Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[John Roberts {JR}]

I guess these are the things I'm talking about John. You can get close, but certain things just can't be modeled and sound like the original. I hate it when the Chinese take a chip and replicate it,

Perhaps better than these chips being completely unobtanium.

and I hate it when the tolerances for older chips are "tightened".

Don't see how that is bad.

It takes away the character of the original much in the way a 1% resistor doesn't react the same as a 10% old school carbon comp. It's all these unknown variables that add to the character of older hardware.

There is thread right now on a tweaker forum about resistors. Some cheap low quality resistors have a voltage coefficient that causes the resistance to change with voltage . Generally not noticeable for modest voltage levels, but for large voltage swings like feedback for a high power amp, or perhaps inside a tube amp with lots of signal swing it can cause measurable distortion.

I didn't know you were that involved with BBD, I'm impressed, especially since I never really got deep into the world of IC. The link below will take you to a good schematic of the AD-9. Can we make it better?

http://www.dirk-hendrik.com/Ibanez_ad9_analog_delay.pdf

Let me count the ways... 

I'd have to charge consulting rates to give a rigorous review but quickly in passing.

#1 U1A looks like a simple pre-emphasis circuit. Like FM broadcast it boosts the high frequency going into the delay path, then rolls them off after the delay, to effectively scrub some HF energy out of the noise floor (I've done this too). The de-emphasis is applied in U1B.

#2 U2a is the compressor half of a companding NR so input signal is compressed 2:1 before the delay, then expanded 1:2 by U2B after the delay. In an ideal world this expands the noise floor down 1:2 and doubles head room by compressing the input.  The Ne571 is actually the cheaper version of ne570 with relaxed specs, so just dropping a ne570 in there would be better.

The primary time constants for the compressor/expander is  set by C17 (C30). It looks like 22uF which seems very slow to me (I'd probably use 1/10 that value but they may keep it slow to introduce some transient distortion and perhaps mask tracking errors that would be more noticeable if faster.)

I notice that they short the gain element input to the side chain input . Pins 2 and 3. I would typically add a low pass filter to the side chain input to reduce errors from clock leakage and high frequency response errors in the expander decoding. The encode and decode side chain want to see the exact same side chain signal, but that compressor is not only seeing a wide band signal, but wide band plus pre-emphasis  so a very hot signal. The decode side chain has input and output filters changing the frequency response not to mention some output clock leakage. Putting a similar LPF in series with both of these side chains, will reduce errors caused by different HF response. As connected frequency response errors in the delay path from all those LP filters get expanded 1:2. 

There are many more tricks available to companding NR but this is the cheap tour.

#3 the MN3205 is an old Matshusita chip, with it's dedicated companion clock driver  MN3102. I used a smaller matshusita BBD in my old kit article, but used better IMO Recticon parts in my studio and consumer delay lines. (Good luck finding Reticon parts today).

You can adjust R27 (input DC bias) for symmetrical clipping.

You can adjust R17 for minimum clock noise. The BBD only passes one sample per half clock, so the output has a one stage sample and hold source follower to repeat each sample, for the second half clock, so there is a continuous output. These two mosfet source followers have different Vgs, so the trim balances them out to reduce a clock component from any DC error between them.  This clock noise is generally harmless but can cause tracking errors in the expander, or be audible if clock frequency is low enough.

#4 T3 and T4 are each two pole LPF to anti-image (smooth) the sampled waveform.

I don't see a discernible method to the filter alignments, but I'm not going to do the math.  Looks light on anti-alias input filters and heavy (normal) on output side. But feed from a guitar pick-up may not go very HF. 

The very slow companding time constants may be to conceal tracking errors that I expect to be present because expander side chain gets more filtering than compressor side chain.     

So in conclusion I see several things I would do differently but that doesn't mean you would prefer my more accurate version.. it's a guitar pedal after all so they probably designed a bunch of it by ear.

Back in the '70s I actually wrote my own computer program to help me design complex mutli-stage filters for BBD delay lines. Since this was the early days for computer aided design, i used the tab function on my dot matrix printer to make crude response plots.

I could write a short story about filter alignment tricks but I won't waste the bandwidth, for an audience that probably did even follow along this far. 

JR   

[Bob Leonard]

I'll trade you a box of clown caps for an overhaul.

[Stephen Kirby]

That may have been true in the past, but not so with the Axe-FX. It can be as dynamic and expressive as you like, but to get it to match a regular amp takes some tweaking.
It comes down to whether the convenience of having an infinite number of amps and fx in a 2u rackspace is worth the (possible) fiddling required to get the tone you're looking for. If you typically use 1 amp and a pedalboard for all your songs and don't mind dragging it around, then there'd be no reason at all to buy an Axe-FX. Just like the guy mixing just one band on an analog board with an outboard FX rack. If he doesn't need the size and flexibility advantages of a digital console, then why buy it?

I'd agree that the current stuff is far beyond Line 6 kidney beans.  I'm curious as to what tweaking can be done to make them feel like a good conventional amp.  Typically when confronted with one, I turn down or off all the effects and turn down the gain from the laser shred presets.  But while there is a noticeable dynamic response and you can get more or less distortion with the pick attack, I still can't hear my fingers the way I can with a good conventional amp.  I do know a fair number of folks who believe that the tone is in the gear and/or that they can buy tone.  Maybe I was fortunate to grow up playing really garbage stuff.  If I was going to get any sort of sound similar to my heros, I was going to have to fight to force it out of the gear.  So when I got to the point of playing though good gear and getting stuff of my own, I had pretty expressive fingers.  As a sidebar, the guy who used to run GC's west coast guitar division once ordered up a $5000 custom Two Rock largely on my ability to make a demo unit sound like Robben Ford was playing through it.  He couldn't get that sound, but figured if I could, it must be a good amp.  Conversely, I've also played though Robben's Dumble and the feel was entirely different.  And you would have to play it completely differently to try to make it sound like him (which I definitely wasn't going to do in front of him )
Arthritis kind of put a damper on regular gigging but I still play from time to time.  And I still have folks stick a guitar into my hands asking me to make it sound like this or that person.  To the degree that a rig gives me a big pallet, is how I judge it's responsiveness.  Something that just isn't there with the modelers.  You can tweak knobs to get something more like this or that, but you can't just go from one to the other with your hands.
If you've ever been to a Greg Koch clinic, that kind of variation of tone is what I'm talking about.  I could never play like Greg, or get such a range of tones from the pedestrian gear he uses (which is why Fender hires him to walk into a store grab a Tele or Strat off the rack, plug it into any old Hot Rod something, and make all of us consider bonfires) but it's something to aspire to.

[Jay Marr]

I . Lets imagine a $3.5k modeler ( the basic retail price of the Axe FX II XL+ ), an amp which could set you back another $600+ and a speaker cab which for a decent one is yet another $600+.

The Axe FX 2 costs $2500 new.  Less than 2k used, and mint.  Just want to make sure we're working off of facts.

I own the Axe FX (as well as a PILE of tube amps...yes, I'm a junkie) and I can say (and almost every Axe owner says the same thing), that it is hands down one of the best gear purchase I've ever made.
The price is scary when compared to other modelers, which is why lots of folks are skeptical.

[Luke Geis]

It appears they are having a 10% sale right now on it and that the price is listed as 2.5k for the modeler only. The controller that is specifically for the Axe FX is about $800. I saw another site earlier that was charging 3.5k and listed the retail price as being 4k. I have not seen any used XL+ versions yet. Reverb.com seems to be ripping people off with the prices they are showing.........

Don't get me wrong the Axe FX is a great unit. I would also put it down as being the best purchase ever, not because it is the bee's knee's, but because the tech is probably only limited by what I can do to it and because my pocket book says I better love it.

Ask a person who has shelled out the coin for a Soldano SLO-100 if it is the best purchase they ever made? The answer by those that truly love that sound will say hell yeah, some will say I like it, but wish I didn't have to spend 3.5k to get it and the rest will say that they sold it because it was not their flavor. Why is that? It's considered by many to be one of the best high gain amps you can get and it does pretty much what you buy it for as good as it can be done. I think it's because it is pretty much a one trick pony and a user can ask for more sounds than it is truly able to produce. In other words, it is not able to out tech the user. If the amp could produce more sounds than the user could create, it would be the best amp ever.

Contrast that to an Axe FX and you have two different beasts. The model of the SLO-100 in the Axe FX will likely not sound better in all reality than an actual SLO-100, but you can make the model sound pretty much any way you want it too. The amp will still have the basic SLO-100 signature sound, but you can use any cab, mic position and effects along with pre and post EQ at your disposal. You can even change the input impedance to change the sonic character a little too. What the Axe FX won't do is have the same touch and feel as the real deal. I have heard stories of people selling their Axe FX not because it didn't sound good, but because they found that they spent more time fiddling with the presets than playing, or because they found that all they ended up using was the FX because the amp models were un-inspiring.

To test by what is meant by touch and feel simply take your favorite amp and get a sound you like. Turn your guitar volume down with the volume knob. The sound should clean up a bit and also round out a little too ( should get more of a buxom tone ), but not loose definition and feel. It should not sound sterile and choked and should actually sound more dynamic and lively. I find the sound to get more ploinky ( if that's a word ) and the notes should almost bounce off the fretboard and through the amp. I have not found a modeler to react that way. The gain may decrease and the tone may round out a little, but the dynamics, touch and the way the note bounces off is not the same when the guitars volume is turned down with a modeler. The tail of the note also doesn't sing the same as with a real amp. I think this is mostly down to modelers and their heavy use of noise reduction and the higher noise floor?

I know that if I went out and spent 2.5-3k on a two space rack, my wife would choke me if I said it was not the best purchase I could have made. She understands that if you want a Dumble tone, you buy a Dumble amp, if you want an SLO-100 sound, you buy an SLO-100. It doesn't matter what the cost. $2.5K is not chump change and you can get many an amp that all sound amazing for that price. The down side is that the $2.5K only gets you in the door....... You still need a controller to change patches and banks while on a stage. A complete system for the Axe FX runs about $3.5k and that does not include the speaker. I suppose you could get the controller done for less, but it is also more of a hassle. You could also probably bypass the cost of a speaker if you already have one? If you were starting from scratch and going all in you will spend at least $4K on a full Axe FX rig with controller and speaker, new prices of course; used may save you a little coin.

I'm not sure people are skeptical about spending the money as much as they are worried about spending that kind of money for something that doesn't perform like they expect it too. It would have to equal the performance of an amp for me to be happy, the FX are just a plus. I have $3k+ in guitar amps ( after thinning the herd ) and they all sound pretty darn good. Each one does something the other can't and I would love to have it all in a two space rack if I could. That two space rack has to equal the performance of a real tube amp for me to be happy with it. Otherwise it's just and expensive two space rack. If it did, I would spend $10K to get it if each and every amp model was that good. You would have to be crazy to not want 100 amplifiers that all perform exactly like the real deal for that price. That is not where the tech is currently at though. It is so very close, but just not quite there. I will snatch one up though one day. Like I said, in the three instances where I had one through my PA, I couldn't tell if it was a real amp or not, but I wasn't playing it. If I play it, it is going to have to sound and play like a real amp to me.

[Andrew Broughton]

I have been playing guitar for about 23+ years and prefer high gain amps.
.
. (snip)
.
My bet is that within 5 years it will be figured out, cheap enough and superior to current tube amp design to have no excuse to not use a modeler.

I gave the analogy of the LP and CD, I'll give you another, more relevant one to live. Wedges vs. In-Ears.
In-Ears allow a performer to hear what they want and the FOH mix to be cleaned up (if you further eliminate the amps on stage), but many don't like what they hear in their ears. They don't like the "closed-in" feeling and want the freedom to adjust their mix by moving around on stage, and for other various reasons.
Typically with a good monitor engineer and good gear, and if the performer has some studio experience, it can all work out well for using in-ears. Even if it's not as pleasurable for him as using wedges, it allows for a better show, where others on stage aren't being blasted, the out front mix can be cleaner, and the audience can enjoy better sound. It depends on whether the performer cares most about the show or himself.
Everyone has agreed that out front the sound of the Axe-FX is great. If there's an issue, it's only the guitar player himself that isn't happy. Interestingly, I've found it's always hardest to get guitar players on in-ears too.
Is the show for the guitar player or for the audience?

Comparing a guitar amp to a modeler when playing it in your room or with a small band in a bar is not the right comparison. Keep your amp in those situations.
Instead, compare the modeler to amps in isolation boxes behind the stage where the player is on a 40' stage and is listening to his amps via an in-ear mix. Even if the amp is on stage, the guitar player is going to mic it up and put it through his monitor, he's not going to be able to get that small-room 1-amp tone.
Certainly there's no reason in that case to not replace that setup with a modeler, and that is why you are seeing more and more pro bands switch over. That and the other reasons I gave earlier.
Check out the list of endorsees. Are there any players that you respect in that list? If so, do you think they would use it if it wasn't working well for them? They can use anything they want. Fractal Audio doesn't give product away or pay the players to use their stuff. They have to be doing something right.

[Jay Marr]

@luke - you absolutely do not need the fractal controller.  Any midi controller will work.  I have a rjm mastermind that cost me $250.  It can perform the same functions as the fractal controller.

Axe fx 2, axe fx 2xl, axe fx 2xl plus all have the same modeling capabilities.  The xl versions just have more memory.  The first generation axe fx 2 still holds 500+ presets and more user cab files than I could ever use.  You can get one used for 1800 any day of the week...add a used atomic CLR for 800 and you have a crushing, super versatile rig....for 2800.  Not cheap, but it's not 5k.

[Bob Leonard]

Actually $2800 isn't the issue to me, it's paying for 500+ pre-sets that I won't use. Many years ago I opted for a Digitech 2120. I really liked that box and for the multiple genre I was playing at the time it worked out just fine, but when it came to real tone with depth for classic rock and blues I would unplug the Digitech, which by the way cost $1600 with the pedal at the time. It was very convenient to be able and step on a button and have "the" tone for the song regardless of genre. I used that box to push the front end of a 1964 Fender Twin Reverb, and that's where the real tone came from.

That was some 15-20 years ago, and I'm not saying the technology was equal to todays technology, that would be silly. But, for better than $3K the technology would be wasted for my use.

Keith, you refer to the endorsements for the product and I'll agree that's a pretty good list of people using the product, and as a matter of fact I've worked with a few of them. Masters of their craft each and every one of them. The question is this though. Do they use the box exclusively? Probably not, but even using the box at that level speaks volumes.

Like any tool the craftsman will dictate it's use, determine it's need, and determine the circumstances under which that tool will work best. To me it's one hell of a Swiss army knife, But me? I'll stay with the Craftsman socket set passed down by my grandfather knowing it will work for what I do. And speaking of socket sets, I'll sell the 1964 Deluxe Reverb on the left in my picture for $3200. If you buy it, I'll buy the Fractal.

[Stephen Kirby]

Is the show for the guitar player or for the audience?

Comparing a guitar amp to a modeler when playing it in your room or with a small band in a bar is not the right comparison. Keep your amp in those situations.
Instead, compare the modeler to amps in isolation boxes behind the stage where the player is on a 40' stage and is listening to his amps via an in-ear mix. Even if the amp is on stage, the guitar player is going to mic it up and put it through his monitor, he's not going to be able to get that small-room 1-amp tone.
Certainly there's no reason in that case to not replace that setup with a modeler, and that is why you are seeing more and more pro bands switch over. That and the other reasons I gave earlier.

Put me down as one of those guitarists who abhors in ears.  As you said, if someone does a lot of studio work and is used to that disconnected in the cans thing, then being similarly disconnected on stage is less of a deal.  Although I do note that a fair number of folks who are on ears still have amps on stage and sometimes at pretty significant stage volumes that blow past the 25-30dB attenuation of in ears.  John Mayer comes to mind.  And probably countless rockers who have one or two of the 4-12s hooked up and playing at them.  Or special wedges with guitar speakers in them.

I grew up playing live.  Played on TV in 1968 in SF (although it was pretty terrible Silvertone equipment at the time).  I have a hard time doing the playing in the control room hearing myself buried in the mix on NS10s thing.  I want to be in the room with my amp and K240s or even my Grados.  Which I find to be better than the one ear off thing many folks do.  Otherwise it's playing by braille. 

In a corporate band, on ears, such a thing is probably just the ticket.  I tried it for awhile with a Vox Tonelab which was the best sounding and feeling thing at the time IMHO.  I eventually took one ear out (replacing it with an ER9 plug so the levels were even) and went back to my Fuchs ODS and pedalboard.

A buddy of mine played on the original Guitar Hero tracks.  He used a Roland modeler for most of it.  It gave that already produced sound and he was able to closely match the original highly produced sounds.  Being able to clone the original records (and read down shred guitar parts) was why he got the sessions (although at the time he told me he didn't know why a game company wanted such exact reproductions compared to the usual chugga chugga he did for them).  The modeler was perfect for that.  Live he uses that pedal only for effects and as a master volume controller in the loop of a VHT or Marshall amp.

It really comes down to your approach to tone production.  Having had the experience of playing together with Robben Ford and watching his hands unconsciously feel around the fingerboard for the tone while I was trying to shove some sort of decent tone up the wire into his 2nd Dumble, I'm solidly in the "tone is in the hands" camp.

I remember once hearing a Korg M1 at a store with these great sax samples that blew my mind.  I was telling a friend about it who played both keys and sax.  "It sounds just like a sax."  To which he responded "Who?"  That just about sums up my experience of modelers.  It sounds just like X amp but not as played by anybody in particular. 

[Luke Geis]

But if you mix match things it won't be as cool........... I was only pointing out that as a full " rig " it's pretty pricey even if you can get the used prices down to the sub 3k mark. I can get a pretty nice Hughes & Kettner rig for that price and have a very versatile setup. Shoot I can even get an Engl Powerball 2 and cab for right around the 3k mark. 4 channels from clean to mean, noise gate and midi switching if desired.

I have been looking around and researching some more and it seems that the companies are starting to hone in on the decay and low volume input from the guitar. The Line 6 Helix is the newest kid on the block and that seems to be what they are claiming is nailed. I have not yet heard the XL+ supposedly the older versions can support the newer G3 modeling? The lower noise floor of the XL+ interests me the most. If they can truly get that touch part nailed it's game on. while I believe they have it to convincing levels now, I will a bet that it's still not quite there. They seem to keep focusing on what an amp will do at x setting and forgetting about what it will do at X setting when the guitar is doing this. The noise floor and gating interfere with the performance. I could care less if the tone stack is perfectly modeled. Does the sweep of the EQ encoders give me the sound that the amp would have?  The modeling of the core sound of the amps is there, they have it. It's just the touch and feel part they have to get.

[ProSoundWeb Community]