MK 2 - Profiling 2.0

    Quote from sailor

    That's interesting. Do you have details about the design of the new profiling that you can share? What profile processing at playback time would not be running on the DSP chip?

    I have zero insider information. All I know is that DSP has never been cited as the limiting factor related to playing Profiles.


    There is SO much else that can change besides one DSP chip to improve performance.

    Quote from b_ryan

    I’m certain that they will be. I’m pretty sure, that they have have said so already. Albeit in a slightly “downscaled” (in lack of a better word) version. So even if it’s not the actual power - or power limit - it’s quite possible something technical, most of us don’t know about.

    This is exactly what I mean, you will be able to run 2.0 profiles but you won't get the quality you will have on MK2. We are speaking about a digital system where after the first AD conversion nearly all is done by a SW running on a DSP (DSP because they are specialised in functions that are massively used for signal handling - audio here ) and then you have at the end a DA conversion to feed the outputs. 2 things have major impact on the performances and capabilities: the processing power (DSP here) and the memory (here RAM and Flash but if we look for performances more the RAM). Telling that we do not care about the power of the DSP is just for me a nonsense. The booting time that is more than 2 times faster is a good sign of faster DSP (and probably memory speed... here the flash is very important)

    Just an extract of the presentation of the MK2 you can easilly read on the site itselt of first page :

    " ... At the heart of the PROFILER MK 2 Series works an upgraded processing engine, unlocking faster performance, with boot times clocking in around 20 seconds, and a host of new features that expand the boundaries of what a modern guitar or bass rig can do."

    Quote from Ruefus

    I have zero insider information. All I know is that DSP has never been cited as the limiting factor related to playing Profiles.


    There is SO much else that can change besides one DSP chip to improve performance.


    Sorry, I thought you were saying you're a Kemper engineer.

    For realtime signal processing, the constraining resource is almost always the power of the DSP chip (which is unchanged in the Mk2). Other factors like I/O bandwidth or memory or the MCU (the key component that was upgraded in the Mk2) are important for some things, but relatively unimportant for the realtime signal processing (for example processing the guitar signal with the profile).

    I suspect the "limiting factor" comment was to say that the Kemper has plenty of DSP horsepower to spare, and that's probably true. That's why Kemper was able to add fixed effects. Presumably the new profiles will be more demanding on the DSP processor, so it's good there is processing headroom available. In other words, it's great news for us that there is spare DSP power, but it is the DSP power that is usually the critical resource for realtime signal processing.

    DSP power is a finite resource though, which probably explains what Damian mentioned, or why some effects were excluded from the fixed effects.

    After a certain amount of time keeping an old DSP is a bad bad idea because this old DSP becomes too much expensive. This is the story of all siliciums. You become the only user and then the silicon vendors are just asking you much more to keep it in production. There is no good reason to stay on an old DSP. On top for a lower price you will get much more power and this is valid also for RAM and Flash.

    Quote from StratUS90

    After a certain amount of time keeping an old DSP is a bad bad idea because this old DSP becomes too much expensive. This is the story of all siliciums. You become the only user and then the silicon vendors are just asking you much more to keep it in production. There is no good reason to stay on an old DSP. On top for a lower price you will get much more power and this is valid also for RAM and Flash.

    I'm surprised the expert, world leading engineers at Kemper haven't thought of this. Do you build musical instrument processors for a living? You should give them a call.

    I’ve been on “stage” weekly on a Stage since the Stage was introduced and due to all of the benefits, I wouldn’t go back to amps live BUT I am hopeful that 2.0 introduces the missing detail I experience when I do play my amps at home. When I play my Tele through my Princeton Reverb and snap the strings with the meat of my fingertips, that tube responding just hasn’t been replicated by any digital unit that I’ve tried. That being said, I’ll still be using my Kemper live every week. (Recently added a Player too for rehearsal night).

    Quote from StratUS90

    This is exactly what I mean, you will be able to run 2.0 profiles but you won't get the quality you will have on MK2.

    Yes. So whether you get 2.0 on the MKi is a matter of how you define it. It’s a yes and a no.

    Quote from StratUS90

    Just an extract of the presentation of the MK2 you can easilly read on the site itselt of first page :

    " ... At the heart of the PROFILER MK 2 Series works an upgraded processing engine, unlocking faster performance, with boot times clocking in around 20 seconds, and a host of new features that expand the boundaries of what a modern guitar or bass rig can do."

    Very well noticed. I know nothing about DSP and computer stuff. But could it be, that - as others here have said - it’s not all up to the DSP alone; and regarding power for both the 2.0 profiling and the running of the 2.0 profiles that it’s affiliated with something else?

    So when you quote “[…] upgraded processing engine, unlocking faster performance […]”, that could mean something entirely different than just the chip alone?

    As I said I’m not a computer geek, so I’m just wondering if 2.0 is related to the chip alone; and I mean both the 2.0 profiling and the running of 2.0 profiles.

    As mentioned here by myself - and others - the 2.0 profiles will be running in a downscaled version on the MKi. So that could mean, that it’s stuff connected to the chip, which is key to 2.0.

    Quote from Finally

    I'm surprised the expert, world leading engineers at Kemper haven't thought of this. Do you build musical instrument processors for a living? You should give them a call.

    I was still very recently a senior expert for embedded sw for a big car OEM. The DSP used by Kemper was by the way designed initially for Automotive. I know what is a DSP and perfectly know the reasons why we change HW from one generation to another. In automotive an OEM of this size is selling 6 to 7 million cars a year. The HW is in each car and has to be paid by each customer (impacts direcly the car price). This sometimes generates some SW porting activity but it is not a question because the R&D effort is amortized on millions of cars and then the equation is simple: 1 euro saved on HW is 1x6millionx4 years is 24 million euros. You can easilly imagine that it is more efficient to put the pressure and effort on sw side until this effort stays below 24 millions. Beleive me this makes the life of automotive embedded sw guys hard... really hard.

    Here the DSP was designed for audio processing and brings hw support for audio features simplifying the work of sw guys.

    To answer another question: in many of those systems the DSP is not alone and there is either a micro controller or a micro processor. In the last car generations we have very powerfull processors (nvidia, quacomm, ...) in the multimedia system that covers the audio with many cores and you can delegate to that processor some audio handling or other tasks that the DSP was executing to generate again some room on it and sometimes to remove some (we were using sometimes several DSPs) Maybe it is the case here also. This is easy until the functions are not linked between processor and DSP. If they are real time (or lers say very low latency expected) and in a row the split becomes more difficult to design.

    The split could explain the lack of flexibility of the new "fixed effects" in terms of positionning.

    Edited once, last by StratUS90 (November 25, 2025 at 9:36 AM).

    The discussion of DSP for effects/location/features is outside the discussion of Profiling 2.0.

    The Mk1 Stage has the same Profiling abilities as the rack/head. Identical.

    It’s way faster to boot, react….etc. That extra performance has nothing to do with Profiling.

    Quote from StratUS90

    The split could explain the lack of flexibility of the new "fixed effects" in terms of positionning.

    The MCU in the Kemper is nowhere near powerful enough for realtime signal processing.

    The reason for the DSP chip remaining the same in the Mk2 is what Ruefus said above. Their code is inextricably tied to the Freescale architecture.

    Quote from JC MUSCLE SHOALS

    I’ve been on “stage” weekly on a Stage since the Stage was introduced and due to all of the benefits, I wouldn’t go back to amps live BUT I am hopeful that 2.0 introduces the missing detail I experience when I do play my amps at home. When I play my Tele through my Princeton Reverb and snap the strings with the meat of my fingertips, that tube responding just hasn’t been replicated by any digital unit that I’ve tried. That being said, I’ll still be using my Kemper live every week. (Recently added a Player too for rehearsal night).

    Each year we get improvements in digital simulation of guitar ‘amp-in-the-room’ by all the major companies in this field.

    However I believe it will never get all the way there because a ‘live amp’ is able to deliver sound (and what we feel in the way of response to starting the the string vibrating and sustaining) without anything altering or filtering what a real guitar speaker cabinet delivers in that scenario.

    Simple laws of physics dictate those limitations for digital simulation. The real (or simulated) mic and sometimes mixer preamp as well as all that goes into making up the monitor playback solution are not in the chain when it comes to a real amp sound.

    The best you can hope for is a perfect simulation of a recorded ‘real amp sound and feel’.

    Compared to us civilians who are accustomed to the real speaker cab in the room, players who are accustomed to standing at the studio desk playing a real amp, isolated in another room, mic’d and sent through the recording studio playback are easily satisfied by digital. Because at that point there is little to no physical difference in the signal chain…

    Kemper does a really good job of getting close. The Kones are perhaps the best yet.
    However, nobody has found a way to completely break the laws of physics.

    Edited 6 times, last by Enchilada_Jones (November 25, 2025 at 4:25 PM).

    I’m sure you’re right but I can always hope. 🙂