You're right I don't think AWM2 based mono legato has gotten any update. The mono legato behavior can work as most expect when using the right kind of waveforms designed for it (no attack recorded in the sample). So waveforms provided could be one way of improving mono legato for AWM2 without stirring the feature pot. I haven't looked that closely to see if that happened or not.

Certainly allowing for more elements per Part gives one a lot of flexibility in creating sounds that utilize waveforms with no attack recorded.

The other saving grace here is that mono legato is often used for synth analogish sounds that are also now covered by the new AN-X engine. That engine has been "built" with higher resolution for lots of it and stands out as leap frogging the other two engines in many respects. This takes the heat off of making any changes to AWM2 mono legato.

I was wishing for AWM2 mono legato updates as well but am pretty happy with what's there to take the place of those improvements.

Even if the preset waveforms haven't been significantly "saturated" with new attack-less waveforms, the new AWM element options offer 3rd party content providers LOTS of options to provide these in Libraries (promotional or, more often, at a cost to us).

I was hoping that they will add that mono legato like Genos and some PSR series have and also most of other brands on old montage as update, and later maybe on new model. Its deal-breaker for me. I have Montage and tried in many ways to create that smooth legato transitions with samples, with different custom samples for offset (lot of peoples also tried), but it's impossible, there is always that reclick more or less and its sounding pretty bad for mono sax sounds, clarinets, woodwinds ...
If this is still the same I will probably sell current montage and buy used Kronos or Fantom. I don't think that they planning to do something about that ever with OS update. 🙁

[quotePost id=123622]
So it's Virtual Analogue circa 90's.
Odd that it's limited to 16 voices. [/quotePost]
I thought so at first too, but didn't realize that 16 voices is supposedly quite a good number for VA synthesis, as others have pointed out. There must be a big difference in processing from physical modeling engines..!? Also, Paul from sand, software and sound has talked about VA synthesis and AN-X in particular, and he appears to say it's quite processing intensive.

Yamaha might want to point this out in the various demos as they discuss the number of voices, because many people may not realize the difference, and could view 16 vs 128 voices as being quite low when it's actually not...