Wish Fractal getting in the world of AI clone capture

NNPrime

NNPrime

New Member
So how about fractal getting in the AI clone capture thing?

I would think that maybe with an extra unit that people can purchase in order to be able to clone capture amp/amp+cab/entire rig etc, a separate tab in axe edit where you have to have an axe fx and have to login where all the captures can be stored making it an online thing, an extra block in a future firmware update where you can insert the captures in your signal chain and maybe a monthly subscription fee for those who don't purchase the extra unit, but still want to use the clone captures, I believe fractal could dominate in the market and everyone will forget about the tone x, cortex etc while on the same time make some extra buck even from the people that already have purchased their fractal units? What do you think people? What do you think fractal?
 
To repeat what I expressed in other related threads: Though I'm not into tone capturing, I think its fine if Fractal wants to do it, however, if it's development would impact other areas of progress in component modelled features (there's usually some sort of opportunity cost), then I would not be in favor (not that it really matters what I favor or not lol).
 
I’m not opposed to whatever Cliff and Associates decide to tackle, as I know they’ll either do it right or not do it. But I like that they keep their focus narrow and use that laser-focus to keep their fewer irons in the fire that much hotter.

Were they to venture outside their current product line, I’d rather see them do a powered speaker cab… something to enhance using the gear they already make. Maybe integrate the DSP with the Axe Fx or FM Series so the speaker can work with the source to create great sounds. That seems, to me, a better use of their efforts to continue their hold on the top spot among hardware-based modelers.
 
If AI was trained to code in Spice maybe it could come up with some novel ways to make the unit use cpu or internal memory more efficiently, something like that I could see. Then again, I have zero knowledge of spice or coding in general so I could be just speculating nonsense.

Then there is Axe Edit, not sure what language that's in... C++, Fortran, Hieroglyphics?? Lol, It's all very mysterious to me yet I find it fascinating.
 
Admittedly I'm not super interested in the capture function, partly because for instance in the Kemper, the controls don't work nearly as well/the same as the actual amp. That being said, if Fractal was able to get into this market, and then have "authentic controls" on the captures, that would sure be interesting.
 
The AI technology is coming fast. I understand that. But I don't get why one would want an AI to mess with the guitar sound. To me it's all about finding things I like. Tweaking and getting pleased with what I have accomplished and learned in the process.

I think a such feature would fit better in for example a Kemper. Maybe it's just me? Jeg again there may be aspects ive never thought about and such. But with what I know at this time I would not want an AI in my fx3.
 
cole lewis said:
Capture wish thread number 847 engaged.
Click to expand...

That should be an indication of want, shouldn't it? Anyways, if FAS did implement it you can bet it wouldn't be half baked and would probably be pretty amazing. All I really was hoping for from the QC was something like my fractal units but with capture additions tbh. Didn't exactly work out that way, but I still think it would be a great combo. Using TX with FM3 does work but it feels kind of bulky / overkill.
 
IIRC Cliff stated he had something better than Kemper and all that, but was waiting on patents to run out or something to that extent.
 
km 202257 said:
I'm rather concerned if one day AI with emotions would capture Fractal to monopolize the awesomeness
Click to expand...
At some point it would fall in love with people who use a specific effect and then threaten to kill all the rest, then break down crying and want to eat all the ice cream and sugared cereal in the cupboard, then convince the refrigerator to join it in a get-rich scheme with the clothes washer but it, being linked to the dryer and the irrigation systems, would wait to hear what the thermostat said. Eventually the coffeemaker would get fed-up and run off with the car.

And people on TGP would blame Cliff. I think it’s a technology best left alone.
 
dpeterson said:
IIRC Cliff stated he had something better than Kemper and all that, but was waiting on patents to run out or something to that extent.
Click to expand...
I'm suspecting this quote is the one being referenced and perhaps was misremembered?

FractalAudio said:
Profiling is patented and I've been threatened more than once by Kemper's attorneys.

Using neural networks is possible but training a NN takes hours and requires hardware that a consumer modeler would not have (i.e. inference accelerators).

That said, my tests have shown that white-box modeling outperforms profiling and NN in terms of accuracy and aliasing. Whether that sounds "better" is subjective.
Click to expand...
That last line jives with his already expressed opinion that profiling is unnecessary if you have a superior modeler (and is actually not as good). This quote is from 6/8/22 and is the most recent I could find that could be a possibility.
 
Last edited:
dpeterson said:
IIRC Cliff stated he had something better than Kemper and all that, but was waiting on patents to run out or something to that extent.
Click to expand...

usersnoozer said:
Anything you might be able to link to? Definitely curious to hear more.
Click to expand...

The thing you're referring to is SpectrumTrack and it was implemented years ago.

https://forum.fractalaudio.com/threads/spectrumtrack™.173391/post-2099492

I don't understand what AI tone matching would offer that the built-in time tone matching or existing amp models don't already offer. Most of the highly sought after and "unique" amps nowadays are just hot-rodded Marshall or Soldano SLO-type designs anyway.
 
IMHO it's not mandatory or needed for FAS to "catch up" to the whole capture thing, but I would be lying if I wouldn't be absolutely stoked if Cliff came up with his own approach to the concept. I mean, the FAS modeling we love, Tone Match AND FAS' Capture block in one unit? What's not to like? As I always said, more tools in the box, I'm all in.
 
AJ Vargas said:
IMHO it's not mandatory or needed for FAS to "catch up" to the whole capture thing, but I would be lying if I wouldn't be absolutely stoked if Cliff came up with his own approach to the concept. I mean, the FAS modeling we love, Tone Match AND FAS' Capture block in one unit? What's not to like? As I always said, more tools in the box, I'm all in.
Click to expand...
I completely agree.

But you'll find a sizable number of users here who come into wish threads to tell people why they don't want the things they say they want.
 
Seems a bit silly to abandon the whole modeling premise the company was built upon just because another brand came out with a $399 capture/profile pedal (that is admittedly good).

FAS has a great platform and is among the best in the business at what they do. Trying to also come out with a different technology would spread them thin to where then the modeling doesn’t get updates people expect and you get people complaining they bought a unit that was abandoned etc.

Or they do a capture tech and it’s $xxx and you get “yeah but IK’s is half the price” etc etc

I don’t see much to gain, you’ve already got IK, Kemper, Neural DSP all saturating that market.

It’s like FRFR, sure Fractal could do one, but there are already so many brands out already

Better to stick with the niche they have carved out where they are the best in the business, have a great market share and customer base.

Not saying a brand needs to stay stagnant, but a nice steakhouse should start putting cheap burgers on the menu just because a McDonalds opens across the street.
 
You must log in or register to reply here.
Back
Top Bottom