Meet the creator of AmpLion
Hello Audified friends,
Today, we will ask a few questions to our senior product designer Jaromír. The reason being that soon one of his most personal Audified products AmpLion will get its next incarnation. We bet he almost can't wait for it to be out! :)
Hi Jarda, how are you today? Still thorough-going yet with a big smile on your face at all times, as when we spoke last time? :)
Hi, thanks for asking. I feel great today. It’s always fun to talk about guitars and amps.
Tell us, when you first started programming and what things drew you to it the most? Were they music-related?
I started with programming at university, but I took just the basic courses. At that time, I was more an analog guy. The first real project I worked on and had to use programming was a control unit for an analog mix console controlled via MIDI. Shortly after, I found a job advertisement for an internship at Dsound/Audiffex. They showed me their overdrive and distortion plugins, and I was then wondering whether I could improve the sound to be more realistic. And if it is even possible to meet the sound quality of analog effects. I think it was in 2006, and that year I started to take programming seriously.
You are a passionate guitar player. What were your first piece of guitar software and hardware? Do you have fond memories of it, or do you despise it nowadays? :)
I built a treble booster pedal and played into a hi-fi amp. Not a good start indeed, and very quickly, I bought my first combo, 30-watt transistor Marshall. But it was not loud enough when you played in a band, so I bought a low-quality copy of 50W Marshall Plexi and replaced its preamp with a preamp I built according to the Dual Rectifier schematic from the internet. It was really good sounding but not reliable enough. Then I found an Engl Thunder combo, and I fell in love with the sound. Since then, I have always played via Engl amps when playing live. The software came much later, I mentioned the Dsound effects, and I tried the AmpliTube, but frankly, I was not happy with the sound. That’s why I decided to start working on the first ampLion software.
It's getting close to the release of your latest software project - AmpLion 2 Rock Essentials. Could you share with us what the most demanding moments that you experienced during the development phase were?
I had a perfect model that sounded exactly like the hardware amp, but I was not able to run it in real-time. It took me a long time to tame it and run it on a standard computer. I am very happy I managed to preserve the original quality of the model.
What are the main challenges of developing, let's say, a guitar-specific software product in comparison to software for non-musicians?
Standard software programmers interleave their job with drinking coffee, but guitar-specific software programmer interleaves with playing guitar. I am just joking, but there is something about that. The main challenge is that you often don’t have a specification or requirements. Either it sounds good or bad. You have to pay attention to all details in the sound, and it can be pretty exhausting. During the last project phase, I sit in front of my computer for hours with a guitar in my hands, wearing headphones. It sounds like a dream, but in fact, it is hard work.
What do you think is the key to the long-lasting success of your ampLion Pro guitar software?
The key is the technology used there. ampLion Pro was built as a real amp circuit schematic solver, fully capturing the amp dynamics. In contrast, other guitar processors of the time used just a series of distortion blocks and filters between them to match the sound of a real device. The sound quality was there when playing high gain riffs, but when it came to dynamics, there was a huge gap between the guitar processors and real amps. But not with ampLion Pro :)
How long does it usually take for you to fine-tune a model of a real amp? And is it even possible to make it 100% accurate? (I know, what a sacrilege question that is!)
It is a huge difference if you try to capture one sound, I mean one amp setting, for instance, all knobs at noon or, if you are trying to match all knob positions. To explain the fine-tuning process: I don’t change the schematic of the model. I only change the component values within their tolerances (typically 1% for resistors and up to 5% for capacitors, but much more for vintage amps), measuring potentiometer taps, and most of the time, I’m playing with parasitic properties of the components. I usually spend up to one week just doing AB comparisons and playing the model and real amp.
Do you believe in the magic of much-coveted old valve amplifiers? And do you then think guitar software can be relatable to its wide fan-base?
The old valve amps gained a lot of reputation. Musicians have their idols in the previous era, and those idols used to play these amps. And sometimes, you want to have exactly the same sound as your idol. It is a significant part of that magic, but at the same time, it is a fact that the old valve amps sound different from new ones. There are many reasons for that - components, materials, and technologies changed over the last 50 years. There is a lot of value engineering on modern amps to deliver them at a lower cost on the market, too. All these small things affect the final result. But whether old amps sound better or not depends mostly on the taste of each guitar player.
For the second question, if you use in-ear monitoring or record the guitar in DAW, I believe the guitar software can be as good as the real amp. But it would be extremely hard to convince someone playing the full-stack at full power that these two things sound the same. When you play loud, the acoustic pressure adds the extra dimension to the sound, you can feel it in your stomach, and it’s the reason why even I take my heavy rig onto the stage instead of a small pedalboard with a guitar processor. In other words, you cannot always compare those two things. It depends on both the concept and the context.
What is your favorite old-timey guitar amp? Have you been able to model it yet?
Personally, I prefer modern amps where the distortion comes from the preamp but when it comes to old-timey amps, I would choose a treble boost channel from AC30, and yes I have had the chance to make the model of this great amp recently. And talking about effects, I got the original vintage Deluxe Mistress flanger from my father, it's a more than 40 years old unit with a lot of corrosion on the chassis but it sounds incredible, and I made a model of this guy as well :)
When modeling classic guitar gear, what is your motivation behind the process? To make it 100% accurate sounding or to make it feel great for the player? Or do these even contradict?
I believe that if you make it 100% like the amp, then it must sound great. But at the same time, there are no two same-sounding amps out there. The components and tube tolerances make each amp unique. I prefer the raw sound and a good amp or a good amp model, which will reveal your mistakes when you play. The flat-sounding cheap amps with infinite distortion and no dynamics hide your mistakes, and I am not fond of that. And in case you do not particularly like the color of a raw amp, you can always do post-processing in DAW or a mixing console. That is why my models don’t contain any additional pre or post-EQ sections.
As I had the opportunity to come to know you, I have realized how meticulous you are - in the best possible way, of course! :) In my book, that is an exquisite quality for someone who models gear. Do you think this tendency of making things perfect helps you in the process?
You really have to pay attention to every detail. It happened to me many times that I decided to skip some details, hoping they are not so important and you can guess what happened. They were, and I had to rework the model. So in a way, you really have to consider everything.
Do you believe there is some uniqueness to analog gear that just can't be touched by digital tech? Or do you think that it is possible to recreate anything in the digital realm with sufficient capacity and the fitting process applied?
As long as the gear obeys physical laws, you can model it, and it only depends on how complex the model is. There was huge progress in analog modeling in the past ten years. I am happy we are not scratching the surface of possibilities anymore. But there is still a way to go. We need to deep dive and repeat all the work we did on a schematic level inside each component.
The easy question at the end, what is your most desired piece of gear nowadays? To make it a bit more specific, what guitar do you want next?
Vintage Fender Telecaster, Mark V combo, and implement my algorithms into a hardware unit :)
Thank you very much, Jarda, for giving us such an insideful overview of the thinking behind your projects. And all the best to the new and soon released AmpLion 2 Rock Essentials!
Interview led by: