There’s always a piece of gear that has guitarists talking at NAMM , and this year the Kemper Profiling amp has dominated that discussion. Back in the mid-1980’s when Christoph Kemper was writing code on a Sinclair ZX Spectrum (the British equivalent to the Commodore 64); the technological advances that would eventually allow him to develop the Access Virus series synthesizer must have seemed a long ways away.
Since the development of that iconic synth, Christoph has been looking further at the horizon and now with the forthcoming Kemper Profiling Amp tackling the ultimate challenge, truly digitizing the tube amplifier.
His profiling process, which he details below, is a fascinating departure from the modeling amps and pre-amps currently on the market. The initial first hand reviews all seem to point to this potentially being the next big thing for guitarists.
Christoph was kind enough to give Guitar-Muse readers an in-depth overview of the profiling process and the amp itself (but for those of you who wish to whet your appetite, you can see some initial videos on the Kemper Video Resources Page). With that, you might want to grab a cup of coffee (and your credit card) as here are:
10 Questions with Christoph Kemper of Kemper Amps.
1. Thank you for your time Christoph. How did this product come about?
My company has been making the Access Virus synthesizer for the last 15 years. Throughout this time I have been fascinated by distortion, which was one of the main sound aspects of the Virus synthesizer. As a producer, I was into guitar sounds as well, and a couple of years ago I started developing my own guitar amp.
When it came to the task of creating an accurate simulation of a distorting tube circuit I realized, after a while, that my previous approaches to distortion did not match the dynamic character of the real thing. The tube is not only special because of its soft distortion curve, but especially due the fact that it sucks current from the surrounding circuit, when it is driven into its saturation. This changes the behavior of the circuit dramatically, but only for the moment of distortion. This is the crucial point for a tube simulation, and some modeling amps obviously have great approaches for these dynamic sounds. I also found a great way to parameterize this circuit and behavior, but it took a very long time, and was very cumbersome, to match these parameters to those of a specific tube amp.For my circuit, there were simply too many interdependent parameters, and it would have taken ages to model just one or two dozen amps.
As a basically lazy person I spent my time trying to find an automated method, rather than modeling amps by hand. The problem is there are many equations, with even more variables, that need to be solved; I wanted these equations to be solved by the system listening to the original amp.
It took me almost two years to have the first promising profiler method up and running. The actual profiling algorithm is not based on known techniques, so I’m afraid I won’t be going any deeper and revealing the secrets behind it.
However, many people have asked about the purpose behind our weird-sounding test signals, so here’s a rough explanation:
The pulsing white noise modulates the saturation and thus the current of the distorting tube as explained above. By checking the residual distribution of the noise and the slight changes in the frequency response, a number of circuit parameters can be solved. The “UFO” sound does roughly the same, but seen – or heard – from a different angle: it is a fast sweeping group of sine waves creating interference signals in the distortion stage. The analysis of the interference results solves another handful of parameters.
2. What is the difference between modeling and profiling?
Personally I have two views on modeling and profiling: one is technical and the other is philosophical.
Modeling, in the first sense, is bringing the physics of the real world into a virtual world by defining formulas for the real world and letting them calculate on a real-time computer (such as a DSP or a plug-in environment). I don’t know how other companies model their stuff, but I tend to listen more to a circuit rather than studying its theoretical background. By treating models on a theoretical basis, one tends to oversee some very important side effects that can later be heard clearly, so you end up listening to it anyway.
I did our models of the distortion pedals just by listening to them, and matching my model to perfection just by ear.
Profiling is an automated approach for reaching a result that is probably too complex and multidimensional to achieve by ear, or by capturing the behavior of individual components in isolation. This is the case for a tube amp.
By philosophy, “modeling” was used as a marketing term by some companies. It says: “Here is a valid virtual copy of a valuable original”.
What I have rarely seen is an A/B comparison between the original and the virtual version. Why is that? Profiling, in our sense of the word, (yes, take this as a marketing term, too) is a promise to create a virtual version of your original, but with the ability to qualify the results by a fair A/B comparison. You get what you want, and you can check what you have just got.
3. If understand correctly – the profiling process allows for controls to extend beyond the range of the profiled amp (i.e. a Fender champ cranked to 20)? Does this extend to other settings as well?
This is true. The gain control allows for a gain range of 80 dB, which is from absolutely clean to insanely distorted. This gain control will be applied to every profile, so you can easily go beyond the analog boundaries.
The equalizer is another obvious tool to go beyond (the physical range of the profiled amplifier). We are building up a library for passive tone stacks so every amp can be equipped with the corresponding equalizer, even after the profile has been captured. The tone can be shaped then as on the original. But you can even choose another passive tone stack for your profile, or even a studio equalizer, that goes far beyond the boundaries again.
Our unique amp controls that change the basic sound and the distortion character of the amp are really exciting. Whenever we demonstrated these features to guitarists, they were so blown away from the impact of these controls, that nobody ever asked for a specific EQ or tone stack. Within two dozen demonstrations we touched the tone controls three or four times only.
4. With combinations of mikes and gain stages it sounds like multiple profiles could be made of the same amp. Is it possible to combine profiles into a single sound?
This is actually a plan for a future function that we are considering. It might be possible to stitch a number of profiles to one single profile in such a way that it would add a new parameter for morphing thru this new dimension. I could think of profiles that represent the amp with a different mic angle, mic distance, different temperatures, etc.
This is an issue for future updates; with our synthesizer line we have a 15 year tradition of free feature updates via the internet and we’re planning to create a similar legacy for our guitar amp. We definitely have plans for a (centralized) profile exchange platform, but these are not specific yet.
5. What profiles (amps, cabs, and effects) does the unit ship with and how many can it store?
It is actually too early to release a listing of amps now as we are just building our library. At the moment we’re collecting and selecting profiles from all kinds of directions. In addition to standard amps (lo gain and high gain), we will also profile rare and expensive boutique amps and continue to expand our library even after the amp has been released.
The Profiling Amp can store at least 1000 profiles, so we are going to chase every amp that people are looking for. The profiling process is limited to amplifiers. This includes tube amps, solid state amps, most modeling amps and most distortion pedals in combination with an amp. It is not possible to profile other kinds of effects, as this would require a wholly different technique.
Our amp is equipped with a large number of high quality vintage and studio effects, a collection that will also grow over time. From our synthesizer line, we have a well-established reputation for creating first class effects and this experience has already migrated to the Profiling Amp.
6. Can effects spill over between patches? (For example, if I have a rhythm sound with a delay can the delay spill over if I switch to a lead tone?
Yes, this is a major feature of our amp. The delay and reverb will spillover at any time without certain precaution. There is enough calculation power on board to have the old effects decay, while the effects of the new patch already sound.
7. I understand that the unit has very low latency which makes it a good candidate for live use. What are the recommended methods for using the unit live?
The audio latency is 2 ms and a bit. Due to the AD/DA conversion, audio latency cannot be avoided but our latency is equivalent to the sound travelling 3 feet thru the air. Our amp has a dedicated monitor output that will output an individually controllable on-stage signal. The main outputs will go to the front-of-house mixer and present a perfect studio sound to your audience. The close-mic studio profile works best in the front-of-house.
The monitor output will also feed the return input of your tube amp and cabinet if you still want to bring it on stage. (Of course, there is the problem that the guitar speaker wouldn’t deliver the pure studio sound, as it would introduce its own character, but this live characteristic might be what you’d want for your stage sound.) You would use this amp only for your stage monitoring, so no microphones would be required.
You could also use a linear active speaker or monitor box for monitoring that would recreate the profile even on stage.
8. The amp seems to have a very clean and compact design.It’s hard to determine what will happen in the future but will there be other hardware or software options for the unit available or possibly other additions like Kemper brand Midi control pedal?
Derivates of this kind (like a rackmount or a floorboard model) seem to be possible but decisions have not been made yet. We definitely have no plans for a VST/AU version.
Certainly, we have plans for a dedicated floorboard, but no specific release date as yet. Additionally, we support midi floorboards. In terms of midi patch latency, the patch switching time will be only a few milliseconds. It is so fast that we allow another couple milliseconds for a smooth fade out and in for a nice transition between the patches.
9. Can you talk about what’s under the hood (processors, speeds, sampling rate, A/D conversion etc)?
The main DSP is a Freescale DSP (formerly Motorola) running at an equivalent of 400 MHz speed. The code consists of tens of thousands of lines of pure assembler code. The global sampling rate is 44.1 kHz, while the internal sampling rate is partially much higher. The algorithm for the tube simulation runs on more than 700 kHz sampling rate (!).
10. When is the unit available in the states (and MSRP)?
The production has started already, but it’s hard to say when the first units will hit the USA. We will apply some extended test cycles to the first units. A conservative estimation is: Not before October.The retail price will be around $1800 to $1900. This is also not 100% fixed yet, because of the volatile Euro/Dollar exchange rate.
Thanks again for your time Christoph! Information about Kemper Amps (and Christoph Kemper) can be found on the Kemper amps website at http://kemper-amps.com.