Wish Dual Amps - even if they'd be stripped down versions

Ghost_of_Cain

Ghost_of_Cain

Inspired
Bold suggestion, isn't it? ;)

Nowadays I read many great topics here about how people would appreciate squeezing bits of the new algorithms into old AX8's hardware, while I feel I'd rather go for something different. If anything can be still optimized, when it comes to processing power, let it be to employ second Amp instance in the grid.

It could be kind of "low-res" button, could be stripped-down amp version without some in-depth controls, it could even rule out using nice reverbs / whatevers completely - I don't care.

I can live fine with just one amp in the grid and control the gain / master via scene controllers to get smooth clean-to-dirty tone transitions, but having 2 amps (channels) to do the same, with independent tone control - that would be something.

As for how it would work - I imagine the 2nd amp would be always available, but without pushing this, let's say, "low-res" button, we would just get the same message we are getting when overloading the cpu (that the block is disabled).

Dreaming? Maybe. But if FAS team would ask to vote whether we want them to work on "porting more new epiphanies" or rather "optimizing cpu usage to get more Axe II functionality from the old AX8 hardware" - I'd vote for the latter.

Thanks!
 
Last edited:
One practical limitation to this is that there are two CPUs - only one is dedicated to the Amp block. Freeing up CPU from reverb won’t help. I remember reading somewhere that the Class A type models use most of the available CPU already (for example).

It’s not about how many controls are exposed to the user, it’s about how complex the modeling can be at the component level. It would need a simpler, lesser modeling algo to run two amps.
 
“Stripping down” an amp model by removing controls isn’t going to save much CPU. The real CPU hog is the modeling itself. The only way to reduce CPU would be to take shortcuts in the modeling. In other words, make a model that doesn’t sound as good.
 
Rex said:
“Stripping down” an amp model by removing controls isn’t going to save much CPU. The real CPU hog is the modeling itself. The only way to reduce CPU would be to take shortcuts in the modeling. In other words, make a model that doesn’t sound as good.
Click to expand...

Gentlemen, I know all of this.

But at the same time, I remember, when ZX Spectrum was getting old, there appeared a dozen methods of defeating its graphic modes' limitations. With astonishingly good effects ;) :D
 
Ghost_of_Cain said:
Gentlemen, I know all of this.

But at the same time, I remember, when ZX Spectrum was getting old, there appeared a dozen methods of defeating its graphic modes' limitations. With astonishingly good effects ;) :D
Click to expand...
ZX Spectrum was a 35-year-old budget personal computer from a time when color computer graphics and animation were a major hit on the CPU. The AX8 uses modern technology to drive a monochrome display of about the same resolution and almost no animation. There’s not much CPU to be saved by removing controls from the screen.
 
Rex said:
ZX Spectrum was a 35-year-old budget personal computer from a time when color computer graphics and animation were a major hit on the CPU. The AX8 uses modern technology to drive a monochrome display of about the same resolution and almost no animation. There’s not much CPU to be saved by removing controls from the screen.
Click to expand...

Not entirely sure if you're getting the metaphor here... No. I'm sure you're not. :D
 
Rex said:
Pretty sure you're not explaining it, either. :D
Click to expand...

I'm not sure I'm able to explain it further. All was said in the first post - I meant "optimizing the code", doing miracles on non-expandable hardware, as-it-is, only by and due to the fact, that the hardware is well known for developer, and one has been programing it for years.

History of 8 and 16-bit computers reveals many great examples on how impossible could be made possible on presumably obsolete hardware. Thus tricks with ZX Spectrum were perfect metaphor background. And it had nothing to do with graphics / displays / whatsoever. Your comment about "controls" and "screen" just scared me off - I simply cannot imagine, how you couln't see under this layer. Sorry!
 
all of the following "amounts used" are assumed. i don't know how it all works exactly.

the way i see it, if it a full CPU is used for an amp block currently, to have a 2nd amp with the same quality, the other CPU would be needed for a 2nd amp block, assuming there is some CPU left over to run the device itself (screen, controllers, audio, etc). that would theoretically mean ZERO effects. i'm not sure anyone would want the AX8 to do that.

from there, the thought of reducing quality to save on CPU enters the equation. hypothetically speaking, if the Amp quality we have now uses 1 full CPU, to get 2 Amp blocks, we'd "halve the quality" of the current Amp block to squeeze 2 Amps into the 1 full CPU. that's probably not actually possible, you can't just cut it in half. but assuming it is just for argument's sake, would you be ok with "half the quality" in the amp block? any clue what that'd sound like? if it's anything like going from a WAV file to a low quality MP3 with sparklies in the highs... no thank you. i bet "half the quality" would sound exactly like that.

i don't think it's as simple as "reduce code" or "reduce features." optimizations are always being found and implemented. so i doubt FAS is sitting on a 20% less CPU version and not giving it to us. amps are complicated. there are so many modeling companies now, and they all have their own sound. some are obviously more simple than others, and we can hear and feel that. that's what make people prefer one over the others.

additionally, i'm sure there's a threshold of what FAS finds acceptable for their amp models to sound like. i'm sure at some point there was a lower quality sound as a test, but it was just not passable quality.

so is further optimization possible? maybe. we've been getting optimizations over the years though. so it'd take a big breakthrough to make a substantial change. perhaps that breakthrough was the new processors in the Axe-Fx III? hard to say, but that'd be my guess.

finding optimizations in other gear perhaps shows that those were just not optimized to begin with. i don't think it has much to do with FAS gear. but sure, it's always possible that at some point optimizations are possible.
 
Ghost_of_Cain said:
Thus tricks with ZX Spectrum were perfect metaphor background. And it had nothing to do with graphics / displays / whatsoever.
Click to expand...
When you said that the ZX had been optimized by "defeating its graphic modes' limitations," I just naturally assumed you were talking about graphics. :) And graphics — even basic graphics — were a major performance hit on CPUs 35 years ago.


Ghost_of_Cain said:
Your comment about "controls" and "screen" just scared me off...
Click to expand...
"Screen" is all about graphics (see above). And when you said "It could be kind of "low-res" button, could be stripped-down amp version without some in-depth controls," I just naturally assumed you were talking about controls. :)


Bottom line: The ZX was a consumer-grade, general-purpose computer running a general-purpose operating system. It's not surprising that this could be optimized. Fractal products are pro-grade devices whose operating system is custom-built to serve just one application, and the code has been optimized repeatedly. There's not much wiggle room.
 
Rex said:
when you said "It could be kind of "low-res" button, could be stripped-down amp version without some in-depth controls," I just naturally assumed you were talking about controls. :)
Click to expand...

Ah, I see :) I was thinking about low-res as, let's say, the cab block's "normal resolution" option :D

Bottom line: The ZX was a consumer-grade, general-purpose computer running a general-purpose operating system. It's not surprising that this could be optimized. Fractal products are pro-grade devices whose operating system is custom-built to serve just one application, and the code has been optimized repeatedly. There's not much wiggle room.
Click to expand...

That's a valid point, sir. I still do hope that there is at least a little headroom in it - and I hope we'll see some surprise in the future.

That would make sense from the commercial point of view - to put in AX8 something that would clearly convince the large audience, that: AX8 is not dead, and you can safely buy it, AXIII-alike in floor format is still far away. Who wants revolution - should pull the trigger on the newest product; who wants reliable and yet still actively updated one - can still be happy with the "old" thing. But that's another fantasy story :)
 
Last edited:
  • Like
Reactions: Rex
I think FAS has always gone for the best possible quality modeling. If you were okay with one amp block you went ax8, if you needed two you went for the II

I get why some would be okay with 2 lower quality models, but there are lots of other players in that game.

I think FAS does a given job and does it better than anyone. Just need to pick the hardware that had the capabilities one needs
 
Back in the days it was the user community pushing then limits of the unit, not Sinclair. This is not going to happen to a closed appliance once the manufacturer shifts focus to a new box.
 
ChristThePhone said:
...once the manufacturer shifts focus to a new box.
Click to expand...

That's the point! If manufacturer would like to extend the life of AX8, and take their time to develop a decent successor, they might throw in a few "epiphanies" showing, that the current line is still alive. Right now there are surely few undecided customers wondering if they should buy AX8, or wait for the next floor unit from FAS, as AX8 might be "nearing the end of its life" (I'm not bothered by this, just telling about what I often see now on the internet). Any new features would be a good "pros" for those undecided.
 
Ghost_of_Cain said:
... Right now there are surely few undecided customers wondering if they should buy AX8, or wait for the next floor unit from FAS, as AX8 might be "nearing the end of its life" (I'm not bothered by this, just telling about what I often see now on the internet). Any new features would be a good "pros" for those undecided.
Click to expand...

Nobody knows when the AX8 successor will come. All I can recommend to the undecided is: Ask yourself : is the box good enough as is ? If yes, go get it and view updates as a bonus. If not, start waiting. Like a real amp it will still sound great in 10 years without a single update.
 
Buy a III now, enjoy it, if/when a new ax8 comes out, and it’s got the feature set one wants, sell the III and buy the new floor unit

Let’s say it doesn’t come out for a year, well, you then get a year of III enjoyment, can sell that for maybe $500-600 less than new costs, and buy the new unit. “Cost” you $500 for a years enjoyment of a III.

I’ve easily spent more than $500 on shipping and seller fees buying and selling pedals in a year lol.

One can certainly always just enjoy what they’ve currently got too. Nothing wrong with that, but if one enjoys the latest and greatest, new features etc, the III is going to be where the excitement is moving forward
 
You must log in or register to reply here.
Back
Top Bottom