Axe-FX II

[PowerTube]

IMO They need to release AXE-FX Lite. Keep the USB of the II, Make it half as powerful as the II. Use less expensive DACs since the USB will allow straight digital full quality output. Cut out some of the repetitive amp models. Sell it for around $1250. You hit that price point and you will have more hobbyist recording guys biting on it.

Years ago, when my wife and I were financially strapped, I had a major lust for either a POD 2.0 or a Flextone, neither of which I could afford at the time. I always dreamed that Line 6 would make a pedal with just the Modern Hi-Gain model, or maybe mix it with a few others. They finally came out with the distortion pedal, but it modeled distortion boxes instead of amps.

But yeah, there's a market for people who don't want or need every possible bell or whistle.

[Cliff]

@marcus - I've been wondering you'd go about modelling a tube in software. Do you have any info or references for how it might be done? Do you know what the differences are between the Axe-fx approach vs the line 6 one?

[Hellbat]

My understanding Cliff, is the initial Line 6 technology basically models by taking a series of "snapshots" of the output of a real amp as they feed different frequencies and such through it. Then they basically interpolate the results to make the output of the model resemble the output of the real amp as closely as possible. So the quality of the modeling boils down to the quality of the sampling and the number of samples.

Axe-FX method takes it more granular. They model the chain of components throughout the amp individually. ie: If you feed a certain signal into one side of a capacitor you will get a certain different signal out the other side. Since you are modeling the individual components, tweaking the values of the components or the signals going through them will react in a much more natural way than just a bunch of input/output snapshots.

As far as I know, the Line 6 HD series modeling is moving in this direction of modeling as well.

[VitaminG]

I will be able to get an AXEFX cheap soon because all the fan boys will be tripping over each other to sell theirs, flooding the market and dropping the price...

but why would you, since you've already determined that they don't cut it? Even cheap (unless ridiculously would-be-a-crime-NOT-to-buy-it cheap), that seems like you're wasting your money.

[Trem]

I agree with alot you say bro. It would be very interesting to find out the results of a blind listening test. I do know that the clips of the Axe FX on Youtube are absolutely incredible but I have heard alot of modeling rigs that sound great online. I am still waiting to hear an Axe FX live sharing the stage with something like a Powerball or a 101B or a Rectifier or an SLO100. I know its old technology but.. I had several Vetta's. I had a Vetta full stack. It was so loud it would literally crack a sheetrock wall. I really knew how to dial it in too. I had sounds that were the best I had ever heard. I brought it to band rehearsal and spent 5 hours dialing it in with the band before the other guitar player got there. All the guys were digging it bigtime. It was tone to die for. Then.. my bandmate showed with his Mesa Dual Rectifier full stack.
It took him 5 minutes to dial in his tone and my guitar tone literally disappeared in the mix as soon as he started playing. The only thing you could hear was his tone. He crushed my tone dead. That was it for me using modeling live when in a 2 guitar metal band.

Well my experience with it was this. At practice when i used the Axe fx i ran thru a Ada microtube 200 and a Single 4x12 300 watt cab not miked (no power amp or cab simulation) with a second line out built into the patch with mic poweramp and cab simulation ran to mixing board for PA against a Peavey Ultra 120 Full stack miked for PA and i could Totally Drown Him Out. When i first plugged in i had the same channel on the mixing board as my mic ran into when i had my peavey miked. When i hit that first chord with the axe it scared the shit out of everyone. I'm not exaggerating at all, it was incredibly loud. Way louder than the peavey was into that same channel.

We practiced pretty loud since we were in a warehouse in an industrial district no noise problem. I mean loud .. gig loud. The peaveys were on 4 1/2 to 5 on the master (when i used one prior) which they really can't go any louder usefully they just over saturate after that. All miked into the PA with 18" subs, 2x 1200 watt pa cabs with dual 15's each, and horns. Also with full miked drums and bass. My rig had no problem moving air on that 300 watt cab either,.. fact the microtube is cleaner and clearer so it could go louder and stay clean and just add volume. I love that thing too. Perfect little amp for the Axe.

I can say without a doubt it could keep up with a 100 watt or 120 watt tube head and cut through as well as rattle your nut sack and sound great doing it. Without a doubt.

[javert]

Marcus is right. There is nothing that can keep us from making digital signal processing models as good as we want. It's just a matter of having fast enough processors, sampling at a high-enough rate and to use enough bits per sample, i.e., it's mainly a matter of whether you are willing to pay for the equipment it takes. The guy behind the Axe FX, Cliff Chase, designed high-end imaging sonars before forming Fractal Audio, so he basically has a signal processing background, and from an interview I read, it does sound like they go about it by modeling the exact circuits (that's what I would do too if I were to do it right) rather than by using the black box approach that Line 6 and others have used in the past. But you have to realize that they did it this way to make the stuff fast and cheap. Cliff says that the Axe FX was only made possible by some of the new DSPs.

[khabibissell]

but why would you, since you've already determined that they don't cut it? Even cheap (unless ridiculously would-be-a-crime-NOT-to-buy-it cheap), that seems like you're wasting your money.

I wouldn't for that exact reason... I'd be wasting my money

[Shawn Lutz]

there were 30 Axe-FX's on eBay the day the news brok...two before. I havent checked in a while so I dont know how many are being dumped but they prices should fall.

[Animus]

There's a shitload of them on ebay now. THe prices will fall soon once all the people that were on waiting lists and who wanted one buy out the used stock.

[Cliff]

Hey Hellbat,
The way you describe the Line 6 technology is how a friend and I were thinking it might be done. I guess it's the audio equivalent of facial motion capture in some video games. I'd not thought of the second way before.
With the samples and interpolation approach it seems like there's a large parameter space that has to be sampled, especially if you include all the tone controls, eq, etc so as to behave the same as on the original model. I think you'd have to do this in the frequency domain. One thing I wonder with tubes. We know they're not linear - you put in a sine wave and saturate it, and you get new harmonics coming out. So I wonder if there's some sort of cross talk between frequencies - is playing one signal with a sine wave and two of its octaves together through a saturated amp going to sound different than playing each of the three tones separately (adjusting volume to get the same overall energy going into the amp each time), then adding the results back together. If the answer is no, I think this'd be hard to model with a sampling and interpolation approach.
Modeling at the component level - I wonder if this means modeling in the time domain, so at every time sample you're saying, here's the input voltage to this tube, what's the output voltage - then send that answer on to the next component. Then you rely on how well the components are modeled - we're not actually modeling electrons, so there's going to be some mathematical or sample-based description of how each component transforms the signal. Not sure about DSPs, but a modern CPU clocks in at less than 4GHz, which leaves around 100,000 cycles per sample to process if your input is at 40KHz. So there's going to be some limitation on the sophistication of the models used. Not that I'm saying I'd be able to tell the difference.
Isn't it funny, though, when you think the transistor was a replacement for the tube, and now were using - what, millions? - to emulate what a handful of tubes is doing? Seems there's probably a few cool new sound processing tricks to think of.

[javert]

Cliff, Hellbat and others, there are some scientific papers about how to do these kinds of things, and they're not that difficult to read. I can't find the one I'm thinking of right now, but I have it in my office and can provide a link tomorrow.

Cliff, you're right about the limitation due to the speed of the processors. It's not about whether we can do it mathematically - we can. It's about whether we can process it fast enough (or how accurately we can do it for a given processor).

[Jack Napalm]

Cliff, Hellbat and others, there are some scientific papers about how to do these kinds of things, and they're not that difficult to read. I can't find the one I'm thinking of right now, but I have it in my office and can provide a link tomorrow.

Cliff, you're right about the limitation due to the speed of the processors. It's not about whether we can do it mathematically - we can. It's about whether we can process it fast enough (or how accurately we can do it for a given processor).

From what I have read that is the key behind the AxeFx. By using the best DSP you can do the math faster than a modern CPU can. That was they key. With the Axe II a way was found to bridge two DSP's with no noticeable latency. Thus having one dsp doing all amp sims and one doing effects.

[Chief_CrazyTalk]

How the mighty hath fallen, lol.

[Trypp Powell]

At Fractal's online store, standards are listing at 1299 - so it looks like those are going to be sold for at least a while. I wouldn't be surprised to see the standards at that price and the ultra's at 1799 in the next few weeks. Used prices on those units should then drop to $900 and $1500.

For those of you who have wanted an ultra or standard, the wait may allow you to save a few hundred more dollars.

[Hellraiser6502]

Bought a (like new) used ULTRA today...$1500.

Still getting to grips with it but it sure sounds GREAT!