Is it time for Fractal to upgrade the DSP chips in the Axe FX II/Xl

Status
Not open for further replies.
there can be no doubt there will be a more powerful version eventually .thats just technology




that said I can do anything I want to do keeping the cpu below 90% .If i want more sounds I make a new preset or try to use x/y. I dont think its realistic to expect to be able to do everything in 1 preset . it is still just a piece of hardware
 
So, does X/Y usage increase CPU? Do scenes affect it? Seems like some folks preset building methodology leaning toward certain features may cause their presets to run at the ragged edge more often given claims of sparse presets w/ high CPU cost that some are experiencing. Mine have to be loaded with FX before I get near 90. I tend to use premixed cabs and one amp, though. Almost never go nuts w/ verb.
 
zenaxe said:
So, does X/Y usage increase CPU? Do scenes affect it?
Click to expand...

Indirectly. When you switch X/Y state the effect has a new set of parameters loaded in to it. For some blocks, the parameters make a big difference in CPU usage. Reverb or the Cab block for example are highly parameter dependent. Similarly, when scenes switch the X/Y states of blocks they change the overall CPU use of the preset.
 
How does it work the Sharcs?
They still have the same chips as 4 years ago?
No Tick Tock at Adi?
 
I worry sometimes that Fractal's generosity in terms of development and innovation through firmware updates creates some unrealistic expectations and amplifies (pun very much intended) the sense of "needing" more and more from the unit. I also seem to remember Cliff saying in a thread that it would take a lot to exhaust the current hardware in terms of development...
 
jjdpro said:
Ok.. Is it time for Fractal to upgrade the DSP chips in the Axe FX II/Xl? I ask this because, I'm seeing so many post about CPU usage(V17) or CPU amount.

I thought the whole idea with the AXe FX was that it eliminated the worry about "running out of steam" DSP wise while being able to execute large amount of data for stellar modeling and effects. Unlike the other manufactures offerings, which were very limited in DSP power.

I'm sure that with each firmware update, which improves then product with each instance, there has to be more strain of the DSPs. Just a thought..
Click to expand...

Yes it is time. It was time for a CPU upgrade back when the XL came out. You are also right that there are a whole lot of users with CPU overload issues. I'd love to use the high quality reverbs, high quality mic preamps, stereo Ultra Res cabs, or two amps in stereo, but can not because of CPU overload. I am using none of the above and still run into CPU issues. With a preset at 90 CPU turning on the whammy and using the controller pedal will sometimes put it over. Or turning on the flanger. Fractal is fully aware of all the posts by users with CPU overload problems. They read the forums. The next generation Axe FX will have more CPU or they won't sell any to current users. In the meantime if it is humanly possible I think they may find a way to decrease overall CPU usage in upcoming firmware updates. Even by 5 CPU would help. They listen to their users.
 
I really think both sides of this debate are right.

Yes, I would like more DSP power as I do have to make small compromises in the way I set up my patches. This is because I like to use a single patch all night on my gigs and access my effects like a pedal board.

Yes, I believe we have by far the most powerful and amazing platform available at this time. I am so far beyond satisfied with this box it's not even funny, it's just silly.

Yes, it is easy to work around any DSP limitations by using multiple patches for different purposes within a given set/performance. I haven't even had to resort to that personally.

Yes, we need more DSP power, and NO, we don't need more DSP power.

I think as Cliff continues to improve and perfect this most amazing gem of a guitar tone machine, more DSP power can only be a good thing assuming latency and performance isn't otherwise compromised. I look forward to Axe-FX III and I hope it has more DSP power, but right now I don't really need it and I don't expect it to come anytime soon.

I also agree that adding the FX8 to an Axe-FX II rig will be an amazing way to add FX processing power and flexibility to the system. Not only can it apparently be added to an Axe-FX rig as a basic controller and FX machine, but it would also be an amazing FX unit for use with traditional tube amps. This prospect really makes me want to get an FX8 and a tube amp as an alternate and complimentary setup to my Axe-FX II rig. That's saying a lot considering that my Axe-FX II/MFC/CLR rig has brought extreme satisfaction to me on my gigs for a long time now and I really haven't used my amps much since I got my Ultra 4 years ago.

Long story short... you are all right, and Fractal, I have to say it, you guys just continue to blow my mind with nearly every firmware and product release. Thanks to you and to everyone on this forum for the whole Fractal experience.

Ok, I'm done. Cheers!
 
Rex said:
For all practical purposes, those are unavailable. Example: last time I checked, Mouser had three of them in stock. If you wanted more than three, lead times were nearly half a year. You can't build a product on that.
Click to expand...

cool thanks for the info! I didn't even know if they were released
 
You can cramp a lot in a preset by designing a preset smart. For example, the grid below requires between 80-90% CPU.

17XLgrid.png


COMP: the ""Pedal" compressors require less CPU than the Studio type. In this preset a Studio Comp is used.

PEQ: the first PEQ has the low-cut and high-cut characteristics of a "clean boost" Klon Centaur. It isn't the same as a Drive block but it requires a lot less CPU than a Drive.

PHASER, FLANGER, DRIVE, DELAY: all using X/Y. Also note that some Drive types take up more CPU (up to 4% more) than other types.

AMPS: I don't use the advanced stuff. Example: using OUT COMP increases CPU usage (depending on its setting).

CAB: I'm using UltraRes, mono. Saves CPU when compared to stereo. When I want to use multiple IRs I mix them into a single one, using Cab-Lab (to be released soon). A Mono cab suffices because all my stereo stuff comes after the Cab. Also, I'm not using the preamp simulation, saving CPU.

REVERB: I'm using the Spring Reverb. It has no High Quality mode, but it has been improved some firmware revisions ago and sounds awesome with my clean and lead tone tone. The Spring requires not nearly as much CPU power as other reverb types. If I need a specific other type of reverb I put in a dedicated preset.

The noise gate in the Input/Gate block is off, saving CU again.
 
Last edited:
somata said:
I worry sometimes that Fractal's generosity in terms of development and innovation through firmware updates creates some unrealistic expectations and amplifies (pun very much intended) the sense of "needing" more and more from the unit.
Click to expand...

Sooooo ... what you're saying is that Fractal's products spontaneously create demand for future improvements to Fractal's products. We should pause here for a moment to marvel not only at Cliff's rather extraordinarily successful implementation of a derivative of the crack dealer business model but at the fact that some of his customers worry after his having been too successful in generating future demand. May I be so bold as to presume that he likely does not see this as a source of worry? :mrgreen
 
FractalAudio said:
No. An FPGA can be used to do some bit-slice or word-slice DSP processing but pales to a dedicated DSP for the type of signal processing tasks we require. Typically you use an FPGA to off-load repetitive tasks from a host DSP or CPU. Where an FPGA shines is in massively parallel processing but they don't work as well as a dedicated DSP for sequential processing. This is why they are favored in things like SDR where you are doing a simple, fast operation on many simultaneous channels. Something like the Axe-Fx would perform worse using an FPGA solution, probably much worse.

If you have many tracks, as in a DAW, an FPGA can be a better solution due to the parallelism of the FPGA but the algorithms will be necessarily limited to simple things like EQ and compression. Complicated algorithms like reverb, distortion, etc. are better suited to a dedicated DSP. For example, the Xilinx "DSP Slice" is very crude in comparison to a dedicated DSP. It's limited to add, multiply and some logic operations. In contrast a dedicated DSP can do far more complicated processing.

I imagine the Avid cards are being utilized as dedicated coprocessors where some of the algorithm is native and certain operations are offloaded to the FPGA. Since this processing is necessarily parallel they can take advantage of the massively parallel architecture of an FPGA.
Click to expand...

The company I work for builds Broadcast Video systems that run on intel, nvidia, windows. FPGA is used for IO processing and routing. The rest is host based using CPU and GPU.

The Avid stuff may be similar. Complex effects are host based while FPGA handles routing and less complex stuff.


Sent from my iPad using Tapatalk HD
 
If you are willing to spend the cash for a AFX3, why not just get a used Mk1 and put it in the FX loop? You can run hires everything and still have room for another 15 blocks. Or get on the FX8 waiting list.

I would rather continue trading CPU for the stellar upgrades we are getting, than have Cliff and crew become focused on design of a new unit and get distracted from improving what we have. Think about that for a minute.

If you are out of CPU put your presets on a diet or get another unit and stop posting these whiney threads. (Just kidding of course. LOL)
 
been thinking about this thread..
it's kinda self defeating..

it's like going to Aston Martin and telling them "your new Vantages need an extra 100bhp"

now if the nice folks at Aston took note of this and thought "yes.. he has a point.. let's do it"
everyone buying a new one would get the benefit of your suggestion.. but you wouldn't
you'd still have your Aston with it's current horse power..

unless of course you'd have enough cash to really not care about taking the $loss on the sale of your current one and then stumping up the $extra on the latest greatest model..

I do reach the 87%+ CPU mark from time to time..
when I do, my solution is to get creative and build a more efficient preset, or strip out any 'nice to have' fx / features that can be deemed "non-mandatory" [given that presets primary function]..
additionally, I do have some presets that are in the mid / high 80's% area - which for me is fine for practice..
but when I play live, I use real cabs, so my live presets do not have cab blocks.. that drops the cpu a little further and guarantees good performance..
 
randolfo said:
why not just get a used Mk1 and put it in the FX loop? You can run hires everything and still have room for another 15 blocks. Or get on the FX8)
Click to expand...

Though linking units will work, I don't see it as a very ideal solution to constrained CPU. It's an expensive fix, it's physically awkward, duplicates AD/DA conversions, requires more cabling, and is trickier to level. Instead I'll just trade up to a an Axe3 when the time comes (though I'd love to get an Fx8 for my tube rig!! gas ... gas ...).


clarky said:
unless of course you'd have enough cash to really not care about taking the $loss on the sale of your current one
Click to expand...

Given great resale value, the Axe is very cheap to stay current with. Once you have your first unit, only a 30-40% additional investment every 3-4 years keeps you in the evolutionary upgrades (which is 1/2 the enjoyment for me, though others may choose to get off the bus at any time and still be left with a great timeless unit (i.e. Standard, Ultra)). The pace of evolution makes the Axefx cheap cheap cheap compared to anything else out there.
 
toasterdude said:
The company I work for builds Broadcast Video systems that run on intel, nvidia, windows. FPGA is used for IO processing and routing. The rest is host based using CPU and GPU.

The Avid stuff may be similar. Complex effects are host based while FPGA handles routing and less complex stuff.
Click to expand...

That's exactly what the Avid cards use the FPGA for: mixer routing. All the effects processing is done on the onboard DSPs.
 
Status
Not open for further replies.
Back
Top Bottom