More Thoughts On Tube Amps v Modelers

László

Too Many Notes
Joined
Apr 26, 2012
Messages
34,607
Location
Michigan
Last night I decided to try out Universal Audio's new Tweed amp model for the Apollo and UAD platform. I have a lot of the UAD plugs, and use the effects processors in my work. Most are very good.

As many here know, I'm not a big fan of amp modelers, so some of my remarks may be attributed to confirmation bias. However, last night I recorded a demo of my new McCarty with this model, to see how it'd work in the context of a track where I might use a Tweed style amp, so I composed something. I posted the link elsewhere, but here it is as well. I think it turned out pretty well:

https://soundcloud.com/lschefman/tweed-model-mccarty-4

Frankly, if you listen to this recording, you might think it's a real amp, you might not. I think the outcome is pretty close, but not 100%. Warmth is missing. However I could get away with this recording on an ad, no problem.

I did have a hard time getting the amp to respond the way my tube amps respond. It felt noticeably stiffer. It didn't respond quite as naturally to the pick attack, and I found myself stumbling a little more than usual (stumbling when playing is pretty much the name of my game!).

Nonetheless, a useful track.

But you know, there's nothing like a listening comparison.

So take another listen to that modeler demo, and listen to things like how the notes bloom and sustain, whether they sound rich or thin, soft or hard, etc.

Now here's a link to a recording I did with the DG30 with the McCarty Singlecut, an amp I think falls nicely into the Tweed style camp, and as you can hear, there's a nice warmth, fullness, and looseness to the amp that really shines through compared to the model; it's also much more easy to play through, the notes sing. And it has that elusive elasticity that I think the models lack:

https://soundcloud.com/lschefman/hammer-dg-mstr-2

And something also in the Americana vein with the maple neck McCarty into a Lone Star, a stiffer amp than the DG30, but still a joy to play through compared to the model, and probably a better tube vs modeler comparison as both employ the bridge pickup on a McCarty:

https://soundcloud.com/lschefman/americana-ls-guitar-v2

Listen, I'm no paragon of guitar skills, I realize that. Other players may feel differently, but listening to those modelers vs the real amps, something is still missing, and playing through the Tweed model didn't give me the colors I get with the tube amps. The model also wouldn't sing on sustained notes the way a tube amp does, even when it's in a recording booth and I'm in a control room.

I didn't include any HXDA clips because it's unfair to compare a Marshall style amp to a Fender-ish amp. I used only the SM57 modeled mic with the modeled amp, and both real tube amp recordings were done with SM57s against the grille.

Hopefully when you hear the real amps in my clips, you can hear the differences between them and the amp model in the first clip.

This isn't to say the model is useless - I'd use it on a track in a pinch. But for me it falls less into the realm of "inspiring instrument" and is more in the category of "workable tool."
 
Last edited:
My personal experience has been that I find a modeller coupled with outboard analogue effects pleases me best. I use a Kemper with fairly low gain settings and then a good quality 'transparent' overdrive pedal in front of it (e.g. Bogner Wessex/Klon clone etc). To my ears I get the best of both worlds; good playing dynamics and great amp tone straight into my PC.
 
My personal experience has been that I find a modeller coupled with outboard analogue effects pleases me best. I use a Kemper with fairly low gain settings and then a good quality 'transparent' overdrive pedal in front of it (e.g. Bogner Wessex/Klon clone etc). To my ears I get the best of both worlds; good playing dynamics and great amp tone straight into my PC.

I've done that, and find it helps.
 
Hopefully when you hear the real amps in my clips, you can hear the differences between them and the amp model in the first clip.

This isn't to say the model is useless - I'd use it on a track in a pinch. But for me it falls less into the realm of "inspiring instrument" and is more in the category of "workable tool."

I can certainly hear the difference - there is a richness and depth to the tone on the real amps that the model misses. On the other hand, the model is a fairly distinctly different tone from the two amps - and this is the point I made in the other thread. A modeled amp is great when none of your real amps can make the tone you hear in your head. In my experience, hearing a mix later on, no one but you will realize it was a modeler - probably not even other guitarists.
So your assessment of it as a "workable tool" is spot on.

If a painter has the best oil paints available, but he is out of red and only has a can of Krylon red, does he do without red? or use what's available? (OK all the stores in the world are out of the red also).

Tom
 
In my experience, playing modelers can sound very good. As good as a real amp, sometimes even better IMO. But not exactly the same as a real amp. They tend to get a bit more "idealized" tone, if that makes sense. But often it's the non-ideal stuff that makes the amps sound good.

Much more than the tone, what I notice is that I find the dynamics such that I can generally get the tone and dynamics I want, but I have to work harder to get them. Modelers tire me more quickly than a real amp.
 
If a painter has the best oil paints available, but he is out of red and only has a can of Krylon red, does he do without red? or use what's available? (OK all the stores in the world are out of the red also).

My brother's a real painter who heads up the fine arts department at a college, so I asked. He said he'd get out the linseed oil and beeswax as used back in the day when artists made their own colors, add some vermilion (natural cinnabar pigment), and mix his own paint.

He's kind of a stickler for detail...he wouldn't use Krylon on an oil painting. It wouldn't blend properly. Here's one of his works, so you can see the kind of thing he paints:

1422647951_Robert-Schefman.jpg
 
In my experience, playing modelers can sound very good. As good as a real amp, sometimes even better IMO.

I don't agree.

I may prefer a model in a given track because it fits the song better than the amp I have on hand; however., if I have model of a particular amp, and play it against the real one, I have yet to come across a situation where I didn't prefer the real version of that amp.

This based on 28 years of studio experience, and many, many models and modelers, going back to the days when things like the Sans Amp, and others were out there before digital modelers.

I really would love to have a modeler that does it right, but so far, I haven't found one.

The Kemper is the closest.
 
Last edited:
I think the difference here is that you generally prefer the sound of a real amp. For me, sometimes the modeler sounds better to me for some reason or another. I don't think the modeler can IMITATE the amp accurately. Quite the opposite, so far. But I don't necessarily hold the amp as the standard by which I judge the modeler.
 
Incidentally, the only paid session work I've ever done was for a video game developer. We recorded my guitar parts through a distortion pedal and a Marshall Micro-Stack (that little 9V amp) and an SM57. Sounded surprisingly good after the studio magic.
 
Incidentally, the only paid session work I've ever done was for a video game developer. We recorded my guitar parts through a distortion pedal and a Marshall Micro-Stack (that little 9V amp) and an SM57. Sounded surprisingly good after the studio magic.

Ha! I've done that with a mini-amp, too!

Sounds a little weird but in a fun way. :)
 
I think once you record it, the differences become much less audible, especially to anyone else who weren't there when you recorded it. It has more to do with responsiveness. I mean, you can't stand in front of the speaker and have the guitar feedback on itself with a plugin.

And I'm not going to downplay the importance of responsiveness -- if you're getting a great tone, and you're getting into it, I think that comes across in the music, which is the most important thing.
 
My brother's a real painter who heads up the fine arts department at a college, so I asked. He said he'd get out the linseed oil and beeswax as used back in the day when artists made their own colors, add some vermilion (natural cinnabar pigment), and mix his own paint.

He's kind of a stickler for detail...he wouldn't use Krylon on an oil painting. It wouldn't blend properly. Here's one of his works, so you can see the kind of thing he paints:

1422647951_Robert-Schefman.jpg
That is amazing!

"Kind of a stickler for detail"...... run in the family? :D
 
New mix posted above and here --

https://soundcloud.com/lschefman/tweed-model-mccarty-4

I changed the EQ and drums a bit, also the keys part is changed. I think it hangs together a bit better now.

I guess it's never a great idea to mix down a track and not come back to it a few times, because there's usually room for improvement, but I'm so used to working fast that I rarely do what sane people do.

However, now I am. ;)
 
Last edited:
sticker for details? Perfect nipple placement in that painting and it looks so realistic it is amazing.

My thought is that the modelers like fractal and Kemper make it so easy to develop sounds in a live environment but in the studio, the musician would go with more traditional effects?
 
sticker for details? Perfect nipple placement in that painting and it looks so realistic it is amazing.

My thought is that the modelers like fractal and Kemper make it so easy to develop sounds in a live environment but in the studio, the musician would go with more traditional effects?

Boobs are wonderful things, and my brother doesn't miss painting the pieces/parts. ;)

As to playing out with a modeler, I think it'd depend on the player. I find I don't play as well with them, so I doubt I'd use one, even live. But there are still a few new ones out there I haven't tried, and the new Helix is one of them.

I was recently talking with my son about his upcoming tour to the UK, and he told me that the band was renting a backline. All they're taking are their instruments and pedalboards.

He told me that for a major tour, cartage can be six figure money, but even for a week in the UK for small venues, the figure was so high it was out of the question! So for something like that, I could see a modeler. Who wants to throw money away on cartage to ship amps? Of course, you'd still need to ship the modelers, and you'd need a backup, so you'd have a rack of some kind regardless.

Still, he won't do shows with a modeler at this point, so for him, it's academic.
 
Is it reasonable to try the amp model through the same speaker cab and mic? That would eliminate the real versus emulated speaker + mic differences. And you could experience the same speaker + room for playing dynamics. I guess the only uncolored way would use a high power linear solid state power amp with lots of headroom.
 
Is it reasonable to try the amp model through the same speaker cab and mic?

Yes, and I've actually tried this with amp models that allow you to separate the model of the amp from the model of the speaker cab (you don't want to hear the speaker cab model on top of the sound of a real speaker cab).

I didn't think the results were satisfactory - whether that's due to the way a linear transistor amp "sees" a speaker load of varying impedance differently from a tube amp may be a factor - and at that point, it's just a whole lot more sensible to mic the real amp, because what's the dang point of modeling if you have to mic up a cab?

What was more satisfactory for me over the past few years has been using a real amp, and a load box, and running the load box signal into an impulse response model of the speaker and mic. I have some of the Ownhammer and Red Wirez IRs, and I feel these are an improvement over doing things the other way around. In fact, this is the way Pete Thorn records most of his gear demos now. He will use mainly a real amp, with a load box and IRs and not his Axe-FX, Kemper or Helix, unless he's demoing those products.

In other words, in my opinion you get a better result with a real amp into an IR-emulated cab and mic, than you do with a modeled amp into a real cab and mic.

Why this should be so is above my pay grade, I just know it's so.

For my work, however, I don't think it's all that hard to just stick a 57 in front of a cab and hit "record."
 
Back
Top