Why are different amp channels treated as different models?

When I was brand new to AXE FX, I used to pine for each amp to be "fully modeled" into one block - so that when I pulled up an amp I had all of the channels, controls, etc just like the real amp. To me (at the time), THAT was a big part of "super realistic amp modeling."
I have grown past that though. Now I am more of the belief that the AXE FX is THE greatest amp of them all, in its own right. I no longer look for it to exactly copy what was already out there. Heck, if I really wanted the best model of a Mark V ever, I should just get a Mark V. (and... yep... I own one of those too.)

Rather, I treat the Fractal unit for what it is... which is the same as any other amp.... it's a tool to get the the SOUND I want. It has a lot of power, so that means it will take a lot of exploring - but in the end I can get things out of one rig that I could never get from another single rig, so I am more than ok with that trade off.

** The CHANNELS feature, along with scenes, is much more powerful than any 3 channel amp, and using that I can have all three channels of an amp and still only use ONE amp block. This actually gives the OP exactly what he asking for in terms of function... just maybe not by following the same workflow he desires. In the end though, by using two amps, and four channels, you could have 2 different 4 channel amps in one preset. Do any of the other devices mentioned do that? What's even BETTER in my mind, is I CAN use 3 channels of a particular amp - but I don't HAVE to. I can mix and match them to build a dream amp.

In my experience, with very few exceptions amps that have stellar high gain don't normally do clean all that well, and vice versa. The AXE FX cures that issue like nothing before ever has that I have tried or used.
 
I like more having an amp model and switches from clean to dirty, in that sense interfaces like S-Gear are amazingly intuitive, easy to use and close to the "real thing".

That being said, the Fractal workflow is not that difficult, you get used to it quickly. I have always open in tabs of the browser the wiki's Amplifier Model and Cab Model list to navigate thru the amazing variety of options.
 
yoitsmegabe said:
I like having the channels modeled separately. It allows you to make a "super amp" using the amp block with different amps as your channels a,b,c,d.
Click to expand...

This would be really useful for me, as I'm not familiar w/all the real amps. So, when selecting an amp model in channel A of the Block, I could select the Amp and a Checkbox in the Amp List would allow me to select all related channels modeled in order Clean, Crunch, and Lead (example)...or Normal, Vibrato...these would auto load in the Amp Block Channels A, B, C...etc. I would think this would be really useful and not disturb the existing Fractal workflow/design...
 
well it would if you already had amps in those other channels that you had selected previously and just wanted to change what was in A.

this is trying to solve a problem that doesn't exist, imo. the amps are already named after their various channels, if they've been modelled
 
chris said:
yeah there are a lot of cabs, and you'll just have to spend some time going from cab to cab to see what you might like
Click to expand...
And IMO this is exactly the part that causes problems for a lot of people. There's about 2300 IRs in the Axe III, and selecting one that matches the preset can be a more daunting task than you might realize - especially for users who are not very familiar with the actual hardware (cabs+mics as well as amps). For example, I have no idea what to expect from half of the mics used to shoot IRs as I never had any contact with them in real life; same for the differences between certain speaker models. It's also a bit of an hen-and-egg problem with amp and IR. For a new preset you start with an amp at its defaults and the 1x4 pig IR and need to tweak the amp tone and find the right IR, but of course changing the IR also affects how you need to tweak the amp and vice versa. While you can get a decent tone with a lot of IRs with enough tweaking, it's often far from obvious if the IR is the weak link in a preset.

I agree that ultimately it comes down to each user selecting what they like best, but I think there could be better tooling in support of that goal. One way to get there could be an overhaul of Axe-Change to allow for easier sharing and selection of individual blocks in addition to full presets. Then users could search for an appropriate preconfigured cab block to use as a starting point, instead of sorting through 2000+ IRs to find a sensible default.
 
I guess I'm just simple. Most of the modeled amps have a matching cab type....pretty easy to grab one as a jumping off point. Probably a solid finishing point too, really.
 
The reason different channels get different models is because your model has to simulate every different part of the audio path from input to output to simulate an amp. The easiest and most performant way to do that is to only simulate the parts that are actually used in the path and not simulate anything else. All the circuitry for Channel 1 and 2 are sitting unused when Channel 3 is in use, but to allow toggling between them in a single model they'd all have to be loaded and running anyway. Plus you need to start modelling all the switches and toggles that change when channels are changed. In the Mark V there are over a dozen different places in the circuit that toggle in and out when you change channels or mode. You could try and make one giant model that has all those pieces and duplicates the logic and switching, or you can just model all the parts that are engaged in one mode of one channel, and do that for each. Now you have a handful of (relatively) simple models instead of one giant complex one. And the channels system then allows you to easily combine them together as if they were one model.

Helix doesn't have a channels analog, so allowing more things to be toggled within their one big model is a higher priority for them.

As for the little mode switches, that's a similar difference in approach. The Fractal amp block is a really specific structure that is fixed in shape and almost every amp has the exact same controls. That allows them to make the bright cap logic once then just have it on every amp in the ideal tab.
1589905716047.png

Helix seems to make their models a little more one-off, where each model has potentially more complex toggling logic, more custom controls and more to debug and test. The amount of time it takes them to add an amp model is also considerably higher seemingly.

They are trade-offs, but Fractal's choice (a really cookie-cutter amp block where they add the required fixed circuit simulation to the pre and power amp sections, and set the right defaults for all their parameters) seems to be working well for them, and they quickly get an amp with a lot of controls and functionality out of the box. But more complex toggleable changes to the circuit mid-stream don't fit, so they become new models that are largely duplicates with the required changes.

Helix's amps are more self contained, they can add all the knobs and switches they choose, but that also means they have to choose to add one each time. Bright switch? Only if the real amp had one, otherwise it's extra work that the real amp didn't even have. Or they might not even add it even if the real amp did (their newer Friedman and Revv models have lots of toggles, but their Mark IV has none). So their amp can be more flexible for custom toggles, but it's more work to add each part.
 
IronSean said:
Now you have a handful of (relatively) simple models instead of one giant complex one. And the channels system then allows you to easily combine them together as if they were one model. The Fractal amp block is a really specific structure that is fixed in shape and almost every amp has the exact same controls.
Click to expand...

Yes, the Axe-Fx has a specific template structure but it's generic thus allowing adding/subtracting/changing features on amps that normally don't have those (e.g. all the ideal + advanced parameters). In some cases a particular part of an amp circuit (e.g. bright or depth control) might be unique so a new amp model is created while still maintaining all the standard generic features (bright switch, depth control, etc.). FAS AMPs also allows circuit modifications like changing the tone stack location or changing from power tube grid bias from Class A to B operation, which modding the amp on the spot.
 
yyz67 said:
Yes, the Axe-Fx has a specific template structure but it's generic thus allowing adding/subtracting/changing features on amps that normally don't have those (e.g. all the ideal + advanced parameters). In some cases a particular part of an amp circuit (e.g. bright or depth control) might be unique so a new amp model is created while still maintaining all the standard generic features (bright switch, depth control, etc.). FAS AMPs also allows circuit modifications like changing the tone stack location or changing from power tube grid bias from Class A to B operation, which modding the amp on the spot.
Click to expand...
Exactly, so much of it is configured instead of hard coded, it means that all those dials and adjustments can be exposed to the users and we can further tweak them ourselves. It's one highly configurable amp model where the only parts that aren't really really just configuration settings seem to be the chunks of circuit programmed between certain parts of the model. It's really flexible, quick to add new models if you have a good schematic, and highly tweakable for the end user. And you get a lot of additional "mods" in terms of extra EQ controls, saturation switches, presence, depth, bright caps, etc that might not have existed on the amp for free, since it's built into the Uber-Model.

The only limitation is that there doesn't seem to be a mechanism for adding additional custom switches into those sections of the modeled schematic, so you get USA LEAD, USA LEAD BRIGHT, USA LEAD DEEP, and USA LEAD BRIGHT DEEP instead to represent the combination of switches. That's a pretty small price to pay for everything we get in return.
 
IronSean said:
The Fractal amp block is a really specific structure that is fixed in shape and almost every amp has the exact same controls.
Click to expand...
yyz67 said:
Yes, the Axe-Fx has a specific template structure but it's generic ...
Click to expand...
Yes, but only at the level of the block diagram and what controls we can access. At a component level I'm pretty sure that the actual coding of the model changes depending on the amp/circuit topology. For example, Cliff talks about how xyz amp has an extra gain stage in the preamp on the red channel, or has another capacitor somewhere, or sends part of the signal down feedback loops, etc. All of those are unique to that specific amp (or amp's channel) and would add a lot of programming complexity in an "all channels in one amp model" scenario. These are all "under the covers" and obviously affect the tone we hear, but don't affect the controls we use to interact with it. Simple analogy: a Fender circuit is not the same as a Rectifier, which are both different than a Marshall, even though we see all the same controls in the interface and the block diagram would look the same from our perspective.

As I said, I'm good with the way it is now. But if it was going to change, I'd imagine the easiest way to code it would be to just have another level of hierarchy - instead of each amp channel being separate, they'd create a higher-level container and then each channel (each current "variation" of the amp) would just be a selectable sub-module under that. Changing the entire coding model to have one topology for the complete amp, and that switched all the internal components would be a massive undertaking. I'd rather they put their resources into other things...
 
chris said:
hmm, the process of selecting cabs is simple: click the cab block, click the Picker and choose from the pop up list. that part shouldn't be more difficult than any other product.

from there, yeah there are a lot of cabs, and you'll just have to spend some time going from cab to cab to see what you might like. as i mentioned earlier, after i did that, i just stuck with a few cabs and moved on. i honestly don't spend any time in the cab block anymore.
Click to expand...

As Chris mentions, there are a lot of cabs, but one way to reduce the time going through them is to put the Looper in the front of your signal chain, record a riff that you are trying to dial in, open the cab picker, pin it and then arrow through the IR's until you find some you like. You can even color code them so you can find them quickly next time you need them.
 
What controls are configurable at the user level has nothing to do with the underlying code; only the parameters we as users can access. Assuming how the code is assembled or how Fractal models amps is... fool's gold.

The initial assumption that the OP makes is flawed to be kind.
 
Scott Peterson said:
The initial assumption that the OP makes is flawed to be kind.
Click to expand...
What assumption are you referring to? I feel like I only asked a question and ended by musing that it would probably be unruly to try to tweak/debug a comprehensive (say, Mark IV) amp from a developer's standpoint since, as others have confirmed, this would be a huge undertaking rather than each channel of the modeled amp at a time. I think the thread kind of veered away from the starting point, but that's cool - it's exciting to see the discussion! I love reading people's insights into my question.
 
decreebass said:
Why couldn't, for instance, the JP2C+ model simply contain the green, yellow, and red channels accessible with the flip of a switch? Why do these need to be different "amps?" This goes for some of the pedals as well that have different modes or toggles. Why not just include that toggle or mode in the pedal (Timmy, for instance, rather than having two Timmy pedals).

I guess it doesn't really affect anything, since you'd most likely have them all on their own Block channels were you to use, say, the JP2C and all its channels in a preset, but it's curious. The Helix does the same sort of thing. The only thing I've ever seen have the "full" amp in one model was Amplitube. The Helix has all the toggle positions in one block, but it still tends to break up all the amp channels into their own amp models.

I get it if, for instance, only one or two channels of a given amp are modeled, but when the whole thing is, it just seems like the whole amp should be in one amp model. Is there a practical reason why companies do this? Would the file inside the firmware be too big or unruly to work with or debug/improve?

Just curious :)
Click to expand...

That's just simply how Fractal has decided to present their amp/drive UI - I'm sure it makes it easier for them given the amount of amps/drives available. From a certain perspective - it presents a consistent view for amp controls - so even if a user choosing very different models, you sort of know where the knobs are as they are consistent.

But I understand the flip side as well - presenting a view of a modeled amp that is closer to the original controls.

I would create a wish list item - it may not happen - but you can put it out there. Nothing wrong in suggesting potential UI improvements.

For example the "Authentic" page - something to that effect was requested since the Axe-Fx I days... it eventually happened.
 
They could have maybe done a simple, purely visual approach to organizing where you chose an amp then a channel, but that would really just be adding a folder type structure to represent the same list that currently exists. Maybe it's mentally simpler to wrap your head around, but it's also more steps to step in and out of these folders to change models.
 
Scott Peterson said:
What controls are configurable at the user level has nothing to do with the underlying code; only the parameters we as users can access. Assuming how the code is assembled or how Fractal models amps is... fool's gold.
Click to expand...

It has a lot to do with the underlying code. As a software developer, you abstract to the degree you can to accommodate the needs of the user, but in large part, a GUI is a window into the underlying architecture. However, if the architecture is totally different than the GUI, that's a sign the architecture is based on an incorrect model of how it will be used.

The initial assumption that the OP makes is flawed to be kind.
Click to expand...

Wasn't he just asking about why the AxeFX presentation doesn't more closely mirror the real amp? There are pros and cons to both approaches, but it's hard to see how either could be considered "flawed".
 
You must log in or register to reply here.
Back
Top Bottom