I agree about investing in a good library(s) to begin with. The bottom line? You can't polish a turd. I invite anyone to post some pieces composed with free VST's that sound like the real deal. Of course learning the "under the hood" aspects are essential, but if you don't have good tools to begin with, it's not going to sound convincing.
The problem with your invitation is that 'the real deal' runs the whole spectrum from sugar-coated film scores and impeccable recording studios stocked with every 'U' on the planet to budget novice recordings done with $20 mics (there are even some genres and some libraries which are intentionally cheap sounding). While for the most part it is true you can't do all that much to improve upon inadequate recordings or poor performances, there always are things you can do to avoid the bad parts of the instrument and bring out the good.
The benefit of sampling is economy of scale- one solo instrument sample library might mean one 1-20+ hour recording which can then be sold or distributed to hundreds or thousands. A $100 piano might represent $10,000 in recording costs. The same benefit falls down to freeware, especially freeware samples extracted from a commercial sample set, where tens of thousands of dollars in gear, performers, and studio time can reduce the only variable to sample detail.
Sample detail in and of itself (i.e. number of pitches, rr's, vl's, articulations, mic positions) is often overrated. You probably use commercial sample libraries recorded in wholetone and minor thirds on a daily basis without knowing it (while there is some freeware even that is recorded chromatically), and there are countless expensive sample libraries with just two or three velocity layers. A sample library with low round robins can be used to great effect by avoiding ostinatos and repetitions in writing, and whether or not we really need articulations like
7 arco sul ponticello violas up-bow at pp with a paperclip 1/3rd the way down the D string is up for interpretation.
I won't argue that the better quality recordings and performances you have at your disposal, generally the better quality your end result will be, but I will contest that
all freeware is incapable of sounding convincing
just because it's free. There are some bad freeware libraries, just as there are some bad commercial sample libraries, but there are also some excellent freeware libraries. Again, you can make great sample libraries at any price sound bad by abusing their weaknesses (try super fast legato/trills in a sample library with slow true legato transitions), but you can also make sample libraries sound fantastic which otherwise have zero good reason to do so.
(the rest of the orchestra is real, but the trumpet is a very simple sample set (3vl, 1rr, chrom sustains)- what makes it sound remarkably decent is the phrasing and programming. While it appears fake in a few places to a brass player, I have no doubt this would easily fool a non-musician. This trumpet, however, fails in some other contexts, so again- my point that it is more about how well and in what context you use the tool rather than its quality is backed up)
&
https://vi-control.net/community/th...-with-a-funky-kawai-piano.70049/#post-4212483
(I've heard live recordings that sound more 'fake' than those)
(there's a reason I say most of the reason people buy sample libraries is the reverb on the demos... that's a ~10-sample harmonica patch)
https://instaud.io/PUD
(I had to double-check this one was samples. Just goes to show how important good release samples are; honestly more important than transitions)
