Right.
Here's a quick and dirty generic guide as i don't use Kontakt, and i last used Cubase v5.
If you are using a stereo signal into a single channel, then you would record the three notes at once. (generally speaking). That means, if your three notes for trumpets are E4, G4, and C5 (with middle C as C4), then you will use a 'sample for each note.'
(I don't know if VSL Opus is recorded by semitone or tone, but using the notes i described, this wouldn't be relevant anyway.) And with a single channel that also means you'll have to try widening or narrowing your stereo width to 'fill the correct space'.
Even then, that means you'll have all three trumpets occupying ALL the trumpet section space equally, so the sound will possible sound 'false', or even too big, hence a need to readjust the balance, with other sections.
If you choose three mono signals in, each with it's own channel (and you'll find a thread in the forum about mono versus stereo, in which i express my preference for mono over stereo, and in which others enthusiastically disagree, so you'll have to decide for yuorself which method to use.) then you'll have three channels with which to place correctly. (See the previous post about placement.)
Either way, if you use three different notes, then your DAW will use three different samples.
If you wish to use three trumpets in unison (all playing the same note), then i'm afraid i can't help further, because i don't know how well Opus or Kontakt will handle using the same sample three times in the same space. Again, this is only my opinion, but if i were forced to do this, i would record the first voice (instrument), convert it to audio and save it to an audio track, record the second voice and save to a second audio track, and then record the third voice and save it to a third audio track.
That way, you can, if faced with any phase or other issues, you can manipulate the audio recording in each track to counter this. And, in addition, you can then place your instruments more precisely, instead of relying on a single channel with stereo width to do the same job. Again, this is about what sounds right to you, and is only my opinion according to my working method.
(Dedicated mono channel for each instrument.)
I hope those chaps who use Opus will step in here, and give you a more precise answer specific to the library, and with the unique working methods of using kontakt, but from a generic point of view, maybe this will get you started towards a better brass sound. It's also worth saying that, in my modest software effects experience, having separate channels for each instrument has also enabled me to brighten and 'fatten' brass sound more easily and quickly, and have greater control over each instrument, including EQ to brighten or dull one instrument in order to obtain a better balance, according to my perception of what i want.
Remember that if you apply an effect to a sample THEN record it to audio, you're stuck with it, imprinted in the audio segment. If it's not suitable, then you'll have to record the same line again, so it's worth making sure you get the inital recording right first. And again i refer you to the podcast, with Michael's comment about applying a little pan to each voice you record before applying 'verb, recording to audio, or adding other effects. I'll only speak for myself, but i've done this for some time, and Michael's advice is, in my opinion, sound.
Regards,
Alex.
p.s. The Vi magazine is excellent, with a LOT of great tips, tricks, and examples of working with software instruments. Well worth the price.