Making sense of clean sense (audio clips added)

  • i know you have much better and refined ears than me but I don’t hear any clear difference between any of the three clips. If there is a difference it is relatively subtle. However, in you example the difference was vert clear. I don’t doubt the difference is real in your example but I’m struggling to explain why it should be dramatic wheres it is barely perceptible on any profiles I tried. I wonder If there is something else in your settings that is having an impact as well as clean sens? Or whether there is something specific to that profile? Presumably you get the same effect with other profiles too though which is why I did the test with a factory content profile.


    I haven’t had a chance to do any further tests yet but I wonder if Per’s observation could be explained by something other than clean sense. The Kemper does a lot of level matching in the background so maybe clean sense is affecting the signal at more than one point in the chain.

  • i know you have much better and refined ears than me but I don’t hear any clear difference between any of the three clips. If there is a difference it is relatively subtle. However, in you example the difference was vert clear. I don’t doubt the difference is real in your example but I’m struggling to explain why it should be dramatic wheres it is barely perceptible on any profiles I tried. I wonder If there is something else in your settings that is having an impact as well as clean sens? Or whether there is something specific to that profile? Presumably you get the same effect with other profiles too though which is why I did the test with a factory content profile.


    I haven’t had a chance to do any further tests yet but I wonder if Per’s observation could be explained by something other than clean sense. The Kemper does a lot of level matching in the background so maybe clean sense is affecting the signal at more than one point in the chain.


    Alan, in your clips, I notice the same thing as mine with the most extreme examples where there is less compression or smoothness in the lower (higher?) settings. This makes the first clip sound drier and the second sound more liquid, if that makes sense to you.


    As far as I know, the only use of clean sense is to level clean tones and distorted tones. That is the explanation given by Kemper Amps.


    However, there are some niggling doubts:

    1) The fact that the input LED should change at all when reducing clean sense. What is the explanation for this behaviour? (Indeed, what is the purpose of the LED at all)


    2) The fact that there is a difference (subtle or not) with distorted profiles when changing clean sense.


    I'll try and demonstrate this with another clip tonight. Will also try to video the change in setting so that it's all above board.

  • yes, it is confusing in many respects.


    i completely agree with you point 1 above. If the clean sens doesn’t affect gain at all why does the input LED change and if it doesn’t then what is the point of even having an input LED.


    Maybe it it the same as the standby switch on tube amps. It is totally unnecessary for the protection of tubes. Early amps never had it. It was added to allow the amp to be silenced during breaks etc. However, people now believe it is absolutely necessary and shy away from buying a valve amp that doesn’t have a standby switch. Maybe Kemper got the same feedback when user testing the first versions of the KPA. If enough users felt scared by lack of any ability to measure input signal, maybe they just added a dummy LED to pacify them (also like telling the bass played to tweak the magic knob or fader on the mixing desk. The one that isn’t actually doing any :D)


    it would be really great if Burkhard or ckemper could answer this clearly once and for all though.

  • I've just carried out another very quick test on Per's scenario.


    Setup a really dirty rig (used the same OLA rig as above because it's in everyone's RM already).


    Place the default Green Scream Settings in Slot A and Slot X.


    Use just the Stack ,with both slots turned off, and move Clean Sens from -12db to +12db. As Per pointed out this makes no difference to the sound at all (or at most a very subtle difference).


    Turn on Slot A and move Clean Sens as before. Again little or no difference.


    Do the same but turn on Slot X instead of Slot A. Again little or no difference.


    Now load a clean rig for ease I just picked the cleanest rig in the same folder OLA - Edvaha Clean. And do the same thing.


    This time with only the amp active there is a substantial volume difference.


    But with Slot A active there is no difference. It would therefore appear that the introduction of distortion in Slot A is removing the effect of Clean Sens because the KPA now knows it is dealing with a distorted sound (?)


    Now with Slot X active instead. There is a marked change in volume and gain. Does this suggest that the Clean Sens is working on the output? I don't think so. To me it seems like the KPA knows that it was dealing with a clean sound and behaved accordingly early in the chain. However, as clean sens was not doing anything after the Stack section the OD in Slot X behaves as you would expect an OD to behave when hit with a louder signal.


    To take this even further. If you turn off everything including the Stack. Clean Sens still makes a massive difference. This would seem to rule out the idea that it is being applied to the output section of the Stack block too, or performing just a parallel path type operation between the clean and distorted parts of a profile (there is no profile being blended).


    Having said that, if you turn off the stack and apply the OD in either of the slots it still does nothing so the KPA is clearly able to detect that it is a distorted signal and apply its compensation processing in many situations. If you apply an effect other than Distortion in slot A and X (such as Reverb) the behaviour is the same regardless of position in the chain. This suggests that it is the inclusion of a Distorion FX that is making the difference NOT that clean sens is working after the Stack section.


    My guess is that Clean Sens is indeed performing an input gain task of some sort but this is then going through some sort of make up gain process at some later point in the signal path and this make up is dependent on various other factorat that point. I don't see any way to reach the conclusion that even though it it in the input section this is confusing because it doesn't apply to the input but rather the output. That doesn't make sense from the tests I've described.


    I would truly love to hear someone at Kemper give as detailed an explanation as possible without giving anything away trade secrets. :)

  • I don’t think they will though.


    My guess is that if ckemper  burkard or any of the team are reading this thread they are probably having too much fun laughing at how wrong all our guesses are ?


    Nope, they'll probably chime in with exactly what the manual says, i.e. clean sense only serves to balance clean and distorted rig volumes.


    They need to listen to my clips though and try it for themselves. Something definitely not adding up.


    EDIT:


    Thinking out aloud. What if you have to first adjust clean sense to balance volumes, then adjust distorted sense in order to ensure that there's no change in "tone/sustain/saturation"?


    Of course, do make a note, if this is the case, out goes the argument that clean sense does not affect distorted tones.

  • Despite the black-box nature of the Kemper I think it really is as written in the manual based on my experiments.


    The most difficult to grasp part is how Clean Sens is utilized in the amp block. If you take it as written, it really is operating in a similar fashion to the loop mix pot on most amps, it mixes between a parallel clean path that doesn't go through the pre-amp and the output of the amp. At max therefore it should make the Clean signal twice as loud.


    The thing that makes this more complex is that this is also attenuated by the "Gain" control on the amp itself. So when Gain is at 10 that parallel clean path is completely muted, at 0 the parallel clean path has a value of 100%.


    The tests I did in the other thread bear this out. With extreme settings the blend itself is audible as you adjust the Gain dial.


    The reason I suggest using a distortion stomp to test with is that this is reactive to volume. Most other FX aren't. This makes it very easy to audibly tell whether the signal reaching that stomp is lower or higher in volume than before as you modify the Sens values.


    I continue to believe that Clean Sens is just the volume/mix of a clean signal in the amp block output, and that Distortion Sens is just a modifier for the Gain dial position on the raw amp (not including the clean blend which I believe is locked to the actual gain dial value itself).

    Therefore I currently believe this is the case :

    PreAmp gain level = Gain knob + Distortion Sens value.

    Amp output signal = Mix(Amp Block Input signal * Gain knob setting, with PreAmp output, by Clean Sens setting).


    I believe that if you utilize a loop then you can substitute loop for amp and preamp in the second line.


    In any case this means that Clean Sens will result in a net volume gain from the loop/amp block when the gain setting is lower. Distortion Sens will not result in any change except to the drive level of the PreAmp.


    The input LED is anyones guess, I don't have my Kemper next to me to test out but it could either be just the raw input signal multiplied by the Clean Sens setting or it could be the the magnitude of the signal going into the amp block or loop (if it comes first) after it has been multiplied by the Clean Sens setting. In either case it doesn't actually represent the magnitude of the signal that actually goes into the signal chain which remains constant.


    To sum up - the Kemper does not have a control for attenuating input.

  • To sum up - the Kemper does not have a control for attenuating input.


    I don't know, Per. There's definitely a difference with different clean sense settings. Not so much tone, but more in terms of the smoothness or compression or saturation of distorted profiles. Listen to the first part of this video, which is at -12, now compare it to the second or third part, which are +12 and 0 respectively.



    Here's a lower gain tone:


    External Content www.youtube.com
    Content embedded from external sources will not be displayed without your consent.
    Through the activation of external content, you agree that personal data may be transferred to third party platforms. We have provided more information on this in our privacy policy.



    Here's a high gain tone:


    External Content www.youtube.com
    Content embedded from external sources will not be displayed without your consent.
    Through the activation of external content, you agree that personal data may be transferred to third party platforms. We have provided more information on this in our privacy policy.




    I can definitely hear in the first case that the amp sounds more dry and less compressed. In the other two, I hear more compression, which gives the tone a smoothness to the distortion. The tone is exactly the same, but the way that the tone reacts, i.e. more liquid at higher clean sense, more dry at lower settings, makes me think:


    1) This was always how it works


    2) During some firmware update, something has gotten screwy.

  • Where is the actual gain knob in either? I don't notice anything other than the expected result of what the manual and I've described. Any difference can be accounted for by the mix of "clean" signal in with the distorted signal if the gain isn't at 100%.

  • There's another test you can do to determine how much of a factor that is. Set Clean Sens to its minimum, then adjust the Direct Mix in the amp block to be slightly higher than zero, see if it gives you a similar effect on those gain laden profiles.

  • Awesome thread!


    have you guys seen my thread here? Kemper FX loop noise


    To add to the discussion: The clean sense on my Kemper certainly does affect the input LED. And it affects some kind of gain staging when I use the fx loop. With nothing connected to the Kemper - literally nothing!! - the noise floor changes drastically JUST by changing the clean sense.

  • I kay he wrong (it has happened before ?) but I just can’t reach the same conclusion based on your tedts or my test Per.


    The most difficult to grasp part is how Clean Sens is utilized in the amp block. If you take it as written, it really is operating in a similar fashion to the loop mix pot on most amps, it mixes between a parallel clean path that doesn't go through the pre-amp and the output of the amp. At max therefore it should make the Clean signal twice as loud.........


    ........I continue to believe that Clean Sens is just the volume/mix of a clean signal in the amp block output,


    if that were the case then Clean Sens would have no effect in the tests I did with the Amp Block bypassed.

    Quote

    The thing that makes this more complex is that this is also attenuated by the "Gain" control on the amp itself. So when Gain is at 10 that parallel clean path is completely muted, at 0 the parallel clean path has a value of 100%.

    i’m not 100% on this but I think that once gain goes above a certain number (I think its 4) the Kemper sees the rig as a distorted profile so there is no clean parallel portion. At 0 it is a clean profile and between 0 and 4 there is som sort of blending happening.

    Quote

    The reason I suggest using a distortion stomp to test with is that this is reactive to volume. Most other FX aren't. This makes it very easy to audibly tell whether the signal reaching that stomp is lower or higher in volume than before as you modify the Sens values.


    yes, but the reason this provides a poor test is because the KPA uses some intelligent processing when it detects a distorted sound.

  • The KPA isn’t that smart. What’s the reason you think there’s a complex intelligent processing with distortion rather than a much simpler approach? If it were that smart why would it only apply up to a certain pint in the signal chain?


    You can hear the crossover mix of clean if you make the sens parameters extreme enough, it’s possible that clean sens isn’t the full throw of the dial but you’d need to test that somehow t figure out how much, e.g. perhaps make a profile with the gain set to 100% (but with a very clean sound) and then just measure when the volume started to increase with a high clean sens setting.


    if you have the amp block bypassed I imagine it simply engages clean sens at 100% at the location of the amp/loop. Again just check by using a distortion stomp to see at what point in the signal chain it applies it.

  • Another thing that does suggest the Clean Sens is acting as an input trim is the Input Led. As nightlight already pointed out above,


    Quote

    However, there are some niggling doubts:

    1) The fact that the input LED should change at all when reducing clean sense. What is the explanation for this behaviour? (Indeed, what is the purpose of the LED at all)

    The LED is reflecting the input signal. This can be tested by rolling down the guitar volume and watching the LED.


    Start by setting Clean Sens to +12db and play hard. The LED goes Red. Now turn the guitar volume down and play again. The LED glows Green or only just lights up depending on how far you turned the guitar down. This is the expected behaviour if the LED is measuring input level.


    Now turn the Clean Sens to -12db but turn the guitar volume back up the LED reflects the lower input as would be expected from reducing the input level.


    It isn't conclusive that this is being reflected elsewhere in the signal path but it certainly shows that Clean Sens is affecting the measured input as shown on the LED. If it was only affecting the output of the Stack section and beyond the LED wouldn't respond when reducing the guitar volume.



    If it were that smart why would it only apply up to a certain pint in the signal chain?

    That could well be a conscious decision based on the fact that adding a boost in front of an amp is generally intended (by the player) to push the amp without necessarily increasing overall volume whereas when players put the boost/OD in the loop (in the physical world they can't really put it any later than that) they do it to achieve a volume boost. If I was designing the KPA to behave the way my customers want to use it, I wouldn't allow distortion to smooth volume post Stack.

    if you have the amp block bypassed I imagine it simply engages clean sens at 100% at the location of the amp/loop.

    I know what you are saying and it is definitely one possible explanation but it is a huge leap of faith to "imaging" it works this way when there is no other evidence to support that theory. That looks like confirmation bias to me.


    I don't know the real answer and am just surmising in the same way you are but based on the evidence so far I can't see anything that suggest the Clean Sens doesn't act as an input gain in some way in addition to it's main function of balancing clean and distorted levels.

  • Well obviously only the Kemper team really knows how it works but the evidence is there in the tests that I did. I think you’re over-complicating a simple thing.


    I just don’t see a reason to believe in a complex solution with special cases for every distortion, compression, ducking etc, delay dirt function, combinations of stomps etc, when there’s a very simple answer that matches the observed behavior.


    If a boost were expected by the user in front of the amp wouldn’t you expect clean sens to actually modify the input volume? But it doesn’t because if it did then the amp gain (distortion) would respond accordingly, same as if you adjusted the volume pot on your guitar. Clean Sens doesn’t act like that. Try it, roll back the volume on your guitar and see how the amp sounds, then bring it back up and lower the clean sens setting. They’re not doing the same thing.


    Instead the only location where stomps are affected by the clean sens setting is post amp/loop block.


    Then you can actually hear the mix itself if you use extreme settings on the Sens settings.


    Why would there be a complex reason for this?

    There’s a very good reason that I don’t think there’s a complex reason for this - It’s made by humans. We are lazy creatures, the only reason we invent or engineer stuff is to avoid having to do a much more laborious thing. So we apply smarts instead. The smartest most intelligent most elegant solution to a problem is the simplest one. I think Christoph and his team are very smart, and I just don’t think they’d expend that much effort to deliver the same behavior as a pretty simple solution.

  • I did a search on CK's forum name and the phrase "Clean Sense". He consistently calls it a volume control and in one post even considers taking the input LED out of action, as he feels it confuses people's understanding of the function. Mr Kemper also states that the input is soft clipped if the input signal is too high, though the KPA has masses of digital headroom at the input and the soft clipper only comes into effect when the input LED is consistently red. He also says that this has no effect whatsoever on the distorted sounds, as these are by nature hard clipped at a lower threshold. The only instances where clean sense could possibly be perceived to be affecting distorted sounds is with breakup/low gain tones, where the balance between the clean and distorted portions of the profile could possibly be audible.


    Make of that what you will.