Btw, David. This man plays the SWAM Violin extremely expressive using a wind controller instrument (which I am seriously considering now haha). It also gets even better since he plays THE most beautiful violin theme ever in my humble opinion…the main theme from Schindler’s List.
Those things are very interesting indeed, and I’d assume even more so if you’re already a wind player. I’m not sure what they’d offer over a good keyboard (on that note, can’t wait to get my hands on that Osmose!) + BBC2 or similar. Are they keys on these velocity sensitive, BTW?
Anyway, have to pretty much agree regarding the Schindler’s List theme, although there are some incredible themes in the classical repertoire as well. To think that this piece was considered “too simple” at some point… I don’t remember the details of who said what, but the conclusion was that Itzhak Perlman is going to play it, so it would be fine either way 
Personally, I’m actually leaning more and more towards simple themes with lots of expression. The classical violin gymnastics is certainly impressive, and can be very beautiful, but to me, it kind of gets old eventually. I prefer deeply emotional pieces, played by passionate players on great instruments. I think the only thing that compares to that is a truly great opera singer, but I’m not even sure about that sometimes!
With your knowledge Olofson wouldn’t it be possible to connect an electric violin to a plug like swam? And get the real deal but with the violin of choice?
I do have various ideas here, with some potential designs in mind, but I’m afraid an electric violin has all the same problems as the guitar + guitar synth approach, with latency, misdetections, limited expression etc.
What I want is accurate readings of bow speed, bow direction, bow weight (including bow on/off string), and bow position (sul pont/sul tasto), which can then be fed directly into SWAM, ordinary synths etc.
Same for the left hand (though SWAM doesn’t support any of this I think - yet); ideally, you want not just the position each string is stopped in, but also how it’s stopped, as in, how close to the fingerboard the string is (it’s not always optimal to push the string all the way down to the fingerboard!), how much weight is on the finger, and if possible, which part of the fingertip is used (tip vs pad), as that affects how much vibration the finger absorbs. All of those parameters affect the tone, and those parameters also vary through the cycle of the vibrato, which is a big part of why the vibrato is a unique “fingerprint” of the player, as well as a central part of the expressiveness of these instruments.
So, in short, if it is to REALLY be anything like playing a real violin, it gets really rather complex. Of course, it also becomes about as difficult to play as the real thing as well (I’m afraid it’s either that, or losing some of the expressiveness) - though you still have the huge advantage of being able to edit every last detail of the performance.
Hmm, I did not think about note velocity on wind instruments, I guess not? But that would take away one dimension.
I agree, expressive and passionate on longer notes and “simpler” melodies can connect deeper than fast, agile phrases.
You would use velocity to control something like attack intensity (hard or soft onset of the bow) or portamento speed, for example.
Indeed - though for the BBC2, I typically set these instruments up to use breath alone controls dynamics, and it’s agile enough for creating distinct onsets and all that, so that works pretty well.
Velocity is typically used for controlling slide speed (or fingered transitions or equivalent), but I’m not sure that’s the optimal way of doing it anyway… It’s difficult to control with precision, and it’s of course very different from “just” playing it as you do on a violin, or when singing, for example. (That’s why I’m thinking ribbon controller or similar for pitch on this MIDI violin thing I have in mind… It would literally play like the real thing in this regard as well.)
Have you found a good way to control the slide speed in a way that you can sync them to land on a specific beat? I’d like to call it “Measured Glissando”, but not sure if that is the proper term.
I haven’t even tried that, TBH. The problem with slides like that is they’re part of this whole expressive thing, so I’m not sure there’s much point in trying to time single slides like that. I think just playing a “run” of slides is a much better approximation of how you actually sing or play an emotional slide. There is melody and rhythm in the slide itself, so trying to approximate that with a generic mathematical function is going to be very limiting.
Which again leads to the same conclusion you have pointed out a few times. We need a better way to record transitions than keys. Ribbon controller, Roli Seaboard…some way of physically controlling the motion curve and timing of the transition.
Exactly! Unfortunately, those things tend to be difficult to control, but I don’t really think there’s a way around that… As I’m starting to realize with the bowed strings, as ridiculously difficult it is to learn these instruments, the result goes beyond just being able to play what you intend on them. I don’t think you can have truly free musical expression unless you find and learn an instrument that essentially allows you to translate your expression into sound on a more or less subconscious level. If you can hear it, you can certainly approximate it to a high degree through programming and whatnot, and I used to think that might actually be the ultimate form of expression, if developed enough, but I’m not so sure about that anymore.
The analytical, logical Dave is in a weird place these days…