I am using a Pentium 200mmx with a Midiman 2in/4out interface. I have, in my opinion extensively tested the 2080 in every conceiveable scenario available to me. By that I mean my soundcard midi interface, a mate's mac, my mac and also directly from a controller keyboard. I believe that although software sequencers aren't the tightest they could be (I use cubase VST 3.7 on the PC), that it has nothing to do with this.

An interesting point mentioned by another person here was that the JV1080 is not as bad. I haven't had the opportunity to try this but I have thought of reasons for this in relation to the differences.

1) Having 8 Expansion slots rather than four (I can't see this having an effect-especially as I have one expansion board).

2) Effects: 3 EFX may have some bearing in Performance mode (where timing is worse). This may be because the effects routing is more complex.

3) Effects ii : When I did speak to one half decent guy at Roland- (I did say half!) he was saying that the 32bit Risc cpu handles EVERYTHING. Now if "everything" involves the effects then that is two more chunks of effects processing to handle. I personally can't see one processor doing that amount more without completely screwing the whole machine up - but it is still a possibility.

4) The screen: This may seem trivial but surely the 2080 has more screen processing to do (Full graphical display, lots of info at once etc. This could have an impact on available audio processing power.

Whilst I appreciate that no one really cares what the technical breakdown of why this problem occurs (after all isn't this supposed to be about music!!!!) However I for one would like to see Roland admit to their mistakes and explain to it's customers why they screwed up. Treating people like me with contempt when I phone up just makes the situation worse.

I just want as many owners as possible to read this and take it for what it is. That is the only way anything will happen.....