I was just over at the VSL site listening to some of the demos and came upon a bit of strangeness. The violin solo in Saint-Saens' Danse Macabre keeps popping it's pan position to different places depending on which sample is being played. I've listened several times, and at least on my system, it's happening. The result makes for an odd disjointed line. Can anyone else confirm this?
I hadn't heard that demo before, not the best VSL demo for sure. Most stereo libraries have inconsistent stereo positioning, especially in the close solo instruments. We had a great discussion about the subject a couple of months ago but I can't find it right now.
sounds to me like they just have a stereo mic set up close to the vln so when it crosses strings from upper to lower it goes from slightly left to slightly right. I think this is pretty common in solo vln recordings as I've hear it before. Though in a sample set it's way more pronounced because I'm sure that the player was probably moving around slightly too and with each note recorded seperately.....
It's really distracting. I don't know how the original dry sample sounds, but with the perspective on this recording, it sounds like the player moves over about 10 feet on the stage depending on the sample. Very disjointed.
It's not THAT closely miked. The player keeps moving between the samples and that causes the differences in stereo positioning.
Yeah I get you. That would be very close.
What I'm refering to is the way a vlnist holds the vln and where the strings are place. The low g-string attack would shoot out at about 8:00 while the high e-string attacks would be at about 1:00. Just a thought.
Though I do agree that the more obvious thing is that the player was squirming about impatiently trying to do the 10th take of a single stacc. note. I'm a player. I'll never sit for a sample session(unless they pay quite well. Like tens of thousands.) I don't have the patience for it. No way.
I go mono with the solo violin, and some of the other solo instruments. Often it's actually the better choice, if you want to push them back farther. Narrowing, as Nick suggested, also works just fine.
Even in cases where there's not a directional anomaly, some of the VSL instruments and even sections respond well to being narrowed and focused up. It's part of the design. I would go as far as to say one should narrow down many of the VSL instruments when you're going for a big-arc soundstage.
It's actually a good design (defects aside, of course), because you can always make an intimate recording less intimate by narrowing it and pushing it back with EQ, reverb, or convolution. You can't go the other direction easily, making something less intimate become more intimate. Compression and EQ can make something bone dry get a bit more intimate, but there are practical limits beyond which there's no payoff.
I've seen a better picture of the VSL setup but here is a partial view. The distance between the G and E strings is less than 1.5 inches on the bridge, I just measured it. (Not trying to prove Jose wrong but out of pure curiosity. ) The player can easily move the instrument for several inches even on a chair but I assume that the player was standing so there is even more room for unwanted movement.