Amp vs. Models Contest: The Winners!

WARNING! This post contains the answers to the quiz! If you still want to take the test, stop reading now and challenge your ear here.

Don't worry, it's only Photoshop. We would never inflict anything this hideous on one of our winners.

It was only two weeks, but it feels so much longer—at least for me, after processing some 220 entries before receiving the third perfect eight-out-of-eight score last night, which concluded the contest.

I honestly didn’t think it would take as long as it did. Not that I’m complaining! The Amps Vs. Models Contest was a fascinating experiment that produced many interesting comments. Thanks to everyone who chimed in!

The final entry was from Jessie Nieboer of Walkerville, Michigan, who selected a Seymour Duncan Twin Tube Mayhem pedal as his prize. Several days earlier, Ralf Tyra of Hamburg, Germany, claimed second place. He chose a Twin Tube Classic and a Tweak Fuzz. And last week Colm Kelley of Dublin, Ireland, took first place with the first perfect score. He hasn’t yet decided which stompboxes he wants, but mentioned the possibility of three Deja Vu delay pedals.

And the correct answers?

D, C, B, A, F, E, G, H.

The amps pairs appeared in chronological order: The first two clips were the Fenders, the next two were the Voxes, followed by the Marshalls and the Diezels. In each case, the model appeared before the amp, except in the case of the Diezels. In other words, Clips 1, 3, 5, and 8 were models, while 2, 4, 6, and 7 were amps.

I’ve said all along that I probably wouldn’t have aced this test if I hadn’t made it myself. And most of you were in the same boat. Ten entries received a score of 6. Everyone else scored lower.

Some post-game observations:

• Listeners not only had difficulty discerning amps from models, but also identifying amp types. Many confused the Vox and Fender sounds. Even more mixed up the Marshall and the Diezel. In retrospect, it may have been a bad idea to include the Diezel, since its sound is not as widely known as the others. My thinking at the time was that I wanted to include something non-vintage and metal-approved, and I consider the Diezel an especially good-sounding example of this sort of amp. I bet more people would have ID’ed the amp correctly if I’ve played a more metal-sounding riff.

• The model that fooled the most people was the Marshall. Many, many contestants mistook the real Marshall for the model, which is interesting, since, to my ear, this is the pair with the greatest discrepancy between the tones. I’m not saying it’s any easier than the other pairs, or that I would have gotten it right. It’s just interesting.

• I probably should have come up with an alternate method of scoring that gave separate credit for distinguishing models from amps, and between various amp types. For example, a couple of contestants realized that 1, 3, 5, and 8 were the models, but got all the amp types wrong. It seems to me they should have received a higher score than 0.

• If this had been a playing test rather than a listening test, there would have been far more perfect scores.

But enough of my yakkin’! Let’s hear from three guitarists with ears far superior to mine:

Grand Prize winner Colm Kelly is currently working on a new studio project called Tiny Telephone Exchange. (You can check out their tracks here.) When pressed, he described his listening techniques: “The biggest tell-tale was the self-noise of the actual amp; this was particularly audible at the end of the clips. I am also reasonably familiar with the microphone used for recording the clips and knew what results could be expected. Otherwise, the Bassman was relatively easy to pick out as it was the only one of the four amps with 10″ speakers. A lot of time 12″ speakers contain low end information that gets EQed out at the time of mixing. Depending on what you are after, a 10″ speaker can sound closer to ‘finished’ straight off the bat.”

Colm says he is not a great fan of modeling: “In my experience, modeling takes a lot more tweaking to get it to do what you want, and the ability to endlessly tweak can lead to indecision. A recording with multiple tracks of modeling is a lot tougher to get to stick together. Referring back to my comment above about amp self-noise, I think a lot of times in recording, that dirt/noise is an important part of getting things to sit together right. I also think that the results that can be achieved with modeling depend heavily on the quality of the DI and conversion that you are using. That said, I am not up to date on the latest modeling software so I can’t say how much it has improved in the last couple of years.”

Second-place winner Ralf Tyra is a self-described “bedroom guitarist” who recently picked up guitar again after not playing for a decade. He says he trained his ear by working with Avid’s Pro Tools and Eleven Rack modeler: “I listen to any separated recorded guitar track I can get my hands on, regardless of the style of music, and try to tweak in the sound as close as I can get on my Eleven Rack. This is probably why I did that well in your contest: Listening to and A/B-ing real recorded amps and modeled amps is currently a great part of my leisure activities. I do not use any special studio technique for this, just my ears.”

Unlike Colm, Ralf is an enthusiastic model user. “I am deeply impressed by the quality of amp modeling nowadays,” he says. “It’s a dream come true to have so many great-sounding amps, cabs, and mics at your disposal, and still have some space in your bedroom.”

Third-placer Jessie Nieboer took a less scientific approach. “I have no clue how I did it,” he shrugs. “I just listened for some of the tell-tale signs of amps, like a little bark from the head and the more direct sounds of a model.” Jessie’s musical pursuits consist of “writing my own stuff, and just messing around playing with my friends.”

Any grand conclusions? Not really. Models are cool.  Amps are cool. Sometimes models sound great, and sometimes they sound crappy, much like amps. We’ll be all over both alternatives here at Tonefiend.

Thanks yet again to everyone who participated! I hope you found it as interesting as I did! 🙂

30 comments to Amp vs. Models Contest: The Winners!

  • Dan

    Well, I was way off. But, I’d figured that I was going to be way, WAY off, so now I don’t feel so bad. Congratulations to the winners!

  • Paul

    Much to my surprise, I identified all the models correctly! But also to my surprise, I confused the Marshall and the Diezel?! So in the end I just got 4 out of 8… 🙁

    But just like Colm Kelly, I could only distinguish models from real amps by listening for room sounds and other noises. I can’t say that one sounded better than the other.

  • thanks for putting on the contest it was fun and educational

  • Robert

    I almost got all the amps and models right except the marshall. I guess I shouldn’t have guessed the models on the style of riff played, that part made all my models wrong…

  • John

    It’s great to see that I got it right on the second submission (which didn’t count in the spirit of the contest). I too used the tell-tale noise to distinguish the real amps from the models. Interestingly enough, my only experience with the Diezel amps was thru models but I was able to pick out that tight low-end thump right away. Great job Joe and to all the winners!

  • Nuno Carmona

    «And the correct answers?


    But D and C were Fender. In the post you say the first two are Voxes (which I also think they are).

    «,the next two were the Fenders». But B and A refer to Vox in the post.


    • joe

      Thanks for catching my mistake, Nuno! I’ve fixed the post accordingly, and it now reads correctly. Yes, the first two examples in the quiz were the Fenders, followed by the Voxes. 🙂

  • PazF

    Hey, I got only one pair wrong…all other were right and it was quite easy to tell the models from the real ones because at the end of each track there was that typical noise that tube amps have that models don’t…Because I have almost all of these amps(except for the dizel one…), I recognized them and the track did match the “style” of the amps…in all of them….great contest, and great exercise for the ears!

  • Vince Daukas

    Great comparison challenge! It would be better if you took out the amp brand, and had more tracks to compare to get to the basic amps vs. models test. Too bad the self-noise is such an aid – maybe self-noise could be sampled from the amp track and added to the model track to wash out that difference.
    I don’t know anything about the pure sounds of different amps, as I use models exclusively and choose tones based on experimenting with mic/amp/cab/fx/room combinations. For fun, I tried to group tracks based on the differences in characteristics seen in spectrum analysis, and guessed as to which were which. My guesses were way off as to what characteristics were what, but after seeing the results, I’d say that the distorted sounds (all but the first two tracks) from amps seemed to have more “sharp” frequency drops or jumps in places across the spectrum. Maybe modeling creates too much “perfection” in this respect?
    Of course, spectrum analysis isn’t a listening test, and I could not say that amps or models sounded better or worse. I think your contest showed that a vast majority of discerning listeners (probably mostly guitar players) can’t really tell the difference in sound between amps and models.
    By the way, which modeling applications did you use?

    • joe

      Thanks for your savvy thoughts, Vince!

      Yeah, the results of visual analysis programs can be…um…interesting. Sometimes they can be incredibly helpful, and sometimes they cough up all sorts of random gobbledy-gook. I often use a sort of spectral analysis tool, Logic’s Match EQ, which is an EQ “cloner” along the lines of Waves’ Q-Clone and a number of other EQ-matching tools. Match EQ lets you analyze two passages of music, and generates an EQ spectrum that aligns the second analyzed passage with the first. But since it’s a high-latency process, I use it more for analysis while designing sounds than for realtime playing.

      The models were created entirely using the plug-ins that come with Logic Pro and Logic Express — no hardware modelers. The signal path within each channel is generally Amp Designer (Logic’s amp modeling plug-in), Channel EQ (Logic’s “generic” EQ plug in) and the program’s basic compressor, imaginatively called Compressor. Real simple! In fact, the modeling channels I used were pretty close to preset sounds in the factory library. (Some of these same amp models appear in GarageBand, and GB is powered by the same audio engine behind Logic. But GB has less sophisticated EQ and compression tools, so while it can sound great, you can’t really get “surgical” with it as you can in Logic.)

      As a Logic developer, I’m biased. I know these tools well, and think they’re pretty darn good. But please don’t take that as a put-down of any other modeling technology.Also, it’s been a year or two since I’ve done a comprehensive survey, and I hear great word-of-mouth about a number of software and hardware modelers I have yet to try for myself.

      Again, my goal wasn’t to trick people, just to provide a reality check about some of our current tools. It never even occurred to me, as some correspondents have suggested, to mix in a bit of amp noise with the models. Heck, it never even occurred to me to listen for it!

      Anyway, it seems pretty obvious to me at this point that the amps and models are close enough that any number of other variables in the music-making process can easily outweigh the differences between these methods. And I can’t help pointing out that approximately 1.5% of contestants could consistently differentiate the two. (My guesstimate is that another 10 percent were usually able to do so.) And this is among a pool of (let’s face it, guys) über-geeky, technically oriented players who love to dwell on this stuff. What would the results have been among musicians in general? Or among (gasp) listeners with no technical know-how?

      We’ve all read so many blanket opinions about models vs. amps, often based on wishful hearing and other sketchy methods. There’s a lot of distance between typical gear-head forum comments (“MIcrochips ain’t got no soul”) to the sorts of statements we’ve read here. (“While scrutinizing the tracks on headphones at 96kHz, I perceived a tiny bit more amp noise on the decay portions of certain audio clips.”)

      It should be obvious that the only metric that really matters is, which tools most inspire you? I get inspired by both amps and models. How about you? 🙂

  • Colm Kelly

    +1 on what Vince said. It would be cool to hear some info on the process you went through in the preparation of the clips.

  • Simon

    Sorry if I overread that, but what equipment did u use for the modeling?
    In our Musicians Board (german board) some said it was a Line6 Pod 500 HD. Is that right?

    Thank you for the cool contest. I did not vote, but it was fun to see what modern 1s and 0s can produce.

  • Colm Kelly

    I found it interesting that a lot of people got the Fender and the Vox mixed up; I think that a lot of times the sound we associate with an amp is as much the speaker as anything else, the Fender clips had real alnico chime not that far removed from what could be expected from a Vox amp.

  • joe

    Yeah, it interested me too. But the Fender and Vox amps were all original, so you’re hearing Eminence 10s on the former.

  • Haha, I was one of the 10 6/8 entries and did it in one sitting, so I strangely didn’t listen for the amp noise at the end. (I got the Fender clips reversed.) Dangit!

    As for inspiration? I find myself more inspired by a great tone, rather than the notion of how it was producted, and when you get your monitors optimized and mixed well, modeling seems far more capable in the end; you can go from copying real amps, yet go beyond that as well, into another realm of tweaked and blended sounds, impossible with actual hardware.

    I find it funny how much resistance to this paradigm h/w amp “connoisseurs” put up! I love it when they’re falling all over themselves revising their appreciation of a sound, once they are told it’s modeling. Hilarious!

    Thanks for a fun and revealing exercise Joe!

  • el reclusa

    Heh…for inspiration, I’m often inspired by particularly BAD and/or “wrong” sounds as much as lovely, muscular tones, “real” or modeled. Interestingly, I noticed a modeling amp in a catalog the other day that purports to let the end user- you- do your own modeling, I’d guess much like creating your own impulses in a convolution reverb. That’s a route I would love to take. I’ve been trying to find a distortion device that matches the ferocious-yet-smooth sound of my old Gibson GA-5 when you keep playing after shutting it off. There’s no standby, so you get the most amazing, saturated (but not very loud) tone as it’s poor little 6V6 is starved for voltage. It’d be nice to see more weird choices in modeling gear…like oilcan delays and broken fuzzes. Y’know, FUN stuff!

    • joe

      Yeah, one not-so-great aspect of modeling is the narrowing of the sonic gene pool to a few “canonic” amps. It’s definitely changed the way I think about buying and using amps. I’m now more inclined to pick up something for its weirdness factor than its classic versatility.

      Love that dying-tube sound, BTW. I got something similar by yanking the AC adapter on a tube-based Flip Tremolo pedal. You can hear it on the solo of the Eels song “Agony.”!/id3538324

      (The solo isn’t included in the short browser-based preview, but if you preview it in iTunes, you get the solo.)

      If you get into building stuff, you’ll find that it’s super-wasy to integrate a dying-battery simulator into just about anything. In some circuits, it makes a big difference. That phenomenon is the basis for all those stories about obsessive/compulsive detail-oriented guitarists like Eric Johnson preferring non-alkiline batteries in their stompboxes.

  • Mario

    Damn, had perfect score right away but then switched diezel with diezel model, oh gosh! 😀 Any sparing pedal for me? 😀

  • Eddie

    WHOOO! I successfully managed to get NONE right! ( :

  • el reclusa

    Yep- I agree. It’d be nice to get models of some more oddball effects as well (oilcan delay springs to mind- I finally plopped down too much money on one a couple of years ago and STILL can’t get the damned thing to work!), but I suppose if the demand isn’t there, well…yeah.

    Good point about being selective about amp purchases, though- I hadn’t thought of that, but it makes very good sense to skip a lot of the “classic” stuff that’s been modeled more than well enough in favor of the…unique. Thanks, Joe! You may’ve just saved future me an awful lot of cash! In all seriousness, though, I’m finding lately that it’s much more productive to use what’s already in front of me than waste time lusting after stuff I probably don’t need, especially when I should be practicing or building things, and to that end, I’m really looking forward to trying my hand at some DiY pedalmaking soon…

    Love that Eels track…Souljacker has been on heavy rotation for me this week, but I somehow missed Shootenanny in the shuffle when it came out.

  • ezcomes

    well…i had the first four right…and then i fell into the Marshall Deizel trap!

    hahaha…wicked work! thanks again!

  • gui

    Im glad you kept your promise not to humilliate anyone on the constest, i would be the clown in the stompbox hahaha

    hats off to your tweaking abilities

  • I’d like to suggest the next competition…. a Wah challenge! See if people can identify the correct wah model through the same amp settings.

Leave a Reply

You can use these HTML tags

<a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <s> <strike> <strong>




Click to upload a JPG

This site uses Akismet to reduce spam. Learn how your comment data is processed.