View Full Version : scoring for three trumptes (or horns or clarinets or...)

05-13-2005, 11:03 PM
As I understand things, the samples of 3 trumpets or 6 horns or 3 clarinets will be a possible choice in places where 3 trumpets or 6 horns or 3 clarinets are all playing the same line.

However, it is common in orchestral scores for the 3 trumpets to be playing different notes, and to use the sectional samples on three different notes would effectively create a 9-trumpet section.

So then, as long as three tracks need to be set aside for three solo trumpet voices in an extended symphonic work, whether they all play the same note or they play different notes for some particular passage or not, it seems that three tracks need to be allocated to the trumpet section in any case.

So then, why not just use three tracks each containing a solo trumpet throughout and dispense with using the sectional samples? Ah yes, but three clones of a single sample is not a section.

In other words, how is everyone else (for those of you who care to comment) assigning voices in a project? Three solo trumpets throughout? Three solo trumpets plus another track for the sectional samples? It seems in any case there must be a huge file with many more tracks than orchestral voices. And doesn't there need to be several tracks for each solo trumpet in order to accomodate a variety of articulations? Which means that the three trumpets actually need twelve to sixteen tracks between them? 700 tracks for a Wagnerian orchestra? Or is there some more clever way to assign tracks to samples for scoring a large orchestral work?


05-13-2005, 11:11 PM
I often use a section sample for 2 or more notes, and the results
are good enough for me.

I dont know how to solve the problem of creating sections with solo instruments,
at least within EWQLSO.

But as far as taking up a heap of midi tracks, you can use Keyswitches.
You can make you're own in Kontakt 2.

That's all i use actually. Otherwise i'd have 150 tracks at least.

I've got keyswitches with up to 10 articulations for one instrument/ midi channel.


Nick Phoenix
05-13-2005, 11:15 PM
You just use the solo trumpet. Play a chord and it's realistic. use the 4 trumpets patch for a solo section line.

05-14-2005, 12:56 PM
You can't just play the same sample 3 times at the same time, it phases. That is why you need to use the 3 tpts for unison. If we had tpt 1, 2,3 then you could play all at once.


05-14-2005, 01:12 PM
.. which is why a lot of people are waiting for them - far more solo instruments and smaller sections will be available. In Silver Pro (for example) I believe there is an additional Bb trumpet and a piccalo trumpet - so with the original solo in Silver you have 3 trumpets.


Stefan Podell
05-14-2005, 01:12 PM
If you want to have a section going in and out of unison, another approach is to load the same patch in multiple slots (in Kompakt or Kontakt). The sample data only gets loaded once, so the memory requirement isn't 3 times more. And it also solves the phasing problem (or cutoff problem).

You also, of course, have the option this way to have any separation you want (1 and 2 playing one line, 3 and 4 playing another line, for example).

Someone posted a tip once on NSS to get a little extra separation in the sounds: Have a fixed pitch shift on each patch (for example, 0, +1 half step, -1 half step, -2 halfstep). Then shift the other way in the MIDI channel (which should be a track parameter, so the note values don't have to change).

All that gets you (1) different samples for the same notes, (2) different velocities when playing unisons, and (3) different attacks and releases when playing unison. I'm no expert on this yet, but I like the effect better than switching between the solo and section patches. It's not really practical with big sections like strings, of course, but for woodwinds and brass it can be pretty great.

- Stefan

George Bellas
05-14-2005, 02:52 PM
Wow, what a great inquiry! I have been wondering this too.

The same question would apply when trying to score a Divisi section of strings. For example, if the patch used is a "24 1st Violins" section and you want to orchestrate a Divisi, of say 3 parts, that would effectively result in 72 1st Violins! Add your 2nd violin section and you've got one brobdingnagian string section! :eek:

I am indubitably curious if (and how) the upcoming "Pro Upgrades" contend this orchestration issue.

Stefan Podell
05-14-2005, 03:53 PM
I think the Pro updates have some 3 and 4 violin patches. So you could do pretty convincing divisi that way. 18V div a 2 = 3 3V patches, for example.

05-14-2005, 09:24 PM

Thank you!

The idea of using pitch-shifting and transposition to remedy the unison phase problem is positively ...

Brilliant! Ingenious! That's exactly one sort of thing that I was looking for.

Yes, George, your observation was a major reason for posting my question. Putting it in terms of the trumpets kept the problem simple to discuss, but yes, this orchestrational dillemma is even more pronounced when scoring for divisi strings.

I'm still looking for clever ways to approach the track-assignment problem of not wanting to create a 700-track file. Kontakt and keyswitches, eh? Hmmm. I'll have to look into that.

Any other suggestions on efficient ways to rotate track assignments among the voices in an orchestra?

A major consideration, of course, is that any good method must allow for appending short files together without creating additional problems.

Yes, Nick, thanks for making that point clear. The word 'realistic' is key. Sixteen trumpets on a chord is not realistic, nor is four identical, simultaneous samples on a solo line, which is why there are both solo and sectional samples. Now if I could just figure out how to juggle such a wealth of resources without overtaxing either my system or my brain, I'd be even further ahead.

Part of my question goes toward the following thoughts: What are some of the more efficient production strategies? Rendering each of the four choirs of instruments into audio (and archiving the MIDI in case it later needs to be revised) such that only one choir at a time is being edited in MIDI, rendering into audio all but a current thematic group that is being edited, or keeping the whole orcherstra in the MIDI realm while creating the tracks?


05-15-2005, 08:46 AM
Thanks for starting this thread, I was wondering about this too!

I had another question to tack on, hopefully it isn't a stupid one... but how noticeable is the panning in all of this? Whenever you have 4 solo trumpets playing, they are all coming from the same "player" so to speak (the same position in the panning field), rather than 4 separate positions... I thought about offsetting the close mics to remedy this, and maybe even altering the left/right volumes of the other two mic sets (as discussed in the manual to change panning)... but this seems like a lot of work, especially if it sounds realistic enough to just have the 4 trumpet lines come from the same "player" ... I guess I'll experiment when I finally get platinum installed, but still... would be nice to know how others feel...

05-15-2005, 10:07 PM
For me the ultimate goal is to have sufficient samples and computer power to actually have enough different high quality solo instruments to be able to write freely.
This of course doesn't help much if you are keen on staying within one library only (for understandable reasons - $$$ ;) ).

What others have allready suggested is the way to go - load up the solo trumpet and ensemble patch, write your divisi with the solo and unisons with the ensemble.
Now, if EWQLSO were available for Gigastudio, you could do all this within one line thanks to patch changing. ;)