I've been thinking about this alot. There are enough big differences between "sequencer looping" and "delay looping" to make them significantly different. Technologically, the former requires much less memory. Audially, the decay of past generations is somewhat more interesting with the later, although this is a purely subjective judgement. When you play a note into a MIDI sequencer loop, what is typically recorded is 1) what note you played and 2) how hard you played it ("velocity"). This takes somewhere around five bytes to represent in your sequencer (three for the actual MIDI data, a couple of bytes overhead for timing data) regardless of how long your loop is. Now say you want this to feedback at a nice robust 95%; enough to keep the beat around, but also to let old material leave as you evolve the loop. What this means in MIDI is that the note is played a second time without you doing anything at 95% of its original velocity, and then the third time at 95% of 95% of the original velocity, and so on. When you play a note into a digital delay loop, you're essentially recording digital audio. Just for the sake of comparison, let us say that's 44100 2-byte samples every second of loop (that's somewhere near CD-quality in mono). This means 88200 bytes per second of loop for our delay buffer! Now say you want this to feedback at a nice robust 95%; blah blah blah. What this means to a digital delay is that the relative amplitude of each sample is reduced 95% each time it goes through the delay buffer. That's the kicker. When you reduce the velocity of a MIDI note-on message, you are not necessarily reducing the amplitude of the sound. In fact, in most cases, you're changing the sound in very many more ways than amplitude. The effect is far different than if you had used a digital delay for the same purpose. The selection of what modulations occur in response to velocity is critical in this sort of technique. Also, most MIDI synthesizers can only play so many sustained notes at the same time; with a digital delay there is no such limitation. This is part of what makes MIDI loop sequencing attractive for percussion tracks; percussion sounds don't usually sustain for very long, so you don't run out of "voices" in your synth. Also, in usual styles of percussion track evolution (I'm thinking of the x0xes), the "delay feedback" is 100%; to evolve the track you build a completely new one and cut or xfade to it, or play with outboard effects. In sum; you can't easily sound like a Frippertroid with just MIDI sequencer looping. Finally, at the risk of being a horrid A/D troll, I believe that analog tape delays are likely to have a significantly different sound than digital delays, due to the audio characteristics of tape saturation. I wouldn't be sure though, I've never played through an analog tape delay before. (Y 'jfm3) P.S. Interesting... It was certainly once the case that good recording tape was less expensive than computer memory. Nowadays, I'm not so sure that's the case anymore.