TTS-1, as you say, is a Sound Canvas VSTi. And not even a last gen one. If you want your arranger to sound like a G600, maybe even a G1000, then OK... it's going to do the job. But who in their right mind is going to run a software arranger on a current gen laptop, and be happy with a soundset that is based on something 15 years old or more?!

I mean, what's the POINT?

You are back with the PA3X now. Honestly, are you going to be happy with an arranger that is FAR less impressive than that?

Look, the whole idea of the software arrangers is it allows you to play multiple style formats. So, you are going to want to import the latest styles. But, with the latest styles comes the task of rearranging EVERY SINGLE STYLE to work with a soundset that is completely different to the one the style was written for. Not only the basic Capitol sound, but the myriads of Variations that often are quite utterly different to the capitol sound. Drumkits with quite different note mappings (Standard MIDI, GS/GM and Yamaha mappings all seem a distant memory), utterly different velocity crossover points between samples, velocity response, etc..

Just this task alone is quite a task (look how long and how slow EMC are in responding to style conversion software for newer arrangers with newer soundsets). And now you want Don to consider the task of creating a Soundfont or VSTi based soundset that is not only going to cover ONE manufacturer, but ALL of them, and do it seamlessly?

I don't believe it can be done. Dom certainly couldn't. I have never heard a VSTi yet that, when sent a proper Roland GS SMF, or a Yamaha XG file, ends up sounding BETTER than its hardware equivalent. And Lord help you if it contains some of the more obscure Variation sounds or those lesser used drumkits..!

But the task of creating an look-up table to convert every single PC/00/32 code for a Roland sound, or Yamaha sound, or Korg sound, and pull up the closest equivalent in a Kronos or MoXF is FAR less difficult. The sounds have already been created, they've been balanced and effected, eq'd etc.. The hardware WS is utterly impervious to polyphony over-run ( a major issue in recreating dense styles in software alone) and if certain PC Variation sounds aren't close, it is a WS after all... not that hard to program something closer.

It's all well and good getting an SD2 or 4 to recreate Ketron styles identically to an SD1. But isn't it far more exciting to contemplate the vArranger running into the Kronos or MoXF, and finally have that mind-meld of the arranger and the WS/Arp/Loop player/Karma generator, and move beyond simple recreation of legacy arrangers and move towards creation of something utterly new and unheard as of yet..? But with all the reliability of TOTL hardware based synths?

Just as Kay's Karma software can push a Motif into utterly new capabilities, vArranger engineered to trigger existing TOTL WS's would add yet another degree of capability to a keyboard utterly incapable of doing this for itself. And would take FAR less effort on Dan's behalf than trying to put together some naff VSTi soundfont that would struggle to achieve better than 15 year old sound quality.

Just my opinion... but backed up by experience of watching Dom's project go down the toilet.