Best profiling gain setting

  • The Profiler Model referred to in this thread is ...
    ☑️ Profiler Stage

    1) WHAT IS THIS POST ABOUT
    Since trying to create my own version of amp cloning, I have come to an understanding of how the Kemper may go about making a profile™.

    My process is to do a frequency sweep of the amplifier applying a high volume signal. Then sweep again with a very low volume signal. The difference of these two sweeps will reveal the input EQ of the amplifier. This EQ is where most of the amps sound comes from. IMO, It is the critical piece of cloning an amp.

    Over the years of making profiles in the Kemper, I always had better results if the amplifiers gain was set somewhere near break up. No gain sounded bassy and full gain sounded too bassy as well as undefined. After creating my cloning process, it became obvious why that was.

    Creating profiles near breakup usually gives the most amount of dynamic range to the resulting INPUT EQ result.

    2) HOW DOES MY CLONING PROCESS WORK
    - The input EQ is the fingerprint of the amp.
    - At a low input volume (gain) the resulting freq sweep shows the EQ of the entire amp (input eq, output eq, speaker, mic, etc).
    - At a high input volume (gain) the resulting freq sweep shows the EQ of the entire amp minus the input EQ (output eq, speaker, mic, etc).
    - Doing math on these two sweeps reveals the input EQ. The more dynamic range in this result the more accurate the cloning.

    At this point, I am assuming the Kemper does something very similar when doing its profiling procedure.

    3) WHAT ARE THE RESULTS
    As an example, I made several sweeps of a Kemper profile at varying gain settings. I chose the Lars Luettge profile TH - Mars VM410 20 to be the sample profile.

    WHY I CHOSE THIS PROFILE
    - During my cloning process this amp sounded really good.
    - The higher the gain the more pronounced the Input EQ is. Clean amps have flatter responses. High gain has more of a bandpass peaked response.
    - This profile is in the included Rig Packs from Rig Manager.
    - Everyone has this profile and can get a feel for what the perfect gain setting is by playing the profile at each gain setting.

    I swept this profile from a gain of 1.0-5.5. The resulting sweeps are shown here. I stopped at 5.5 because I can only attach 10 files to this post.

    Each image shows three sweeps. The highest sweep is the high volume (gain) sweep. The next lower sweep is the low volume sweep. The sweep below 0 dB is the resulting sweep that shows the input EQ derived from the upper sweeps. In this example the input sweeps have a little over 20 dB dynamic range.


    From the resulting sweeps we can see that the cloning process worked well thru the whole range. However, the best results were from gain settings 2.0 to 3.5. These sweeps had the most dynamic range and therefore the most accurate results.

    We can see that from the low gain sweeps 1.0 and 1.5, that the dynamic range is lower. This results in a bassy profile. More bass is present in the resulting EQ. The actual input EQ is obscured by the dynamic range noise floor of -40 dB.

    The higher gain sweeps from 4.0 to 5.5 have hit the ceiling of our dynamic range. This again results in a bassy response because the higher peaks are being clipped at 0 dB. But the overall response is now being flattened as well. This results in, not only being bassy but, a wrong sounding clone. The mids and highs are flat losing the fingerprint of the amp.

    The sweeps from gains of 2.5 to 3.5 have the greatest dynamic range and are more accurate at capturing the overall sound of the amp.

    4) WHAT DOES THE KEMPER DO?
    I have not investigated what the Kemper does. I assume it does something very similar. My cloning tests used fixed input values of 0 dB and -40 dB. This provides a maximum of 40 dB dynamic range in my clones. The Kemper may do a high volume sweep and then adjust its low volume sweep until noise becomes an issue. This means the Kemper may provide more/less dynamic range depending on the situation.

    You can see from all of the above sweeps, each provides a useful clone. Some are more bassy than others and some have a flatter high freq response. The Kemper gives you a DEFINITION control in its AMPLIFIER section so the erroneous bass response can be adjusted. The DEF control adjusts the most common flaw in the process.

    5) FINAL THOUGHTS
    These tests show what works for a high gain amplifier. A low gain amplifier should work the same but is not presented here.

    To get the best profiles, the amplifiers gain should be set to a value that sounds similar to the example profile at a gain of 2.5 to 3.5.

    STEPS TO TAKE
    - Set the amplfier to the gain setting your profile should sound like.
    - Adjust the EQ of the amp for the best possible sound.
    - Reduce the gain to a value that sounds similar to the example profile at a gain of 3.0.
    - Profile the amp.
    - After the profile is complete adjust the Kemper Amp gain and Definition controls to match your original amps sound.

    6) DOWN THE RABBIT HOLE
    How does this process work? When sweeping at a low input volume, the signal is well below the tube clipping point. As we increase the input volume, the tubes begin to clip. This clipping limits how loud the signal can get. Which means, it flattens the input EQ response out. The more volume we use, the flatter the input response gets until it eventually becomes a completely flat response. By comparing the lowest and highest input volumes you can calculate the exact input EQ response.

    For best results you want the low volume sweep to never be clipped by the tubes. You want the high volume sweep to be very clipped by the tubes.

    Edited once, last by RosboneMako (January 31, 2025 at 9:44 PM).

  • You seem to be having fun :). Sounds to me though that you are doing EQ matching only? Are you in any way trying to emulate the nature of any distortion going on?

    Kemper PowerRack |Kemper Stage| Rivera 4x12 V30 cab | Yamaha DXR10 pair | UA Apollo Twin Duo | Adam A7X | Cubase DAW
    Fender Telecaster 62 re-issue chambered mahogany | Kramer! (1988 or so...) | Gibson Les Paul R7 | Fender Stratocaster HBS-1 Classic Relic Custom Shop | LTD EC-1000 Evertune | 1988 Desert Yellow JEM

  • You seem to be having fun :). Sounds to me though that you are doing EQ matching only? Are you in any way trying to emulate the nature of any distortion going on?

    The point of the post is to get better Kemper profiles. EQ matching would be something that gives you the CAB portion of a profile. This is all about the amps input EQ.

    The input EQ is the hidden or secret part of profiling that creates most of the sound. This post explains one way to obtain the input EQ, so we can better understand how to get the best profiles from the Kemper.

    An amps sound comes from the input EQ, the gain structure, and final eq/speaker/mic. I would put the Input EQ at 65% of the sound, gain structure at 10%, and the final EQ/speaker/etc at 25%. So tone matching the final EQ only gives you that last 25%.

    You can see that the input frequency response of a 5153 profile is nothing like a Fender profile. If you applied a lot of gain to the Fenders response it would be flubby bassy garbage. The bass would overpower the highs completely.


    GAIN STRUCTURE
    I have not worked on the nature of the tubes/gain stages etc. I have just now only begun to think about that. In my mind it is not a crucial piece in high gain amps, which is what interests me most. I have tried several different clipping techniques in my VST and they all sound very similar at high gain.

    I also watched Jim Lills YouTube video about what makes an amp sound like it does and in his tests the tubes/diodes did not make a huge difference either.

    But I am guessing the clipping method makes some difference because the Kemper works it out when profiling. And that is why the Kemper is so great. They have done the due diligence to make the best possible sound that can be had. They didnt stop at 80% there, they went the whole way.

    For my VST I settled on hard clipping the high gain stuff and using hyperbolic Tangent function for soft clipping. In the VST you can adjust between those two methods. But Kemper is doing more. They are looking at the whole signal from zero to full scale. Probably mapping it to a look up table or something.


    HAVING FUN
    Am I having fun? I am having a ton of fun. I went from just trying to make a distortion pedal to making an amp cloning VST with effects, etc. I only planned on the pedal.

    I figured most dist or OD pedals are a bandpass filter around 700 Hz. So I made a VST with an adjustable center freq and Q. Then hard clipped it. Boom distortion pedal. Took minutes to write. But to really hear how it sounds I needed to pass it thru a speaker. Then how IRs work just popped into my head like someone explained it to me in my sleep. Next thing you know I have effects, etc.

    I was totally happy with the VST at that point. Then one day while thinking about the next pedal I would make I realized that the input EQ goes flat as the gain goes up and boom, now I understood how amp captures are made. I already had IR code written for the speaker/cab IR, so again it took minutes to get the amp cloning working.

    Since this is just something for me to do when I am bored, I may or may not go much further. Time will tell. I may try to spend more time getting Kemper to rework their tube code cause it sounds too compressed to me ;) And that will lead me into looking at gain stuff so I can explain it better.

  • “If you applied a lot of gain to the Fenders response it would be flubby bassy garbage. The bass would overpower the highs completely.”

    Doesn’t that assume that gain is flat? I would have assume that power tubes (especially depending on whether it’s el84’s, 34’s, 6v6 or 6L6) are not flat and amp designers (at least the good ones) look at the eq curve of various tubes at least to some degree and its impact on overall eq as a contributor to tone.

    Your work is fascinating! Thanks.

  • “If you applied a lot of gain to the Fenders response it would be flubby bassy garbage. The bass would overpower the highs completely.”

    Doesn’t that assume that gain is flat? I would have assume that power tubes (especially depending on whether it’s el84’s, 34’s, 6v6 or 6L6) are not flat and amp designers (at least the good ones) look at the eq curve of various tubes at least to some degree and its impact on overall eq as a contributor to tone.

    Your work is fascinating! Thanks.

    I know next to nothing about tubes. As a guess I would say tubes themselves are pretty flat relatively. All of the passive components required to get the tube working probably have a much larger effect on frequency response. And it is generically those components that create the input EQ voice to a large extent.

    My comment about the bassy stuff was to say if you put that much bass into a gain circuit, the highs will disappear. The lows will become a square wave and the highs will no longer exist. They will be replaced by the harmonics of the square wave (clipping). And it will be muddy farty garbage.

    I am no expert. Just a dumb guy playing with code.

  • You can see that the input frequency response of a 5153 profile is nothing like a Fender profile. If you applied a lot of gain to the Fenders response it would be flubby bassy garbage. The bass would overpower the highs completely.

    That is why Mesa Mark series (Fenders on steroids) get their high gain sounds with the bass set below 2 or even off. This is also why the GEQ was such a common option (now pretty much standard) on the old Marks series - to add back bass after the distortion stages.

    Having said that, try a Deluxe Reverb liquid profile. Crank the gain to 7 or higher and roll the bass almost off, mids quite low and treble high. You get a surprisingly good overdrive rock sound. (I got this from a video by Justin Ostrander where he used a real Deluxe).

    I would have assume that power tubes (especially depending on whether it’s el84’s, 34’s, 6v6 or 6L6) are not flat and amp designers (at least the good ones) look at the eq curve of various tubes at least to some degree and its impact on overall eq as a contributor to tone.

    I have a THD B9Valve which is a great little amp. The main USP is that it can accept almost any valve type in the preamp and poweramp sections. You can even have different valves on each side of the power amp (1 EL34 paired with 1 6L6 fpr example). I though this would give a mass9sange pf tone opti0ns when I bought it. However, after swapping valve and recording and comparing the results the differences were “subtle” to say the least.