SYNTH ZONE
Visit The Bar For Casual Discussion
Page 2 of 2 < 1 2
Topic Options
#254138 - 01/23/09 12:33 AM Re: Roland and Korg Chord recognition thread
Diki Offline


Registered: 04/25/05
Posts: 14266
Loc: NW Florida
I completely agree, however, an awful lot of Korg's success with the style doesn't come from unique OS features... Most arrangers are capable of playing longer styles. Most arrangers are capable of at least three different patterns (subtle or gross, it makes no difference) per style. It's just that the manufacturers rarely make as inventive use of these features as Korg does...

I give a lot of kudos to the Korg style team. They certainly make better use of what their arranger really is capable of than some others...
_________________________
An arranger is just a tool. What matters is what you build with it..!

Top
#254139 - 01/23/09 12:50 AM Re: Roland and Korg Chord recognition thread
TedS Offline
Member

Registered: 04/28/06
Posts: 834
Loc: North Texas, USA
Diki, I like the idea of a user-adjustable "window" for intelligent chord recognition. When a style is running, Yamaha does a great job of using portamento to sneak the correct notes in if you play them late. But I often play lower backing with the style stopped. As a single-finger chord player, I can tell you that my midis are loaded with false notes (and even false chords) because the recognition is *very* quick. By analysing the tempo and "ticks", I determined that if you press the 2nd note even 0.008 seconds later in some circumstances, you may get lingering overtones from a premature incorrect chord. This is especially apparent with left-hand voices rich in harmonics when the style is stopped.

Ironically my old PA-50 was very forgiving in this regard, apparently using a "batch" recognition strategy. If I had to guess, I would say the Korg waited about 25ms (0.025 sec) or more for you to press a 2nd or 3rd key before sounding the chord. So unless I was very clumsy and indecisive it rarely played a premature chord. Just to make sure there was no internal quantization going on, I midi'd it into the Yamaha for a quick record. I never found any false notes in the score!!

Most of you play in the mode where 3 notes are required to change the chord, so I can't imagine there would be a problem with premature recognition. The chief problem with "easy" or "intelligent" chords is that major chords require only one note, but other types require more. So the programmers have to decide how long the machine should wait before sounding a one-note major chord. I am aware of the problems with the Technics KN-6000 at its release, so making this window user-adjustable is the best approach.

While they are redoing this part of the OS, they should also allow the user to customize the "intelligent" recognition. For example: minor 3rd below the root could be 6th (Korg) or minor (Roland). C+F should be Csus4. (Korg recognizes this, Roland doesn't.) C+F# is Dim on Roland but not on Korg. On the better Yammies, CMaj7 must be played in root position with a low C. But on cheaper boards like PSR-E413 you can play B2+C3+E3 (which is easier fingering for me!).

Chord recognition (and styles) are the defining characteristic of an arranger. Yamaha's weird "AI Fingered" mode was introduced about nine years ago, and not much has happened since. More customizability in this area, and a degree of Karma-like randomness in the styles could herald something truly new and different (as opposed to the mild upgrades we have been seeing.)
My $.02,
Ted

[This message has been edited by TedS (edited 01-23-2009).]

Top
Page 2 of 2 < 1 2

Moderator:  Admin, Diki, Kerry 



Help keep Synth Zone Online