As others have said, one has to play the samples, literally checking each note, and thinking in very detailed terms.  It takes a great deal of work.  One thing that might help would be to match the volume of the different samples in your matrix.  The loudness of the different articulations vary, and while a player can automatically adjust bow pressure/speed, etc., to create a pleasing line using different bow techniques, that is very difficult to achieve simply by using raw samples.

Try using some of the different controllers to sculpt the line.  If you are only using velocity, the sound will likely be too static.

As a double-bassist, IMO, the VSL sample libraries (even in SE libraries, and the ones that I have), are as realistic as I have found in terms of the articulations offered, and techniques used in traditional orchestral and chamber music.

 That is a difficult question to answer as required time will vary from piece to piece.  IMO the key is to think as a performing musician for each instrumental part.  The difficulty comes in that a computer cannot naturally do what a player has learned to do... a good musician doesn't have to think consiously about how to do technique "x".  In essence you have to tell the computer how to play each line - raw samples in and of themselves do not equate to a musical performance.

BTW, based on this thread, it is not clear to me if you play an instrument yourself or not.  If it happens that you do not, or perhaps you play only piano or guitar, it might not be a bad idea to also take some lessons on an orchestral instrument of your choice.

Note also that VSL provides some ways to automate how a given sample is chosen, particularly if you are playing the parts in (though in my own situation, I prefer to use a more detailed method).

This is one of those questions that is absolutely not answerable.  It is something you simply cannot understand right now.  The fact is that wisdom comes after the experience.  No one can tell you how long it is going to take for you to work on a piece (programming etc).  You're seeking an answer that you'd be better off just investing the time "diving in" rather than wait around for an answer that you will find "suitable".  Because trust me, the second someone tells you "Oh it's easy, you can program one song per day easy!!  I do it all the time!!  yay!!" you're going to think ok cool!  I can do one song per day!  This is wonderful!

Then you buy Solo Strings 1 and find out you have absolutely no idea what you're doing or where to go.  Then you become discouraged and find yourself blaming everyone but yourself for not "giving you the answers you need to give you the results you need".  If you want real world results that are astonishing and realistic sounding, plan on investing a LOT of time.  If you want to create boring, static results that sound like a 1970 midi synthasizer then you can do a few per day.  Your best bet is to tweak and tweak and tweak until you find a sound you love.  Then build templates around that sound so you can just open up a new project and load that template and begin making music.  If you have no idea what sound you're going for, this is another thing that will put a huge wall in front of your forward progress.  You have to know what you want and be willing to work towards getting it.

Being a musician may help you write a score, but don't think that just because you know how to play the guitar, that it's going ot help you "program faster" in any sample library.  Those skills are completely different.  To do a complete song from start to finish in the sample world you need an ear for mixing and mastering (unless you plan to pay someone else to do it for you).  You need some understanding of how an instrument is played so you can pick the best samples to fit the passages for each instrument in a song.  You need to program and work relentlessly to smooth out the breaks in between samples.  Apply just the right amount of reverb, EQ out bad frequencies.  These things are overwhelming at first, but like anything else get easier and easier over time.  But it will feel overwhelming at times if you're just starting out.  One day though, you will look back and wonder why you struggled so hard, because now it seems second nature.

Ask yourself, "Am I willing to work HARD to get the results I want?  Am I willing to invest how much ever time it takes to get what I want?".  If so then you will succeed.  If you're looking for the "no bake cookie recipe" that you can stick in the microwave and be done in 5 minutes, it's not going to happen.  You will fail and become discouraged and blame everyone (especially VSL) but yourself. 

Even to this day, the more I learn, the more I go back to my older pieces that I thought sounded great and realize, how my new skill sets and templates would make them sound even better.  So some pieces can take years to get them where you dreamed they could go :).

Maestro2be

The idea that programming should not take longer than writing - that is unfortunately not the case with samples most of the time, unless you want to ignore the possibilities available and do a mediocre quickie performance.   Programming can  take ten times longer than writing the piece, because you are essentially doing all the performances of the instruments.

I also noticed recently what cgernaey mentioned, listening to an earlier VSL First Edition recording I did that I thought was really good at the time.  Now it sounds very artificial and weird, because I did not have all the articulations that are now available.  There is a huge amount of expressive possibilities with the various articulations and dynamics in the complete Symphonic Cube and it takes time to use them. 

This got me to thinking about how the nature of a musical performance - though it is obviously intuitive and emotional, is still based on a QUANTITY of possible sounds.  In other words, good music is a matter of numbers - massive amounts of data -  not just beautiful sounds.  For example, the old Miroslav Vitous library which is still being sold with new packaging had really nice sounding recordings and performances.  But it was tiny in quantity - like ten samples spread across the entire range for sustain violns(!) no alternations, no articulations for legato, only a couple velocity layers,  etc.   So even though the original playing/recording  was good, it could never even begin to do a performance like VSL.  When the First Edition appeared, it was a revolution in sound, but still inadequate to do a complex symphonic performance.  Now however,  the VSL concept of applying raw power of numbers in a consistent way has proven to accomplish what couldn't be done in the past and you can actually do a serious performance of complex music at least as good as a live professional orchestra.  But this ability takes a large amount of work to harness.  You can ignore that work and do a quickie performance that will sound o.k., but to really use the potential of VSL at the highest level is no easier than playing a live instrument which takes years of practice or - perhaps more accurately - learning how to conduct a live orchestra.   

The larger collections will certainly give you more articulations as well as more velocity layers (I hope to eventually get some of them).  However, my suggestion would be to learn to use the current tools that you have first.  Simply spending money on additional libraries might help, but it also might not, as the issues you are facing seem to relate only indirectly to the sample content.

Looking at your original post again, did you make any alterations to the tempos in your original sequence?  One of the more important steps, IMO, is to program many tempo changes within the sequence file.  A straight midi file will sound very lifeless and mechanical as the tempo will be static.  A good conductor normally makes constant subtle changes, literally almost measure by measure and even within the beats of any given measure. 

This is one of those time consuming steps, as you have to listen to the sequence over and over thinking in terms of how you would conduct each phrase of the piece.

It is impossible to help much without hearing the actual music - you ought to post it and specific things could be suggested. 

@matteo said:

ok i search a shared web space.

4shared [:)]

I listened and had the impression that you were using velocity crossfade which makes the solo morph into a psuedo-ensemble at times.  Did you do that?   If you avoid that and use dynamic samples it might sound more espressivo. Also, more cc11 expression at the starts and ends of notes perhaps. 

For SE and SE plus I think the only dynamic samples you have are sforzando and fortepiano.  I have SE but not SE Plus so I could be wrong on that.