Menu
What's new
By:
Filters

Your favorite MIDI orchestration tip

I tend to play it in LIVE, then I drag that to the right and remake it note by note on the left by taking the live note length and rounding it up or down, where it doesn't change the timing, it keeps everything equal, in time and less confusing when dealing with sustain. Then randomize the velocity between X to X then I think 80% of the human touch is then MIDI automation. So with this style you get timing, cleanliness (visual simplicity), velocity changes and MIDI automation (all of them). I have learned that MIDI automation should most likely be last... as if you have to remove, add or reprogram any notes, you affect the automations.
 
My best tip when using midi is to play it live if you can and only quantize it partially. You want to keep things slightly humanized still.
 
I have a new one that I accidentally came across. This won't apply if you don't layer libraries.

I had a timpani playing two hits fairly closely together. I used two libraries: Hollywood Percussion and Abbey Road One Foundations.

Now, Abbey Road One sounds pretty glorious, but if I only used it, the second hit gets somewhat muddled by the built-in tail/reverb from the first hit.

What I ended up with was HOP hitting both notes using a fairly dry mic (no added reverb or eq), and had the Abbey Road timpani only hit the 2nd note. See the image below:

dTYPF.png


You get the clarity of the first hit from HOP, and in the 2nd hit you get the space from AR1. If you'd only use HOP, it would be clear but there would be no space. If you'd only use AR1, every 2nd hit would be obscured by the tail from the hit preceding it.

Sound example (point being here that you probably would not have guessed it's two libraries. Both clarity and room are maintained):



In other words: when notes are rapidly succeeding each other, consider using a dry/drier sample to prevent tail buildup and only use the "roomy" sample(s) when the final note rings out.
 
Previously missed this thread somehow?

My midi programming tip is simply.... start trills late/after the beat.

When a player plays a trill they often attack the note(s) with gusto and the physical prep for this usually means they are late.

Honestly - it sounds nuts, but I find it helps realism.
 
Very interesting thread.
Where I'm "usually" at now, is playing everything in with a keyboard. I'll play to (but more "around") a click track where there is nothing there at the start, but once I have the first instrument or 2 in place I most always turn the click off and play "with" the first players that went down. I'll do tweaks with a pencil if I can't get it right in a few passes, but I never snap to the grid, even though I'm watching the grid. I'll look at where the played part sits, and adjust the part I can't quite get right, by ear, with the help of seeing the grid, but making my ear and feeling about it be the boss.
I'm **always** in Piano Roll view when recording parts.

I really like velocity sensitive instruments, but when they're not I find a way to make them feel the part with CC01 or CC11. I sometimes play my controllers in on a second pass.
I often draw in tempo variations, during and later, as the piece progresses, but totally based on how it feels. Not because I think I need to do it.

I'll occasionally drag copy an area of bars to save time, because I hear the (general) repetition sneaking up on me right around the corner ... but I always change or add something different to the copie(s) to make them different from the first ... extra harmony, different harmony, variation in the melody ... or drop things out. It depends on what the piece "tells me it wants."
Which brings me to the main thing that I've learned that is most important for me ... is to trust my inner voice, and feeling ... my internal clock, my core sensations ... that place from which it comes that still seems to be a magical gift from the unknown.
I find for me that there's a delicate balance of thought and sensation that makes it happen for me.

I listen to a lot of old dead composers. Hindemith is this weeks go-to.
The more I listen the more it then effects how I compose and play.
I've always been a feel player.

I agree with a lot I've seen in this thread. One that I remember most is
learn your instruments ... how they play and feel under your fingers.

It's not a midi tip but the best thing I can suggest to anyone is always trust your inner voice.
The quiet one ... that isn't the verbal stream.

LOL ... I think Hindemith just played the lick from "In A Gadda Da Vida"!
I'm sure he quoted "blow the man down" a few times earlier.
 
Do you think samples are harsh? When I listen to live recordings, even during loud passages, the sound is still pleasing. However, with samples I have to constantly fight them. Quiet or loud, they are harsh. I'm really, really tired of doing this.
Is this just a matter of careful orchestration and midi/audio polishing, or is this just the nature of samples? Maybe I simply need better sounds? What's your experience?
 
Ivan M. said:
Do you think samples are harsh? When I listen to live recordings, even during loud passages, the sound is still pleasing. However, with samples I have to constantly fight them. Quiet or loud, they are harsh. I'm really, really tired of doing this.
Is this just a matter of careful orchestration and midi/audio polishing, or is this just the nature of samples? Maybe I simply need better sounds? What's your experience?
Click to expand...
I believe even Cinesamples has a video showing to cut the mid for this reason. Not so much they’re harsh, but if you hear a live orchestra, you’re not hearing it up close. Maybe the conductors perspective would be different, but from an audience perspective, you won’t hear the rosin on the bow. So depending on you’re needs, yes…cut the mid that are causing this.
 
I think also you can refer to a talented mix engineer, because they often know and hear things we don't and can make a big difference to a mix
 
novaburst said:
I think also you can refer to a talented mix engineer, because they often know and hear things we don't and can make a big difference to a mix
Click to expand...
True, and there aren't only files that are meant for streaming or CD production that are boing through a mastering process after the mixing.
 
Ivan M. said:
Do you think samples are harsh? When I listen to live recordings, even during loud passages, the sound is still pleasing. However, with samples I have to constantly fight them. Quiet or loud, they are harsh. I'm really, really tired of doing this.
Is this just a matter of careful orchestration and midi/audio polishing, or is this just the nature of samples? Maybe I simply need better sounds? What's your experience?
Click to expand...
My suspicion is that it has to do with instruments or sections being sampled in a room with no one else in it but those players. For that reason most sampled instruments cover more of the frequency spectrum than they would in context. That is to say: sampled instruments are generally more full sounding than they would be in context.

It might seem somewhat ridiculous but the body mass of other people and their instruments would certainly impact the sound. There is a difference between sound waves bouncing around in a barren room vs sound waves bouncing off of other people and (resonating) instruments, even if they're not playing.

If that sounds insane to you, consider that in some spitfire audio samples from AIR, you can hear the organ in that space resonating.

So the somewhat simplistic but truthful answer to your question is to use EQ. Now it's impossible to make specific recommendations but in general I find 200hz and 400hz to be problematic in almost all libraries because that's the area where there is a lot of low-mid build up.

Harshness is likely to be in the 2k range or higher, but that really depends on the instrument of course. Cinesamples has a whole EQ series on their channel that may help out:



Joël Dollié also has a lot of very good videos on EQ (and other processing) specifically for orchestral music.

The best thing you can probably do is get a reference soundtrack, something that is as unprocessed as possible, preferably only has the decca tree/AB sound in it and that isn't too dense in the orchestration, then try to match to that.

Narrowing the stereo image can also help in some cases because (hall recorded) samples generally occupy too much space both in the frequency spectrum and in the stereo width.

In addition to EQ you can also look at the attack and release settings if they're available. A shorter release can result in a drier sound which may be desirable if there are annoying frequencies in the tail. This Metropolis Ark 4 video shows that quite effectively:



Depending on how far you want to go, your final resort is to reprogram/resample instruments. It's not good for your sanity, but it will allow you to remedy serious imbalances in some of the commercial offerings and will allow you to make the necessary EQ, ADHSR and stereo width adjustments at the source.
 
Last edited:
You must log in or register to reply here.
Top Bottom