Liquid Profiles

  • I think I understand how this will work. So were you to know the original tone settings of a profile, you could opt to switch to one of the modeled tone stacks and input what the original tone settings were. But what is you didn't know the original settings?

    Hypothetically, you have a fender twin, profiled like this: B2, m5, T8, P5.

    In this case, the treble was bumped up and bass was cut some, presence and mids left alone. But you forgot these were the settings, so you just tell the KPA that the profile'd settings were all flat at 5.

    So when you bump up the treble on the kpa, what happens? If I understand correctly, the bump on the treble would sound like bumping the original amp from say 8 to 9 or 10, even though the knob on the kemper shows a bump from 5 to 6 or 7 or so. (assuming 5 is the 12:00 position) On the flip side, turning the kemper's treble to zero would only get the original amp to sound like the treble is set to 3 or so.

    If that's confusing, what I'm asking is, if you don't know the original settings of the amp and just tell the KPA they are all at 5, the changes on the KPA tone stacks will of course will not match the original values on the amp when it was profiled. However will an increase or decease on the KPA tone stack sound similar to increasing or decreasing that control on the original amp? Of course, is the numbers don't match, you wouldn't have the full range of the original knob in one direction and could go beyond the range on the other direction. But otherwise you could do modest increase or decreases and have it respond a little closer to how the original amp would have responded to modest bumps and cuts from wherever the original tone was set at?

  • Okay….I don’t know why I was under the impression, that some of it already was included in a possible current beta version. I haven’t clicked the beta-option. Can someone confirm or deny that?

  • Okay….I don’t know why I was under the impression, that some of it already was included in a possible current beta version. I haven’t clicked the beta-option. Can someone confirm or deny that?

    Nothing in beta as of yet.


    I marked the Announcements from Kemper board so I get a notification of updates. Burkhard is really good about updating that board when new stuff drops.

    “Without music, life would be a mistake.” - Friedrich Nietzsche

  • Yeah not quite correct. You can't capture the tonestack, these will be presets available to apply to your profile. So in theroy you could profile a Marshall JCM800 and apply a Vox AC30 tonestack...not sure why you would want to however.


    So there will be the limitation of the preloaded tonestacks - bit like imprints for the Kones I guess...

    Yes - hence why the subject was changed from "Liquid Profiling" to "Liquid Profiles" I suspect :)

  • Machine learning has advanced capturing/profiling to a new level. What's to say AI can't improve upon modeling as well. The ultimate resource will/may be an integration of both technologies. Beware Skynet;)

    Has it?


    Can someone explain machine learning because so far all I've seen is more complex algorithms rather than what I understand as machine learning.


    Machine learning to me is taking multiple complex inputs and then apply this to future calculations. With a capture, you capture the input as accurately as possible, so where is the machine learning?


    Someone please show me I'm wrong as I'm sure I am....but feels like marketing buzz words...

  • I think I understand how this will work. So were you to know the original tone settings of a profile, you could opt to switch to one of the modeled tone stacks and input what the original tone settings were. But what is you didn't know the original settings?

    Hypothetically, you have a fender twin, profiled like this: B2, m5, T8, P5.

    In this case, the treble was bumped up and bass was cut some, presence and mids left alone. But you forgot these were the settings, so you just tell the KPA that the profile'd settings were all flat at 5.

    So when you bump up the treble on the kpa, what happens? If I understand correctly, the bump on the treble would sound like bumping the original amp from say 8 to 9 or 10, even though the knob on the kemper shows a bump from 5 to 6 or 7 or so. (assuming 5 is the 12:00 position) On the flip side, turning the kemper's treble to zero would only get the original amp to sound like the treble is set to 3 or so.

    If that's confusing, what I'm asking is, if you don't know the original settings of the amp and just tell the KPA they are all at 5, the changes on the KPA tone stacks will of course will not match the original values on the amp when it was profiled. However will an increase or decease on the KPA tone stack sound similar to increasing or decreasing that control on the original amp? Of course, is the numbers don't match, you wouldn't have the full range of the original knob in one direction and could go beyond the range on the other direction. But otherwise you could do modest increase or decreases and have it respond a little closer to how the original amp would have responded to modest bumps and cuts from wherever the original tone was set at?

    I think you are right, the settings it was profiled at will have an impact. If you don't know the settings, the tone stack will be "off".


    This is if you apply them retrospectively, the suggestion being is it should change the way you profile. In other words, people focused on the sweet spot as it was a "snapshot". This will not be as necessary I guess ?


    I think the next challenge will be " oh my JCM800 doesn't behave like this", in the same way that was mentioned about the Kemper drive presets - doesn;t sound like my tube screemer...because its based on the "design" and won't account for worn components etc. Not sure how much that actually matters but...

  • The devil is in the details. The pots typically used on most guitar amps have +/-20% tolerance. Some pots are linear while others are audio(log) taper. There is quite a bit of tolerance and variation just in the pots themselves. The tonestack is an interactive system with the pots interacting with the fixed value resistors/capacitors that also have +/-20% tolerance typically. The tolerances can be tighter, but this would not be typical for production amps such as Fender, Marshall, Vox, etc... Also, component values drift over time and temperature. So, no two amps are exactly the same. This is part of the reason, among many, why some amps sound 'special' compared to others. I guess that the Kemper tonestacks will model a generic version. But the claim that knowing the knob positions when the profile was made will lead to the Kemper tonestack to providing the exact behavior of the real amp is not believable or possible. By definition there is a lot of variation. Will it be good enough? Will anyone be able to notice? I believe that will be up to the user.

  • I think I understand how this will work.

    :)


    ...Can't wait to find that out. For me a supposedly new (=authentic?) behaviour of the gain knob is much more exciting than that of the eq knobs (which if course is also very exciting).

    Better have it and not need it, than need it and not have it! - Michael Angelo Batio

  • ......But the claim that knowing the knob positions when the profile was made will lead to the Kemper tonestack to providing the exact behavior .....

    I'm asking because I don't know. Not to be difficult (<- give it time. 8o)


    Has anyone at Kemper said these modeled tone stacks will have the "exact behavior"?

    I would also think that when modeling specific stacks, they'd select prime examples. The 'special' ones, so to speak. No sense using an off-the-shelf AC30 at Guitar Center when you can choose the best sounding one and base things off that, or even multiple good ones to develop an 'ideal'.

    “Without music, life would be a mistake.” - Friedrich Nietzsche

  • I'm asking because I don't know. Not to be difficult (<- give it time. 8o)


    Has anyone at Kemper said these modeled tone stacks will have the "exact behavior"?

    I would also think that when modeling specific stacks, they'd select prime examples. The 'special' ones, so to speak. No sense using an off-the-shelf AC30 at Guitar Center when you can choose the best sounding one and base things off that, or even multiple good ones to develop an 'ideal'.

    No, no one has claimed that. I probably should have written something like 'But the notion that knowing...' instead.


    I have no idea how they did the modelling. My main point is there is variation inherent in the thing that is being modelled.

  • :)


    ...Can't wait to find that out. For me a supposedly new (=authentic?) behaviour of the gain knob is much more exciting than that of the eq knobs (which if course is also very exciting).

    interesting thing HW from Tone Junkie said. That he often doesn’t like how original amp sounds with the gain on max, often too flubby or whatever. So after profiling on the max gain he likes the sound of, say at 8, he then increases the gain on the KPA to get more gain, so that it sounds like a more gained up version of the amp set to 8.

  • I would also think that when modeling specific stacks, they'd select prime examples. The 'special' ones, so to speak. No sense using an off-the-shelf AC30 at Guitar Center when you can choose the best sounding one and base things off that, or even multiple good ones to develop an 'ideal'.

    I'd guess that they use the nominal values for everything rather than "prime examples". That's the only thing that would make sense to me. I also imagine that one man's special sauce magic amp might be another man's crappy paper weight :D


    Also keep in mind that although the values of the components may vary X% from nominal, it will be a normal distribution; so most of the components will fall within a much narrower tolerance band.

  • Do you guys think profile vendors will update existing packs with liquid profile settings? Assuming of course, they have written them down at the time of profiling.

    One vendor has already said as much.

    I think many vendors will follow suit if the have the ability, whereas many others will not.

    Some might also not have exactly written down where the controls were but will offer liquid profiles anyway to the best of their memory/guess/whatever.

  • ...Also keep in mind that although the values of the components may vary X% from nominal, it will be a normal distribution; so most of the components will fall within a much narrower tolerance band.

    This is what you might expect, but higher tolerance parts sell for more. Many manufacturers pull the parts that are closest to the nominal value and sell them at a higher price. The end result is that there is a hole in the values close to nominal.

  • I'm asking because I don't know. Not to be difficult (<- give it time. 8o)


    Has anyone at Kemper said these modeled tone stacks will have the "exact behavior"?

    I would also think that when modeling specific stacks, they'd select prime examples. The 'special' ones, so to speak. No sense using an off-the-shelf AC30 at Guitar Center when you can choose the best sounding one and base things off that, or even multiple good ones to develop an 'ideal'.

    In the video I saw and if I remember correctly, CK said they used the amp schematics to model the tone stack. I took that to mean that no tone stacks were actually "profiled".

    "Faith don't need no second opinion"

  • In the video I saw and if I remember correctly, CK said they used the amp schematics to model the tone stack. I took that to mean that no tone stacks were actually "profiled".

    Yeah - he was pretty clear that these tone stacks are models applied to a profile. I was curious if he'd said "sound exactly like...." or similar.

    I don't think he did.

    “Without music, life would be a mistake.” - Friedrich Nietzsche

  • This is what you might expect, but higher tolerance parts sell for more. Many manufacturers pull the parts that are closest to the nominal value and sell them at a higher price. The end result is that there is a hole in the values close to nominal.

    That is interesting, and a very good point!

    However, I wonder if this was also the case "back in the day" (whatever golden age of guitar amps is often held in high regard, vintage stuff etc) - or if it a later development? I suspect the latter, but I have no idea.

  • Just on tonestacks: it is a mistake to think that all the pots at 12 o'clock or "5" on a standard BMTP set of tone controls corresponds to a "flat" response like a graphic Eq with all the sliders midway up. There is a lot of variation in circuit values - and the result is that some are naturally scooped and need the M control at 10 to approximate a flat response and some are quite treble tilted and will never get the bass level anywhere near the same level as the mids and treble. The design of the bass and treble filters is highly variable between different variations on the basic designs and there are often multiple toneshaping filters in the signal path - examining the various Marshall circuits and you will see why the whole response typically tilts to the treble end.


    I recently read a discussion article on tonestacks that was linked to either here or at TGP which pointed out that the Fender Shimmer relates to the phase relationships induced by the filters in the tone stack and that these do not show up in the typical tonestack analysis software.


    I would expect that CK and his fellow developers are aware of the intricacies of tone stacks and will have applied them as part of emulating amps in software.

  • NAM is free so there is no marketing agenda, it's a free science project with some great results based on machine learning according to it's MIT educated designer. He's not going around telling everybody it replacing anything other than try this cool thing I came up with. JM2C