo.
Results were that I could dial things to sound very close, but felt there was that last final mile missing when it comes to playing feel. I found that by setting the amp models to the closest matching speaker impedance curve. For the Bluetone 4x10, the 4x10 Brit JM45 impedance curve did the job. For the Tubetown, the AC-20 stock curve was good match. I tried to set the curve manually but found it too difficult to judge by ear so I stuck to the closest preset.
This made me pretty confident about the Axe-Fx 3 capabilities as the results sounded and felt just as good whether I was running into the poweramp and real cabs. The IR vs real cab in the room also sounded very similar if I put my ears closer to the center of the speakers where the mic would be.
Real amps are great in their simplicity - plug into a great cab and turn it loud enough, all good. But I don't see a huge need to own them when most of the time my usecases require less volume so the Axe-Fx 3 is just way more practical.