I wanted to build the music like anyone does it on a drum machine, but with real sound in real time, so I did not have to record a basic beat for 20 seconds in front of the public before starting to put harmonies on top of it. Tapping is not acurate enough for multiples, because it usualy leaves the bases irregular and creates clicks in the smooth carpets I often create for a base. I used to calculate a multiple of the delay time displayed on the 2290 while playing and then typed the value into the keypad - scary to remember! 

I went to Denmark and proposed to t.c. to add such functions for looping to the 2290. They thought there would be no market. I wrote to Roland, but they claimed to do this all and have patents - did not understand that! Then I was looking for Lexicon and was lucky to meet Gary Hall. He understood quickly what I wanted, but found that Lexicon would not invest into such a machine (about 4 years before the JamMan). But he sold me his (!) PCM 42 and advised me how to implement 30 sec and I squeezed a board with a bunch of TTL chips into it and achieved tap and multiply functions! 

Still, the functions were not perfect and I wanted digital feedback to keep the loops without deteriorating and still be able to fade them any time. Also I was sick of destroying nice loops with a wrong note or noise. It is necessary sometimes to jump back and undo what you do not like. 

So in '91 I decided to build the machine in my small laboratory at PARADIS in Switzerland, and immediately, fate sent me Eric, a genius programmer. He had come to live in our community when I realized he had just finished his studies. Since he was experienced in 68000 systems and DSPs were still expensive in Switzerland, we started our crazy project on a 68000 with a development system that was badly adapted to our mac IIcx. And we ended up creating 10,000 lines in assembler! I did the engineering of the hardware and matched the cheapest 350 different parts to a manufacturable unit. I learned to do PCB layout and mechanical drawings, front panel design, logo and manual - there was no money to give anything to a third party. A new foot controlled user interface had to be created to call all functions immediately and as intuitive as possible with only a few keys. Instead of modes or "shift" keys, we use long presses and cross functions. 

Well, its impossible to pass the story, it was too intense and incredibely guided.  For us its like a miracle, that the LOOP Delay, as we called it, started to work and keeps growing up to date. Even good we did not know how dificult it would be, otherwise we would never have started. We sometimes spent two weeks to find a bug because we did not have a logic analyzer and did not know whether the problem was in the hardware, in the software or in the development system. But it was an incredible experience to live how all we dream can become possible, and how we executed a plan which must have been there before we started to see it. We had really good times, too, looking at the scope's beautiful pictures and hearing the first crazy distorted sounds coming through. 

In August '92 I traveled to Brasil with the first prototype and gave a few concerts. One of the stars of Salvador wanted to buy this very prototype! I had to switch it on several times until it worked correctly, and when I found the problem later at home, I had no explanation how it could ever have been working! 

In the mean time I built piezo pickups for the AVALON and the IBANEZ AFR bass and other projects, also playing. In March 93, in Frankfurt, the PARADIS booth was crowded to see Ljubo doing his impressive demonstration and competent explanations in 5 languages, and we sold 100 LOOP delays that year, which was not bad, considering the direct distribution and the price of initially US$ 1500! 
 

The ECHOPLEX Digital Pro, as it was called later, had to have more features than the original LOOP delay and to be cheaper. Keith McMillen brought good ideas and Kim did his great job in making it all work. Kim found over 300 software errors along the development, and I fixed them one by one. Eric came down here twice to implement the structural software problems I was not able to solve. He used to complain about the heat and program a powerbook in a hammock.

I am sure, many users are aware only of a small part of the capacity of the machine. There is nothing wrong with that. It takes a time to learn to use it, and each user finds his own way to the loop. It's not like an effect machine you just switch on and it sounds better and you step through all presets to know it. It takes practice like an instrument, study the compositions and experience the kind of melody lines and sounds that dig deeper when repeated many times versus the kind that gets boring.