Menu
What's new
By:
Filters

Your favorite MIDI orchestration tip

mcalis said:
You can do a fake re-bow with most strings sustain/legato patches by adding a quick steep "mountain peak" CC1 curve like so:

Faux Rebow.PNG

The results differ per strings library of course, and it doesn't quite sound like a re-bow, but in a mix it certainly can give the appearance of there being two repeated notes instead of one sustained note.
Click to expand...
I tend to insert a short silence before the peak to emulate this.
I even tried to script it (using Scripter in LPX).
The outcome totally depends on the library used (okayish in OT, doesn’t work with Spitfire, I didn’t try more).
Anyway it sounds a bit fake - especially on ensemble - and one should avoid exposed passages to use the trick.
 
jaketanner said:
This is true...as much as live players use dynamics, it's not that big of a leap within a phrase.
Click to expand...
This is arguably one of the more valuable things one has to learn when driving CC1 automation. I myself am trying to "unlearn" the incessant moving up-and-down of my CC1 fader..
 
berlin87 said:
This is arguably one of the more valuable things one has to learn when driving CC1 automation. I myself am trying to "unlearn" the incessant moving up-and-down of my CC1 fader..
Click to expand...
Since I can't play stuff in, I just do some lazy clicks with mod (e.g. sometimes very very little to no movement) in ensemble contexts. I did this with a few Williams and Goldsmith mocksup. Then I tested it against the "try to impart lots of expression" (whatever that means). The first (lazy) approach sounded much better.

Phew, a nice time-saver it was.

But as Dave says, context-dependant so use your ears. I'm just glad that now I know that in certain circumstances I need to do far less CC massaging than I thought I needed.
 
berlin87 said:
This is arguably one of the more valuable things one has to learn when driving CC1 automation. I myself am trying to "unlearn" the incessant moving up-and-down of my CC1 fader..
Click to expand...
I find that it is easier to make markings on your wheel OR your fader where the crossfades are between dynamics. With sliders, it's easier if you use a strip of artist tape (white....same used on recording consoles) and have various markings on each per library. What would be cool to implement, is a mod wheel that has light indicators when switching between velocity dynamics...similar to how the button lights on smaller keys to indicate the octave you are in.
 
One other thing to consider when using the mod wheel...think in terms of the actual markings on the score...most music doesn't go from p to ff very often...it will be on a specific dynamic for a bit, then gradually move on to another.
 
There's never enough dynamic layers to imitate a score in a realistic manner. Going from mp to mf, transferred into dynamic markings is like moving the CC wheel +10 or something similarly mundane.

Rather than go from 5 to 120 on CC1 to get some life out of samples, I prefer to go from say 60 to 80 on CC1, but also have a similar motion on CC11 and have the velocity match that increment as well. If that's not enough, I also go for CC7 a bit. Numbers vary according to the sample library you're using, also CCs, but the point is that the entire instrument suffers the effects of more intense playing and I have to hear that when I press play, so I take the time to reflect that concept on any controllable parameter I can use. The reason samples sound lifeless sometimes is because they don't have enough dynamics as the real thing, and because of stiff tempos.
 
The thing that we try to capture with the controllers mentioned here is an envelope. If it isn’t there in the sample we have to put it there. It may be very subtle or pronounced. Plus we’re not just dealing with samples but with computers which are going to do absolutely nothing until told to. This sort of default, rigid, stasis is the inverse of musical performance, which is full of both unintended and intended variation. That’s why you have interfaces with options to introduce tuning inaccuracies randomly. Not to mention loops of noises and squeaks in a hall to bring in that imperfection.

Much of controller function is not to get a sample to do something but to stop doing something. A sustained note in the strings will never be at the identical volume from start to finish but many samples will do that every time out. That is, the sample left alone will do things wrong at the rate of 100% of the time. Which is to say nothing of creating expression. So if you want a volume arc in any direction you have to put that in but you can’t stop there because you’re handing things over to the computer which will do nothing - which in many cases will sound very odd. A real player will still add expression after following an indicated one. Or they will simply ease off in preparation for the next bowing. But they won’t do 0. Only a computer can do that.
 
Last edited:
Talking about envelopes: rendering to audio, and then further shaping the samples' envelope for long notes is often a necessity, since original samples were recorded at a certain pace, with a specific vibrato timing which may not work well at a different tempo.
Long notes often have dynamic "bumps" at undesired locations, due to not so great looping.
Evening out those bumps makes the line more controlled and musical.
 
Last edited:
I like using DB to double tuba. I might not do it in a real situation all of the time but in sample mock-ups it fortifies the tuba and low end in general. sometimes a sub-bass ever so slightly underneath.

I also like to double spiccato with pizz accents on the strong beats.

Additionally, I like to add a nonobtrusive pad to the strings to fill out the texture. None of this is news but it's what I like to do. I tend to use step entry more these days and then futz with the notes in the piano roll after. There are phrases that real-time keyboard playing just cannot accomplish well for orchestral mock-ups.
 
I have some shortcuts that I set up only recently that speed up my workflow quite a bit which is:
Shift - C (copy selected notes and set locators)
Shift - X (copy and delete selected notes and set locators)
Shift - V (paste selected notes at locator)

So the same as Ctrl - C etc does, just with setting the locators too. Really simple, but also really useful.

Also I did a video about how you can switch between sections in the key editor in Cubase.
 
berlin87 said:
Very interesting views @jaketanner @MauroPantin and @Dave Connor
One thing I've come across many a time was the notion of mainly riding the CC1 fader/wheel, and leaving CC11 mostly untouched. I am starting to think it should be the other way around — for most libraries, that is.
Click to expand...
Think of it this way...there is no volume knob on an acoustic instrument..it is solely controlled by dynamics and velocity. So you NEED to use the dynamics as a volume...however since they are samples, and a limited number of sampled dynamics, you need the "help" of the expression control to fill in the blanks of missing dynamics (between the dynamics that are sampled)...if that makes sense.
 
jaketanner said:
Think of it this way...there is no volume knob on an acoustic instrument..it is solely controlled by dynamics and velocity. So you NEED to use the dynamics as a volume...however since they are samples, and a limited number of sampled dynamics, you need the "help" of the expression control to fill in the blanks of missing dynamics (between the dynamics that are sampled)...if that makes sense.
Click to expand...
100% makes sense -- Especially for woodwinds libraries, I notice :)
 
It's just a matter of variance. As an example, 5 dynamic layers with a volume curve that covers the 128 CC1 values is one thing. But, that same thing, times the 128 possible CC11 values to control range is so smooth it might as well be analog. It is still discreet in terms of absolute values rather than an analog continuum, which would be more akin to a real instrument, but at that point I feel like it has enough values to fake it.

EDIT: BTW, I gotta say this is easily one of the most helpful threads in here. Thanks everybody for chiming in with awesome stuff!
 
Start with the intention to make a simple piano ballad - not orchestral. When you have filled in all countermelodies, ear candy, horns etc. - which you cant resist - its suddenly grand orchestral. And delicate. It was not meant to be orchestral, but - voila - its perfectly orchestral.
 
You must log in or register to reply here.
Top Bottom