Posts by dhodgson

    +1 to that. That's actually the way I've been doing it, clumsy as it is. I've also found that sometimes the online editor won't allow you to modify the author field when a cab isn't present. Sounds silly, but the KPA lets you do it so the online editor should also.


    -djh

    I'd been thinking about this too. If there really is an "ideal" way perform the rig refinement step, couldn't this be reduced down to a sampled factory performance burned into the firmware and just played out? Or is there something "organic" going on here that can't be performed in a one-size-fits all fashion?


    -djh

    Kojak,


    I remembered when the POD Pro came out; it was the first real change in the Line6 algorithms, which went along with the change in DSP they were using at the time. I got it, and hated it! It just didn't have the same squishy compression for me that made playing so easy. Later on, after having gone through several tube pre's, I came to appreciate the dynamics of my real tube amps a bit more, particularly in terms of getting things to set better in the mix. But, still I have to have a certain response curve to the dynamics, or it feels like the amp is fighting me - and who wants that?


    Much, much experimenting later I got into balancing the feel and tone of my amps through overdrive pedals and stomp EQ's which I'd match to the amp, and between the two I think I got pretty successful at feeling out the interactions between both. What I consider to be the right "feel" is probably about 20-25% from how I have my pedal chain configured and gain-staged. The feel difference between amps don't worry me as much as they used to; as long as the amp gets me mostly there I know I can depend on my stomps to handle the rest.


    My first feelings with the KPA were kinda similar to yours, I think; the sound's there, but the response curve is flatter - not quite as compressive, a little bit of the POD Pro feel sneaking in there. But the fun of the KPA is that you're not stuck with this; between the Distortion Sense, Definition, Power Sagging, Pick and Compressor controls there's a lot of tweakability on tap and I feel myself "settling in" more as my familiarity increases with the KPA. I've also found that as tempting as it is to profile your amp with an overdrive in the chain, more of the 'feel' is preserved if you don't - and I haven't even played around with the stomp section overdrives yet. Anyhow, that's my "new user" impression.


    The "main amp EQ is always post" thing - I can definitely understand that, particularly if all the profiles you're using were created by others as this is one aspect that you can't conveniently re-tweak after the fact. I don't doubt we'll be seeing more in this area, DSP willing because Chris's saturation engine seems almost limitless in this regard.


    Anyway, our favorite amps are like old couches - the new one never feels quite the same, and that's understandable! Making a truly "one size fits all" product is a mighty task, and there's room for only so many knobs. Where's the MarkIV graphic EQ? Dangit!


    -djh

    That's why I performed two different experiments there, guitarnet70. One was to determine whether extracting the cab from a profiled cabless rig (thanks for documenting that technique, sheguitarplayer) was equivalent to profiling a bare wire (which is what every linear amp tries to be - a straight wire with gain.) We've also been encouraged elsewhere by Kemper to profile cabless amps, so in the absence of a manufacturer-provided "no cab" cab profile it seemed worth trying.


    Not to get sidetracked, but the point of even mentioning the loopback cab story was just to confirm my suspicion that something might be off, which I found curious. Additionally, when the cabless amp sounded brighter than the same cabless amp with the loopback cab, it was just grist for the mill. Or not, of course - I haven't done any more investigation since posting this. It may be nothing, so I'm just tossing it out there.


    -djh

    Doesn't matter what Kemper uses - I tried to preempt this response knowing it was the probably the first one I'd receive. :) I blew it. :) I tried to use the IR example as a guide for how one knows the time cap in other modelers that use IRs.


    This is signals and systems problem - it's a math thang - there is a time limit to the "response" the Kemper can capture.


    The response here being the "response of the reference amplifier" - I'll put in it quotes to separate it from the response in Impulse _Response_.

    Ah signals and systems, a man after my own heart. What's your background, Blewis?


    Gazing into my crystal ball, I suspect that the test tone section responsible for determining the cabinet response begins about two seconds past the end of "UFO" swells at the beginning. You've got an impulse, followed by a very brief (1.5mS) step function followed by about 240mS of silence then fourteen swells of a low level surflike noise lasting for about 13.5 seconds. The higher gain the amp, the quieter this section so it could be a statistical MLS-type of chip function that's trying to squeak by for measurement purposes without being clipped by the amp. So, whatever's going on it looks pretty clever and a likely good method for keeping all or part of the "room tone" out of the analysis of the profiled speaker's response. That would put a cap on far-field miking, but if that's your thing I think your best bet would be to an take an impulse of your room and convolve with that later at mix time.


    All of this is just conjecture, of course. ;)


    -djh

    Simpler than that. My E570 is strictly a tube pre, no power amp. Likewise, my "cab" is just a Palmer ADIG-ST. Line level signals all around, Kemper -> ENGL -> Palmer -> Kemper. Both the ENGL and the Palmer have defeat knobs that allow you to take them out of the circuit, making it easy to profile them together or separately. In this experiment however I profiled the ENGL directly (Kemper -> Engl -> Kemper) just in order to be extra clean about it.


    -djh

    OK, this is what happens when I start pushing things. Once I figured out how to extract the cabs from profiled rigs into separate cabs and export those as tarball backups, I started thinking:


    A) "I wonder if the KPA's cab profiling results might be even better if performed separately from the amp."


    Likewise,


    B) "I wonder if the KPA's amp profiling results might be even better if performed separately from the cab."


    If either A or B were true, then that might be a case for profiling the two items separately and combining them later, saving the result out as a new rig. Hi fi! So, I tried this, a task made easier by the "Preamp Defeat" switch on my ENGL. I also created a "null cab" to play with by plugging a loopback cable between the send and return and profiling that. It worked, sounded fine.


    What I discovered though, was that the new rig I created, namely Profile(amp) + Profile(cab) was noticeably darker sounding than Profile(amp+cab). Hmm. Then, curious, I reloaded the (harsh and buzzy) no-cab amp profiled in Step B and then loaded the "null cab" into it expecting not to hear any difference. But there was - it still sounded harsh and buzzy, but some of the edge had gone.


    Has anyone else tried this? Should we always be profiling the amps and cabs simultaneously for best effect, or is there some subtle bug?


    -djh

    New profile up. Best to go with what you know, and this setup has been my hard rock amp of choice for some time.


    Amp: ENGL 570 SE, Ch 3 (Lead II)
    Cab: Palmer ADIG-ST (Brown mode)
    Overdrive: Mad Professor Sweet Honey


    I've uploaded two different versions of this profile, one with the Sweet Honey on and the other off. Use whatever works best for you. My second favorite overdrive, the RockBox Boiling Point, didn't profile as accurately, possibly due to its Tube Screamer ancestry - and a third was rejected for being too noisy! That said, the KPA's profiles are, if anything, quieter than the original signal chains which is a real plus.


    This profile is a bit of a compromise, since I decided to leave off my usual post-Palmer rack EQ and just tweak the rigs after the fact. I hope you find it to be useful and practical, good luck and please let me know if it can be made even bettah!


    Cheers,


    -djh

    Ack! I spent a couple hours trying to conjure up a wah as well. Not claiming these settings are optimum, just that it's sensible with an M-Audio EX-P Expression Pedal. Also experimented with using a tip-and-sleeve Line6 expression pedal for this purpose, (their Stompbox Modelers use these for wah) but to no avail. Stick with the Roland (Type 1) or Yamaha (Type 2)-style TRS expression pedals.


    System Settings
    Page 4
    ...1: Expr. (Type 1) Wah Wah


    Wah Wah Stomp Settings
    Page 1:
    ...Manual 3.5
    ...Peak 8.0
    ...Pedal Range + 27%
    ...Peak Range - 44%
    Page 2:
    ...Mix 100%
    ...Volume + 5.0
    ...Pedal Mode Bypass @ Heel


    -djh

    I'm with CSGBAND here. I'm staring at my ENGL E-570 and thinking that there's three channels, each with several tonal and gain modifier buttons - at least a dozen different combinations worth profiling here. And then there's the question of EQ - should I go with what sounds best, or all knobs flat? And then there's the cab issues; leave it out, throw on the Palmer, or mike up the cab? Attempt to profile the overdrive chain? The post-amp channel strip EQ?


    So many variables - doing this right takes thought and effort. I worry that we're going to be drowning in amateur-grade profiles as the amp increases in popularity, and this could lend a bad rep of the "Millions of sounds, all mediocre!" kind of thing. No doubt we're going to see a handful of professional producer-engineers trying to make a go of being the Redwirez of profile suppliers at some point. Profile rankings on the Exchange page might help sort things a bit, although crowdsourced opinions are only as strong as the background of the people evaluating, of course.


    Thoughts?


    -djh

    I doubt it. I don't have a copy of the full profiling test tone sequence but if there were any crossing sine tones being generated (as has been mentioned) that is when it would happen - not during the refining step. The refining step has more to do with transient analysis, which gets mentioned somewhere in one of the Kemper publications.


    -djh

    Well,


    It would have been a treat if Mr. Kemper had actually decided to comment on my patent link post, but I can’t blame him if he decides not to because that’s the smart money call. I’ve been down the patent road a couple times, and I glad there’s an IP attorney among us (Hi, Adfinitum) here to comment.


    For anyone who’s curious, Line 6’s 1998 modelling patent (http://bit.ly/y90f5u) is pretty much the Mother Of All Amp Modeling Patents. And Douglas L. Jackson’s 2003 patent (previously mentioned, http://bit.ly/yBZtJF) is my nominee for the Mother Of All Amp Profiling Patents. For a while I wondered if Chris’ KPA was actually that design made real, but since it does not incorporate intermodulation distortion analysis I’m optimistic that the KPA is on to something original.


    Now, Cristoph’s patent link was German but fortunately he’s applied in the U.S. as well, and you can read an English version at http://bit.ly/A1rVZD. This application (originally filed in 2007) is still pending.


    In case you’re wondering, I’m a musical patent collector of sorts (I keep a list of these on my Website) because patents tell the other half of the story, namely the part that’s left out of AES papers and design conferences like DAFX. But, they’re admittedly difficult reading; you’re dealing with inventors who are trying to obscure their inventions by trying to cast overly wide nets over their claims (i.e, over-generalizing things through abstraction) or trying to cast overly narrow nets by revealing as little as they can get away with while being generally unenlightening about things.


    But the point is, where utility patents like this are concerned, is to Reveal As Much Prior Art As You Can In Order To Explain What Makes Your Invention Different while Explaining In Detail How To Build A Particular (Physical, Material) Something. Patents don’t exist to protect Ideas – they exist to protect their Implementations. So, whenever Patent claims are brought up by anyone, I ask people to take nothing at face value and digest the actual works in question. And as Adfinitum mentioned, he’s right – I’m not aware of Line6 having ever litigated their modeling patent either, and frankly it makes So Many Claims that I’m not sure if they ever really could. This is the Keep It Simple Stupid rule of patents – more complexity means more holes for others to attack. But at least they’re practical claims, and the same goes for the Jackson patent.


    I’m not at all concerned with any potential patent claims coming from the Axe-FX front, although I would be unimpressed by Cliff Chase as a businessman if he didn't try to cast some FUD in this area. Most people think that patents are an ironclad guarantee of inventor rights, but really they’re not – the “guarantee” comes through active litigation, and ONLY active litigation. Patents themselves are mainly proof that the inventor claimed something by a certain date, and classified it accurately – beyond that, it’s Game On which why the only people who benefit from patent litigation are Large Companies with Deep Pockets.


    -djh

    Awesome stuff. Welcome to the forum, I'm up in Santa Rosa. We should think about getting a Bay Area Kemper User Group together (that's the normal thing us Bay Area and Silicon Valley folk do isn't it?).

    I'm up for that. We could start something on Meetup.com if there's enough turnout. Send me an email Per, and maybe we can, err, meet up.


    -djh