Kemper " Pure Cab " 'vs' Axe " De-Phasing " ....... *very* different !!!

  • Hey all


    Please see here:-



    This is [the excellent]. Mark Day doing a very quick demo of the. " de-phasing " feature in the [excellent] Axe 19.01.


    The Axe. " de-phasing ". feature is Fractal's version of making the Axe. More "amp in a room" ....... as is Kemper "Pure Cab ".


    What struck me upon hearing Marks demo:-


    - in Axe FW 19.01. it s is a very noticeable affect. Whereas Pure Cab is much more subtle


    Not saying this is a bad or good thing but given both "features" are trying to do the same thing, I am quite taken aback at how different they sound.


    Based on my personal use of Pure Cab in. 3.1.1 I find that setting it to " 5 ". brings the mids and lows more in to focus whilst taming the more harsh / ice picky higher end.


    What struck me with Marks demo is that when he goes from 0 to. 6.25, it seems to do the exact opposite - it loses a lot of the body of the tone and makes the top end more brittle. At 0. Its sounds freaking awesome but quite ordinary ar. 6.25.


    This is not a criticism - more just to say how different the 2 approaches seem to be.


    I'm certain many will think that Pure Cab. or De-Phase sounds better ....... but based in that video and my experience with Pure Cab, they really don't sound much at all like each other.


    Is it just my ears. ?


    Maybe its just the youtube audio effect - although I did listen at 720p with good headphones.


    What do others think ?


    Ben

  • I think the big difference here is what they're attempting to "fix." On the Fractal, amp models are created that emulate how a tube amp reacts to a guitar. On the Kemper, you have microphone-based profiles that are created based off who knows what system, but are captured in a similar manner.


    I think comparing the two is like comparing a Fender Telecaster to a Fender Stratocaster. They both sound great, but they have their own role, and people will generally prefer one over the other.

    Guitars: Parker Fly Mojo Flame, Ibanez RG7620 7-string, Legator Ninja 8-string, Fender Strat & Tele, Breedlove Pro C25
    Pedalboard: Templeboards Trio 43, Mission VM-1, Morley Bad Horsie, RJM Mini Effect Gizmo, 6 Degrees FX Sally Drive, Foxpedals The City, Addrock Ol' Yeller, RJM MMGT/22, Mission RJM EP-1, Strymon Timeline + BigSky
    Stack: Furman PL-Plus C, Kemper Rack

  • I think the big difference here is what they're attempting to "fix." On the Fractal, amp models are created that emulate how a tube amp reacts to a guitar. On the Kemper, you have microphone-based profiles that are created based off who knows what system, but are captured in a similar manner.


    I think comparing the two is like comparing a Fender Telecaster to a Fender Stratocaster. They both sound great, but they have their own role, and people will generally prefer one over the other.


    I agree! It's difficult to compare a ''Profiler'' feature to a ''Modeler'' feature. They are both digital devices trying to reproduce guitar amps but with a very different paradigm which require very different solutions to what looks like a similar problem.

  • The solution with modeling is not to tack on an extra "fix" but to model the amp and cab and leave the microphone out of the equation. Line 6's technique is to basically rip apart the amp's circuit board and, piece by piece, reverse engineer the schematic into their software. Speakers are a different story, but... I guess what I'm trying to say is the solution to the problem for Fractal is not what they're doing, but instead to create better models. Kemper has a unique problem in that there is no "model," just a dataset that represents the curves and output of a soundwave, based on a set of microphones, and, with Kemper's approach, there's no way to truly "fix" that 100%, unless there's a different way to profile an amp altogether without a microphone.

    Guitars: Parker Fly Mojo Flame, Ibanez RG7620 7-string, Legator Ninja 8-string, Fender Strat & Tele, Breedlove Pro C25
    Pedalboard: Templeboards Trio 43, Mission VM-1, Morley Bad Horsie, RJM Mini Effect Gizmo, 6 Degrees FX Sally Drive, Foxpedals The City, Addrock Ol' Yeller, RJM MMGT/22, Mission RJM EP-1, Strymon Timeline + BigSky
    Stack: Furman PL-Plus C, Kemper Rack

  • The solution with modeling is not to tack on an extra "fix" but to model the amp and cab and leave the microphone out of the equation. Line 6's technique is to basically rip apart the amp's circuit board and, piece by piece, reverse engineer the schematic into their software. Speakers are a different story, but... I guess what I'm trying to say is the solution to the problem for Fractal is not what they're doing, but instead to create better models. Kemper has a unique problem in that there is no "model," just a dataset that represents the curves and output of a soundwave, based on a set of microphones, and, with Kemper's approach, there's no way to truly "fix" that 100%, unless there's a different way to profile an amp altogether without a microphone.




    HMmmmm... Not sure I agree, don't most AXEfx users use cabinet impulse responses? Those would have baked the microphone into them as well.

  • There is no way to separate the microphone from the cabinet information. Neither from a Kemper cab nor from a IR file on the AXE-FX. They can only take adjustments in the filter or EQ settings that the corresponding curve represents. Though both implementations may use different calculation methods i am pretty sure they are both are aiming at the same thing: Smoothing out the curve to eliminate small frequency band peaks and cuts.

  • Different almost opposite things though. The Axe is trying to sound more like an recording of an actual cab in the room rather than a DI signal, so more natural sounding. It's not really an attempt to solve "amp in the room", but rather to be more like a natural amp like sound that you're used to hearing on record and get rid of the inherently inferior DI through an EQ sound of IR's (compared to the Kemper Cab section).


    The Kemper meanwhile is trying to make your FRFR speaker sound/behave more like a real guitar speaker, and move away from the sound of a real amp recorded in a space, so it's solving the "amp in the room" problem by trying to make it sound like you're playing an amp in your room.


    In short, the Axe wants to sound like the Kemper, the Kemper wants to sound like your amp.


    I think the next stage for the Kemper will be to follow suit of the Axe on this though, although not specifically this same tonal adjustment, general recorded amp tone focus, although more likely working on natural room sound/reverb/sympathetic sounds as the amp sound itself is with a few exceptions already there.

  • HMmmmm... Not sure I agree, don't most AXEfx users use cabinet impulse responses? Those would have baked the microphone into them as well.


    Well, that surely threw a spanner in the works; a more relevant point could surely not be made.


    There is no way to separate the microphone from the cabinet information. Neither from a Kemper cab nor from a IR file on the AXE-FX. They can only take adjustments in the filter or EQ settings that the corresponding curve represents.


    Correct, but although EQ is technically phase-shifting or "controlled" cancellation, I suspect there's much more going on than simple EQ / filtering. That is to say, phase-cancellation detection and subsequent amelioration, IMHO.


    I say this 'cause the nasty (especially high-end) phase cancellation I'm so allergic to cannot simply be EQ'd out in the traditional sense; by the time one's sufficiently eliminated the nastiness there's no top end left to work with. The algorithm employed here 'though appears to target the cancellation and correct it specifically as opposed to the "broad-brush" approach I just outlined.


    Though both implementations may use different calculation methods i am pretty sure they are both are aiming at the same thing: Smoothing out the curve to eliminate small frequency band peaks and cuts.


    This is the end result, for sure, but as I suggested, it's not a matter of simply applying a "reversed" frequency plot to flatten things out. Flattening the response surely isn't the aim after all; it's one of eliminating phase anolamies and thereby reducing comb filtering, which is where your description of the peaks and troughs holds true, again, IMHO.


    In short, the Axe wants to sound like the Kemper, the Kemper wants to sound like your amp.


    Ha! That's funny, Per.

  • i can't find any clips of de-phase on the axe other than the mark day one. I didn't think it sounded all that different from Pure Cab on that (although I do think de-phase doesn't sound very good in that clip), but I wish I could hear more settings.


    I think TylerHB nailed it - at the extreme settings, you lose all the speaker character - the Fractal wiki says that's pretty much what happens. So I think mild/medium settings are the good middle ground - you get something like the far-field sound but you still retain the sound of the speakers and cab. I tended to prefer around 5.5 and that seemed to work best for most profiles - for ones where it didn't work, it seemed the best setting was to turn it off completely.


    I'm curious to see the differences in this feature between the units. There are plenty of Pure Cab clips out there, but De-phase doesn't turn up too many clips. Reading the thread on the new Axe firmware to see how they like it...


    To me I feel like something does get lost as far as the clarity and transient. I want to do more extensive testing with it, while tweaking other parameters like pick and clarity to compensate. While it sounds pretty good through my monitor, I'm not liking it at all through headphones right now.


    One thing noticeable is that the Axe de-phase is available on the cab block, making it a per-patch/rig setting ;-)