Refining, redux: can we have some science here instead of voodoo?

  • In the interest of reliable methods and understanding, I want to know exactly what the "Refine" button does. I've searched this forum and elsewhere and not found anything definitive; all I've found are anecdotal stories and "your milage may vary" advice - a suboptimal state of affairs, to be sure. Like others, I've had profiles that were fine with Refine and some that Refine wadded up into a ball.


    I wanna know:

    1. What is happening during Refine exactly?

    2. What problem is it intending to fix?

    3. Under what circumstances is it more necessary vs. making things worse (and in what I've seen, not just worse, but absolutely destroyed)?

    4. What is the effect of settings before Refine, i.e., the Profiling process, and settings/actions during Refine?

    - Setting Clean vs. Distorted (a distinction I've never quite understood, as all amps will distort when pushed, and you may start with something clean, but crunchifies with more aggressive playing)

    - Return Level during Profiling: when Refine trashes a profile is it because the Return Level was too high during profiling, and Refine just magnifies this overload distortion? I'm using multiple mics, mixed via UAD Neve 1073 preamps, and the output is thus capped at 0dbfs line out; what's the optimum Return Level setting on the Kemper when fed this way (as opposed to plugging a mic directly into the Kemper? This parameter would seem critical, as there must be sufficient headroom on the Kemper's Return to capture their profiling signal without distorting.

    - type of guitar used during Refine

    - playing style during Refine

    - Gain control during Refine - I found that you can change the Gain setting while profiling, and so far it seems to do much more harm than good. What I'd really like to see is the ability for me to set the Gain control (before starting the profiling noises) to where I think it reflects the amount of gain I have dialed into the amp. e.g., if I'm doing really clean, I may set it to 1-2; if I think it's about half-way in its useful range of crunchitude, I'll set it at noon; at the upper end of the amp's useful range, I'll set it at or near the max. Note that I'm *not* asking for the Kemper to somehow blend several profiles done over these ranges (tho that would be optimum, of course) but just to calibrate the Gain knob to what I see as the amp's useful range. As it is, I find that loading a profile (all, not just mine) comes up to some seemingly arbitrary Gain - and that I typically find only a few dots worth of Gain (which may be very far from the initial setting) to be musically useful for my purposes with a given guitar. Typically I'm trying to dial in somewhere in the breakup zone, with crunch that came from the *amp*, not Kemper's synthetic distortion, which, while much better than Line6's, is still too fizzy, IMHO.

    5. If it's a useful "refinement" why not let it be done later, if need be, e.g., with different guitars, players, to taste, etc? AFAIK you can only refine when creating the profile initially


    I've been able to make some great profiles by trial-and-error, but it's frustrating and time-consuming. The Kemper doc on profiling is far too vague, IMHO, on the entire process. And given that there are over 14,000 user-contributed profiles, yet the widely-shared opinion is that the vast majority are crap, methinks that the process and/or clarity on the points above (among others) would go a long way to making the Kemper a better, more useful product, and b) get us all a lot better and more useful collection of profiles on Rig Exchange.


    Thoughts?


    -- jdm

  • This has come up through the years the forum has existed, from what I recall. Most “official” advice I’ve seen other than the manual has been CK saying that one should play “in the areas where the differences become apparent”.


    I’ve had a few cases where refining made things worse. Usually it helps the profile get closer. Typically this involves the bass response mimicking the source better. Not “perfect”, but better.


    Now I myself would not expect much more advice from kemper on refining than what’s currently available. But there’s people who’ve been profiling for years who have their own takes on what tends to work better. Perhaps someone chimes in.