This is an experiment i am doing for a Uni project. To see whether the latest amp modelling software is as good as recording a real valve amp in a blind test.

I am collecting votes on which sample is the real valve amp and which of them are the amp modellers. There are 2 different samples, one strumming and one picking going through 3 amp software modellers and 1 real valve amp.

I tried to set the amp settings, mic type, mic placement and as many other parameters as close to the real valve amp set up as i could, with the features available in the amp software.

The real amp is a Cornford RH30 1x12 combo with an SM57, on axis half way between the cone and the edge. The EQ was set flat across the board as with the gain and master volume.
The signal was split after the guitar, one signal into the amp, the other into the computer. So i got exactly the same performance for the real amp and the software amps.

The Software amps were Peavey MkIII REVALVER, Logic 9's Amp designer and Guitar Rig 4. In no particular order. I would have liked to of tested more, but these were the only ones i could get my hands on.

I am aware that I am not comparing the Cornford with a Cornford modeller, but i couldn't get my hands on a classic Fender or Marshall. I am more interested to see whether guitarists and producers etc can hear the difference between a real valve amp being recorded, or whether modelling software has developed to a point where they can hold their own against real valves.

I apologise about some of the white noise that may come through intermittently. This is because of the software only being a limited demo version and not a fully registered program.


Soundcloud Page

Please post your votes. I'll be collecting them in a few weeks and will let you know the results of who was right!



SAMPLE 1: Real amp - C
SAMPLE 2: Real amp - A
OK, me and my guitarist disagree heavily on these. He thinks

1 - I think valve
2 - solid state
3 - solid state
4 - We suspect this as a valve, but we're not sure.

test 2
1 - solid state
2. solid state
3. He thinks valve I think solid state
4. I think valve, he thinks solid state.

I think for validity of study it's worth pointing out that part of what gives them the distinct sound is the speakers they are through and most people here will be listening to them through crap laptop speakers or little headphones.

I like the idea though.
nothing here is solid state, or am i wrong?
"I came here to chew bubblegum and kick ass... and i'm all out of bubblegum"
@CobenBlack: You are supposed to say which of those are real amps (in this case it's valve) and which are Software amps (-> virtual amps)

To be honest all samples sound pretty good. And I don't play through a valve myself, because I only play inside my small appartment through a practice amp with an amp modeller.

I'll edit what I think into this post after listening to them with better speakers. But I think I have a pretty good idea which is the real amp in sample 1.
Sample 2 is pretty difficult imo.

Sample 1 - B
Sample 2 - A or C
Last edited by Tomura at May 9, 2011,
Yes, there are no solid state amps, just 1 real valve and the other 3 are different software modelling amps.

Thanks for the votes guys!
I know you've tried hard to make it a fair and scientific test but obviously some of the bigger flaws in this are gonna make it hard to take the results as the gospel. Anyway, trying to bear in mind things like hearing the samples through nearfield monitors instead of a speaker cab, and the fact that the amp modelled stuff isn't recorded so will no doubt be different even if you set it up near identically (taking into account you recording the valve amp in an environment different to the one the amp sim developers would have based their sounds on), I'll give this a go to help you out with your research

Out of interest though, might have been a fairer test if you sent the amp sim signal, with the cab emulation turned off, out to the same cab you mic'd for thevalve amp recording, and then recorded that with the same mic in the same position - maybe something you could mention in a write-up as a potential way you feel you could have improved the accuracy of your results, but given the extra time it would take to constantly re-route signals you felt it best to stick to your way for now?

Anyway, here goes:

Sample 1
A: Amp sim
B: Valve
C: Amp sim
D: Amp sim

Sample 2
A: Amp sim
B: Amp sim
C: Amp sim
D: Valve

Edit: Oh, and I fully expect to be proved to be completely wrong - I'm just going by what I feel is a defining characteristic of valve amps over digital modelling, except modelling is getting a lot more convincing these days.

Also, could you PM me the real list? I dunno if I can wait 'a few weeks' You have my oath to secrecy, if so be
Hey, look. Sigs are back.
Last edited by DisarmGoliath at May 9, 2011,
I know I put solid state, but I meant amp sim. I'd also like the answers PM'd if you don't mind. I'll have forgotten about it in a few weeks, and I don't have any particular reason for wanting to screw up your experiment. In fact I'm interested in the results.