Sir Herb,
WOW... The tip about Attack (50 - 64) is simply wonderful...! Thanks.
-
You shouldn't have to "play sloppy" or expect that the 50ms time threshold is universally suited to work with the character of every possible phrase at any given tempo. (Actually, if I'm not mistaken, the threshold varies from patch to patch. If I'm wrong about that, someone please correct me).
If you don't play joined-up notes you can't expect the Vienna Instrument to produce a joined-up line - if you lift your fingers beween notes the note will stop playing and the line is no longer legato. As far as I know the sub-50ms threshold is a constant for all VSL legatos and portamentos, but since you can't gauge 50ms while you're playing, it's better to aim for overlapping notes.
-
They may not be perfect, but at least they prevent legato notes from sounding obviously disconnected.
But that's precisely the problem --- they DO sound disconnected. Listen to the Pavanne demo. Yes, it's beautifully executed (and my compliments to Jay!). It's making the best of the VSL library. But what bothers me tremendously (and this has nothing to do with Jay's work) is the constant stream of "bumps" on the attacks of the slurred phrases. In particular, listen to the homophonic section around 2:15. Every note starts with a "bump". That's not legato. Listen to the bassoon at 4:37. I highly doubt that the score indicates, "play with funny pitch bend on the attack of each note in this phrase". Same with the descending flute + basson line at around 4:26. Again, not a reflection of the beautiful job that Jay did. It's about the samples...
My purpose in posting here was to offer some "relief" to the OP who seemed to be struggling with the legato samples as myself, not to rehash my original point. But hearing what I hear in Pavanne and other demos, at least I can rule out some kind of playback problem with my copy of VI. I'm not saying that to be smarmy. It was a concern at one point, because, despite the documentation I was also able to get legato patches to play polyphonically under certain conditions...
Finally, regarding the 50 ms time threshold, I double-checked the documentation and you are correct.
@ Herb,
Thanks for your reply. But honestly, when I hear C-->B as I do at the top of the .mp3 and I don't see those notes in the score, the last thing I'd think to do is to look halfway into the score to see if those notes occur. In any event, thanks for clarifying where the files go. I suggest that the documentation be made clearer for these tutorials.
-
hm i wonder, does that also work for the older cube and horizon products?
or is there a difference in the way the sample attacks are edited in the VI versions as opposed to the older ones, that calls for the need to have a attack envelope active?
I did try it out though, and it seem like a very good idea for strings especialy :D
-
Hi,
as Conquer wrote: "I was just trying to emphasize how important it is to make the notes overlap to trigger the legato intervals."
Being a classically trained professional pianist but now more towards pop/jazz, I find no problem at all to execute the overlapping playing technique. The trick is to play "deep" legato, where your fingers are kind of glued to the keys so that you release the previous key "late."
Then the VSL approach towards legato makes sense.
I don't mean to be mean, but just a notion that it is possible to play it well.
All the best,
Hannu
-
Yeah... I have to say, Peter, that I thoroughly disagree with your criticisms of the legato instruments. To me they produce more realism than any other library out there, and by a very long shot. But the reason they produce this is not because they are perfectly smooth, but rather because they aren't perfectly smooth. Really, the "smoothest" transition you could get between two sampled notes would be to simply extend the release time of the first note, and have it bleed over into the next note. That's what a lot of libraries do. But the reason this sounds fake is that there's no transition sound. I mean, I appreciate your training, and all, but what's great about the VSL legato is that it's actually retaining the messiness of a real legato - not because the player isn't playing it properly, but because there are physical imperatives within the instrument itself that make a perfectly smooth transition impossible. Even the transition of a fingered ascending major 2nd on violin produces some noise, because you're instantaneously changing the length of the string - there are artefacts introduced into the sound as a result, and VSL's technique captures these. The same is true for the winds and brass... I'm not trying to discredit your experience, or your ears for that matter, but I just don't hear it the way you do, obviously.
Now, I do think that the legato sampling technique is not ideal for large ensembles, like tutti strings. I think this is probably because, although the real legato transition is still being sampled, the nature of the sampler still requires that the transition be realized as a single event - i.e., the transition from one discrete midi note to another. This, of course, is not the case with a real string section, but it can't really be avoided with a single sampler instrument. And I don't honestly think there's any way around this, except for the layering techniques people have already mentioned... Maybe they could internally duplicate the sound a few times, "humanize" the midi events, then mix them all in the output... dunno... something like that...
Anyway, that really is just my 2 cents. Personally, I find very little fault with the legato instruments.
J.
