"Real amps" vs Modelers

brainleaf

brainleaf

Member
I recently watched "In the Room with John Petrucci, Tosin Abasi, and Devin Townsend" on Rick Beato's YouTube channel. Around minute 34, Rick Beato asks what real amps can do that modelers can't. Here is the link

I found the responses pretty interesting. I get the feeling that they are very biased toward using real amps, but I was curious if there is truth to what they're saying. Essentially the argument is that Tube Amps introduce a "randomness" that modelers don't have, and that IRs are simply a "Single instance in time". I don't have any experience playing tube amps, so I can't speak to the accuracy.

I do get the feeling that they are way more experienced playing real amps than using modelers (maybe with the exception of Devin Townsend), and so they are biased toward their preferred rigs. What do you think?

Edit: Regarding Petrucci claiming the Neural DSP "Does it the best" holds very little weight IMO since he has a product with them. I can't trust him to have an unbiased opinion on that point. Also just to clarify, these are three of my favorite guitarists of all time.
 
this discussion is as old as time by now.

the solution for not liking IRs is to use a real power amp and guitar cab.

for many, the randomness that tube amps give is some tubes don't work sometimes, the amp doesn't turn on, it's too cold for the tubes to warm up properly, etc. a bit of a joke, but real as well.

no one is wrong here. we can all prefer what we want, and through our experiences, develop those preferences and stand by them.

amps are great. but they're also loud and not everyone can take amps to the gig or use them at home.

also when someone says "does it the best" it always means "for them." he worked directly with Neural and kept having them make changes until it was exactly what he wanted. Fractal tends to model things as close as possible to the real amp, not to personal preferences of a single player.

so again, there are no wrong answers with gear if you get the result you want.
 
I get all that, and I agree. I was specifically wondering if what they are saying is objectively true. As in, is there truly a randomness in the response of tube amps? What is the cause, and do modelers try to replicate it? And are IRs truly "a moment in time" captured? And if so, what does that actually mean in the result of the tone?
 
brainleaf said:
Edit: Regarding Petrucci claiming the Neural DSP "Does it the best" holds very little weight IMO since he has a product with them. I can't trust him to have an unbiased opinion on that point.
Click to expand...
Looking at the Artist page of fractals main site I see that both Petrucci and Townsend are listed. Do you take that into account too when weighting how much you want to trust in their answers?
 
Old_Soul said:
Looking at the Artist page of fractals main site I see that both Petrucci and Townsend are listed. Do you take that into account too when weighting how much you want to trust in their answers?
Click to expand...
Petrucci only uses the AxeFX for effects as far as I know. They are talking about amp modeling here. I believe Townsend uses modelers more extensively. And to be clear, I do not distrust them at all, nor do I think they are wrong. I am asking about the objective statements because I just don't have the experience of using tube amps.
 
brainleaf said:
I get all that, and I agree. I was specifically wondering if what they are saying is objectively true. As in, is there truly a randomness in the response of tube amps? What is the cause, and do modelers try to replicate it? And are IRs truly "a moment in time" captured? And if so, what does that actually mean in the result of the tone?
Click to expand...
i think there can be randomness in analog hardware. but this usually isn't desired because it's not consistent.

without them expounding on exactly what randomness they like, we can only guess and even so, that depends on their amp and process.

IRs are a moment in time - the time it takes to capture it. again, i'm not sure exactly what randomness they might like with a real cab in a room and a mic on it where nothing is moving around to change things.

they may be discussing what i said above, little "mistakes" that happen along the way due to the analog components that create a random sound they didn't expect. maybe just that unexpectedness is something valuable to them.

the current generation of modelers are very dynamic and respond like a real amp would in most cases. so i doubt it's the basics of what modeling sounds like vs real amps. that's been talked to death as well.

i think Tosin mentioned in some other interview that they used modeling for the demos of their latest album, but then tried real amps in a iso room and it was just more immediate, had more low end, and easier to control for the recording. this could have been just in that moment in time, that mindset, the studio setup, etc. and it could have been that their modeling preset just didn't have that much low end, but the real amp for whatever reason did.

remember that real amps have been used on recordings for many many years by now, and some people are just used to adjusting that vs a modeler. maybe the engineer could just move the mic more easily and they thought that sound was great.

the best way to learn about this is to try a real amp, and turn it up loud. there is something about that feeling for sure. and it is mostly all about the real guitar cabinet producing sound the way it does compared to a full range speaker or near field monitor.

also remember that often we have a tone we love right now, but come back to the same exact thing tomorrow and dislike it.

real guitar cabs do something different that full range speakers. real amps have fewer controls, which means you have to settle for something quicker due to less options. perhaps that's the only reason. the phrase "amp modeling" almost always includes the guitar cabinet too, a very important part of the equation.
 
brainleaf said:
randomness
Click to expand...
brainleaf said:
Single instance in time
Click to expand...
Meaningless phrases created to justify a preference.

Anyone can do their own apples to apples A/B using the looper in the fractal and either a reactive load box with an amp, or good power amp with the fractal. As long as the cab+mic+monitoring setup is held constant the accuracy of the fractal models is obvious. Use what works for you.
 
Being famous and among the best guitar players on earth doesn't automatically make someone unbiased, expert in everything guitar-related and have superhuman ears. Those kind of opinions, for me, have pretty much the same value regardless of who they come from, i.e. close to zero. Full stop
 
brainleaf said:
Tube Amps introduce a "randomness" that modelers don't
Click to expand...
I alway took it (maybe wrongly) that the "non-linearity" of distortion modelling (ie Drive block, amp block preamp, amp block power amp, ...) is somewhat synonymous with the "random" nature of real tube amp circuits (I can't calculate the exact precise result of amp block output content even if I know exactly what went to the that amp block's input).
 
sprint said:
I alway took it (maybe wrongly) that the "non-linearity" of distortion modelling (ie Drive block, amp block preamp, amp block power amp, ...) is somewhat synonymous with the "random" nature of real tube amp circuits (I can't calculate the exact precise result of amp block output content even if I know exactly what went to the that amp block's input).
Click to expand...
Non-linear simply means that the output changes depending on the input level, nothing to do with randomness. Tube amps certainly have some randomness, in fact if you reamp twice the same track with the same amp and do a null test, the two tracks won't cancel out completely, but I think most of that randomness comes from noise, interferences, fluctuations in power and temperature and that kind of environmental stuff, not from some intrinsic property of tube amps.
Btw also Fractal models have some randomness, in a null test they don't cancel out either... unless you set Power Supply Type to "DC" ;)
 
I broke my tube amp out yesterday for the 1st time in over a year. The amp was a Carvin X100B from 82. The chassis and transformers are original but everything else is different. It’s been a few things but right now it’s a 50 watt 1987ish amp. It sounds great and it was a lot of fun to play through it. After using my FM3 almost exclusively for a couple years, I was listening carefully to the artifacts, especially on the decay. All I can say is the FM3 does an amazing job. All the weird fizz and crackle is there on the amp and the modeler. The feel and dynamics are super close, but the real amp has more punch. It’s very close really. And it was good to plug back into the modeler just for the ease and availability of adjustments and the ton of variety. I can see why pros would stick with real amps. They have techs to move rigs around and maintain everything. They have their sound already dialed in and why not use the real thing? I’d bet if someone like Cooper Carter set up their ‘rig’ and blind tested they would have tough time hearing the difference. Especially if through an amp and the same cabs.
 
brainleaf said:
I recently watched "In the Room with John Petrucci, Tosin Abasi, and Devin Townsend" on Rick Beato's YouTube channel. Around minute 34, Rick Beato asks what real amps can do that modelers can't. Here is the link

I found the responses pretty interesting. I get the feeling that they are very biased toward using real amps, but I was curious if there is truth to what they're saying. Essentially the argument is that Tube Amps introduce a "randomness" that modelers don't have, and that IRs are simply a "Single instance in time". I don't have any experience playing tube amps, so I can't speak to the accuracy.

I do get the feeling that they are way more experienced playing real amps than using modelers (maybe with the exception of Devin Townsend), and so they are biased toward their preferred rigs. What do you think?

Edit: Regarding Petrucci claiming the Neural DSP "Does it the best" holds very little weight IMO since he has a product with them. I can't trust him to have an unbiased opinion on that point. Also just to clarify, these are three of my favorite guitarists of all time.
Click to expand...


SIDE NOTE: There's a recording below if you don't want to read anymore about this.

I just listened to John Petrucci speak for about 40 seconds, essentially "the biggest challenge for modellers is not the gain, it's how the amp physically moves air through a cab." I agree. I have been busting my balls for over a year trying to get what I call, the percussiveness, or bounce, or something out of the Axe, it was driving me effing bonkers.

IRs are simply a "Single instance in time" or whatever, I have no idea, I just know they need work. They're not dynamic, so they sound good at one spot on the guitar, and then lack in the other areas.

The amps are great, John Petrucci says Neural DSP 'does it the best' ... referring to what? the simulation of an amp and cab? No way, I've tried every Sim out there, the Axe 3 Amps are in a completely different league.

The problem however, is still, IR's.
chris said:
the solution for not liking IRs is to use a real power amp and guitar cab.
Click to expand...
Yup.

This clip is from today, at first it's completely dry and raw, and then I turned on the reverb, it's just a cheap ( $100 ) mic on a cheap Marshall Valve-state, I plug into the CD in on the Valve-state because it's neutral, it bypasses the VS's preamp, which is awful. So I use the Axe 3 for an amp, out to the Marshall Valvestate, with is mic'd by an Apex 435 mic, which goes into an Axe I/O interface ( because the Apex needs 48 volts to operate ) then out of the interface and back into the Axe 3. The feel is superior to static IR's, and I have the volume super-low because I live in an apartment. It's more percussive everywhere on the neck, and the chord voicings are clear, also consider how much high-end there is in this clip and the speaker handles it nicely, with an IR this would sound nasty. The punch of each individual note when I play a lead line is so pronounced and even, also, when I turn the reverb on the punch remains. Often when using IRs I could dial in an okayish tone, turn on the reverb and it would die and I'd have to EQ things all over again.

 
Last edited:
Dig around Cliff's posts over the last 15 years, a lot of that randomness can be accounted for and is precisely why so many go to Fractal.

I've heard people gripe about modelers sounding too perfect and I never understood until recently. It's hard NOT to dial them into a 'perfect' state because the options are there. Some woofy low end on certain notes in an otherwise great sounding preset? Dial it out. A little fizz on the top that's just kinda gettin' to ya sometimes? Dial it out. Why not, the options are there and it's easy enough!

But these are the exact things about tube amps that could also fall under the 'randomness' umbrella. It's hard to unhear the idealized version of your tone in your head, but chances are, the idealized tone in your head is playing along with some other instruments that are going to have a big affect on how that tones ends up sounding, so if you START with the idealized version, by the time you had everything else in, you've now moved away from your idealized version and most certainly further away from the sound of a 'natural tube amp' on a recording.

There's definitely a lot of recordings as of late where I can tell they're using a modeler because the way it sounds. It's not that it's bad in any way, it's just the progression of guitar tones with the technology available at the time. Surely back in the 90's there were dudes saying "That's not a Marshall with an SD-1 in front, that either a Peavey or a Mesa"

I also suspect that's why you hear some people/bands having tracks re-amped through actual amps on albums, Periphery and Tosin Abasi are known for this, it's hard to dial presets back to being raw and 'unruly' like a tube amp after you've spent so much time making them perfect.
 
You must log in or register to reply here.
Back
Top Bottom