Can't get my profiles close enough

    Quote from fluff191

    I was recently at a guitar show and a guy there had a Kemper rig set up and he was getting ready to start selling his profiles to the public soon. By trade, he is a sound engineer for a living (not a studio engineer, but a guy who gets paid to analyze the resonant frequencies of various products and those sorts of things).


    We got into a discussion about sampling rates and after testing the Kemper for several months, his conclusion was that the reason the low end lacks on the profiles is because the sampling rate is simply too low to properly translate the low end. For the high end, its perfect (as shown in the EQ graphs). But for the low end the sampling rate of the profiler itself needs to be much higher.


    I dont know how true any of that is, because well, I am not an engineer. Food for thought though.


    The food for thought is how this guy makes a living as a "sound engineer"!

    Look at the pic I posted, the low end is damn near perfect, the high end is off a little but it's really close too.


    I can hear a low sampling rate in an MP3 - but it's never on the low end, you usually hear artifacts and weirdness in high hats and percussion. Just my experience anyways.


    Quote from fluff191

    I was recently at a guitar show and a guy there had a Kemper rig set up and he was getting ready to start selling his profiles to the public soon. By trade, he is a sound engineer for a living (not a studio engineer, but a guy who gets paid to analyze the resonant frequencies of various products and those sorts of things).


    We got into a discussion about sampling rates and after testing the Kemper for several months, his conclusion was that the reason the low end lacks on the profiles is because the sampling rate is simply too low to properly translate the low end. For the high end, its perfect (as shown in the EQ graphs). But for the low end the sampling rate of the profiler itself needs to be much higher.


    I dont know how true any of that is, because well, I am not an engineer. Food for thought though.

    okstrat do you profile straight into the kemper or what is your path? Your graph follows it closer than the others shown so is it possibly the style of amp profile? The previous ones have been Orange and Marshall which are more vintage. I don't know that it would but maybe amp type makes a difference?? Either way I appreciate more people chiming in so there is some discussion and possible tips.

    "More Guitar in the Monitors" :thumbup:

    This graph by okstrat is very interesting he was able to get the low end right. This means theres something to be learned here as was mentionned. Was it an amp that was easier to profile then some others, better profilling/refining etc i duno but its worth looking into it! Thanks for chiming in.

    Quote from okstrat

    I just did this as a test in ozone - top is kemper profile, bottom is a boogie I was profiling... they aren't 100% but pretty damned close!

    That's VERY close indeed, yes. However, from the waveforms of the tracks, it looks basically that it's just one chord struck and sustaining for around 9-10 seconds, correct? I get similar results in cases like that - the real difference start showing with more playing, some palm mutes included and so on. Could you try the same thing again with more "playing" and not just a sustained chord? Best is of course if you first record a DI signal and use that when you compare, so you have the same signal in both cases.

    Quote from okstrat

    I just did this as a test in ozone - top is kemper profile, bottom is a boogie I was profiling... they aren't 100% but pretty damned close!


    [Blocked Image: https://fbcdn-sphotos-e-a.akamaihd.net/hphotos-ak-prn2/1149365_10200721040047580_1707167739_o.jpg]



    wow I've done a few tests with ozone and mine have never been that exact (still close)


    could you tell us how maybe what you do to refine? How many times do you refine? 1 or 2 or more?


    Would be great if someone who gets super close results could send a DI of the stuff they play and then some users could run it thru to see how it compares.

    Here's something that bothers me, why is it that so many profiles have a lot of bottom end below 100 Hz? I set up my monitor out with bass at -1.0 because according to my RTA that's the closest to a flat response in the bass area for the DXR10, and even then I have to set the bass in the profile's tonestack between -1.0 and -3.0... Weird. Am I doing something wrong here? ?(

    Quote from Deny

    Here's something that bothers me, why is it that so many profiles have a lot of bottom end below 100 Hz? I set up my monitor out with bass at -1.0 because according to my RTA that's the closest to a flat response in the bass area for the DXR10, and even then I have to set the bass in the profile's tonestack between -1.0 and -3.0... Weird. Am I doing something wrong here? ?(

    I know your this problem, too. But it seems to be typical for the KPA to give a fulminant bottom end. Imo that's not a problem cause you can control this feature i.e. by using the graph-eq in combination with the compressor (or vice-versa) to correct weird differences in sounding. Sometimes the fat bottom is a little nervy but otherwise I don't want to miss it. :D

    I've got same problem too.
    Try profiling a scooped, "modern" type of bass amp with lots of low and high end, and you will have big problems of even getting close to the original.

    Just happened to find tgs's posts on Ultimatemetal, which redirected me on this old thread, I had forgotten about. I could not post ATT, since I was not a registered user yet.


    While the thread seems dead, I'd like to thank OP for his dedication, his willingness to be of help and his approach. And, last but not least, for his manners. :thumbup:
    I hope this will help Kemper to improve their product!


    I also couldn't disagree more with those who tried to minimize, in several ways. As I always say, things don't have to be perfect in order for us to love them :)


    Peace and axes

    Just read through this whole post. One thing that I noticed when watching review videos before purchasing was that the Kemper profiles often sounded a tad thinner than the reference, which is especially noticeable for chug-a-chug metal. Since getting the unit, I haven't actually tried to profile anything (who wants Pod HD rigs?!), but I am very pleased with the low end of the tone.


    To contribute, for those who want to perform the EQ trick tgs describes using the Kemper's Studio EQ, keep in mind that the low and high shelves have rather round knees. I haven't sat down with the pink/white noise and visually checked it out yet, but I know I can set the high shelf to 33 kHZ and still hear it affect the tone (if you don't understand, human hearing doesn't go much higher than 20 kHZ). So my guess would be to set the low shelf to about 50 HZ and set the gain down to about -6 db. You won't match the slope perfectly, but with those frequencies being so low and softer than the peak frequencies of the tone, it'll be difficult to hear. You may get better results using a lower frequency in combination with a lower gain level, for ex. 30 HZ and -10 db.


    Then you can just set one of the mid peaks to dial up the 110 HZ.


    Trying to acheive tgs's EQ solution using the BMTP tone stack is lunacy. There are a couple posts on this forum that demonstrate the frequency ranges they cover, and they are VERY wide compared to what tgs is talking about. I think Bass will affect everything from 300 HZ and lower. Mids affects from like 200 - 1200 HZ.


    To me the crux of the issue though is not the frequency response but the distortion characteristics. it'd be simple enough to compensate the amp settings or some post-eq to force the profiler to get that frequency response. But the roughness of the bass is why profiling is so great - you can snatch up those quirky nuances that a budget modeler just ignores.


    ...as to the "sound engineer" who believes it is due to the Kemper's low sampling rate, this guy obviously knows nothing about DSP. Any frequencies below 200 HZ can be accurately (i mean like 100 freakin %) represented digitally using a 400 HZ sample rate. This is why 44 kHZ is a common rate - it is 2x the highest frequency a human can hear. In theory, a human could never hear the difference between 44 kHZ and 96 kHZ sample rates. In practice, devices are subject to deviation from reference and higher resolution tends to capture things more accurately. Since guitar is primarily a midrange instrument, I think 44 kHZ is more than accurate enough. In any case, he has it backwards - HIGH frequency sounds require HIGHER sample rates to accurately represent. See Nyquist frequency.


    Sampling rate in DSP is a totally different concept than bitrate in mp3 compression. AFAIK, the Kemper isn't using a lossy compression to perform profiling. I doubt it stores any raw audio signals, let alone compresses them.