Thanks to everyone for their input here. Definitely food for thought!

I shall have to think carefully about how I intend to use these, and make my decision accordingly. Some of us lowly composer don't have a lotta $$$ [:'(]

I like the idea of layering solo instruments over the ensemble patch. Of course, the day will come when ensemble samples will be a thing of the past, since more processing power and polyphony will one day make "true" ensembles be just that: collections of sampled solo instruments playing in unison where color is not only defined by the quality of the instrument itself, but also by the number of instruments used in the ensemble. A truly virtual orchestra, if you will, that will allow for as much flexibility as the real thing (or close to it).

Don't forget silence [[;)]]

Emlyn

thats not a dream [;)] There are people looking at developing just that.

I still think tho, that ensemble samples are a need. The sound of a 16 violin ensemble playing together will sound different than a mix of 16 seperate samples of each of those violin players playing. Not to mention it would be wholy different with all the other instruments in a room sitting there and not even playing. Each instrument would resonant with sympathetic frequenices..

all sorts of stuff that will make it never sound perfectly real... [:)]

still I can tell you that the idea you have is something thats being thought about, and early tests are extremely promising

Actually, what I meant was perhaps 10-20 samples of *different* players *for each* playing style and articulation to ensure variability. No doubt an enormously expensive undertaking, but not entirely impossible...

Of course, it will never replace the real thing!

I have to say that although I only have VSL and GOS of these three under discussion, all of the details in this very intelligent and knowledgeable thread pale in significance next to one thing - LEGATO.

Strings ARE legato. That is maybe 75% of what it is to be a string player. And only VSL has the real thing. It is not a matter of comparing many different details, it is an enormous, single revolutionary accomplishment on the part of this company.

As I was recording some more tracks for a symphony I've been endlessly working on, I was thinking as I listened to the startling perfection of sound possible with VSL that this is what has always been missing from sampling. It is the one missing element that is absolutely crucial to true musical expression - the actual connection between notes, not just the notes themselves. That is how a player makes his instrument "sing" and it can now truly be accomplished with samples.

@William said:

That is maybe 75% of what it is to be a string player. And only VSL has the real thing. It is not a matter of comparing many different details, it is an enormous, single revolutionary accomplishment on the part of this company.

This leaves 5% for intonation, 5% Technique, 5% musicality, and 10% money for buying the strings -or Beer [:D]

Iwan

I too have and use all three libraries. and as with the others I agree with KI.
SI is great out of the box. The SI strings are wonderful for sublte and delicate strings. The sound is wonderful. SI also offers a range of string effects and non standard articulations that are very handy for scoring.
VSL is really great sounding too. The legato is unbeatable. Once the rest of the articulations are avalible, it will probably become the defacto string library.
GOS has some good patches, but I find that I don't use it at all with the other two availible.

@Iwan Roth said:

This leaves 5% for intonation, 5% Technique, 5% musicality, and 10% money for buying the strings -or Beer [:D]

Iwan

Hahahaha! I don't think string players drink beer... Isn't it port, or cherry or something? [[;)]]

While we are on the subject, can anyone say anything about the Miroslav Vitous library in comparison to VSL?

Thanks,

Paul

M vitous was great for its time, but definately no comparison to VSL. Some of the winds have a nice character and I still use the solo trumpet - love the tone not the flexibility. Overall the library is a bit sluggish.

What Miraslav had was beautiful - all of the recordings were excellent sound quality, and the performers were probably at least as good as the VSL. However, it was from about ten years ago, and VSL has literally ten times the data (actually somewhat more) and includes multiple dynamics on chromatic samples - none of which Miraslav had. Also, and as crucially, the MV never had anything remotely like the alternation or legato tools.

I am reworking a major recording project I did with the Miraslav Vitous samples to use VSL, and what I notice most is that with MV even though there were many great sounds much of the time one would have to fake something - using filter and volume to get a pp out of an mf sample, or overlapping ends and beginnings of samples to get a "legato" effect. However, with the VSL, you do not have to fake anything, even subtle things, and it makes all the difference in the world.

My biggest problem with Miroslav was it's imbedded musicality - some notes would swell or do a cresc then dim, etc. This made it fantastic to audition, but once I started trying to bend its will to my pieces I found that it invariably would be doing the opposite of what I wanted it to be doing musically. I eventually ended up using AO more often than not, because it was easier to shape the notes how I needed. The only alternative for me was to compose at the keyboard with Miraslov preloaded (instead of in front of staff paper) and let the samples dictate the composition.

The sound of the hall in the ensemble samples was nice, though, especially on the brass ensembles (disclaimer: I'm not trying to start a thread on the merits of samples recorded with concert hall ambience.)

actually, string players like a nice cup of tea [[;)]]
though there may be some national differences in this respect