Artist
 |
As I listen to some of the fantastic big orchestra pieces that many of you create on your computers I really wonder about some details that I struggle with myself.
When you play on a keyboard, you can hit the key hard or soft and that is great for control of picked and plugged strings, mallets, drums, piano etc.
But for the classical bowed strings and winds, the musicians can alter the dynamics also within a single note or chord.
I have tried myself several ways.
I have tried to assign a physical control knob on my midi keyboard and used automation in logic so control volume.
I purchased a very nice Roland expression pedal which allows me to control dynamics with my foot. Better but difficult. And I tried recently to post record automation of a wah pedal only to learn that Logic is terrible at getting the timing right. It was as the sound I heard live when recording was destroyed by a delay of the automation when playing back. I managed to move the automation back in time so it fit. But it cannot be right that this is needed.
I have also tried to draw the dynamics with the mouse. I find it very difficult to get things right with that.
And also, when you only alter volume, it still is not right because the character of the bowed and brass instruments also changes with strength. Louder always sounds brighter as the instruments produce more overtones when played loud.
I would really like to hear how you guys get those wonderful dynamics into your orchestral pieces. |
|
|
Aug 05, 2012 | 7:42 am |
|
|
Artist
 |
The software I use is absolutely remarkable in its ability to create dynamic nuance. It's all done via musical notation. The software has the standard dynamics available for each instrument, from ppppp to fffff. There's also -, and + increments available on for each dynamic, so mf+, and mf- produce slightly differing dynamics. There are 33 different dynamics levels available for each sample of every instrument. The software has the ability to create crescendo and diminuendo using the < and > hairpins. So, I can notate in the score a ppp<f crescendo and it will sound. For example, one will hear remarkable ‘fadeouts’ using a dynamic entry like mf>ppppp.
You are correct, Kenneth, in that character of the attack and dynamics change with strength and attack. This is embedded in the sample in my software. If I want a French Horn line to go from p to f, the result will often be unacceptable, as the sound will not be realistic. This is solved by writing p<f. The crescendo hairpin will usually solve the problem and create a realistic sound. If that doesn’t work, I can try p+<f- for example.
In the sample libraries, articulation and instrument expression also play a role in dynamic profiling. If one enters a ‘sfz’ (sforzando - an instruction to play a note with sudden, strong emphasis), then one will get a sample with the sampled player playing a sfz; this affects the overall dynamics of the note. Also, instrument expression entered during notation, will affect the overall dynamic. If I enter, for example, a ‘brassy’ notation to a trumpet line, the sample will change automatically, and the result will be a ‘brassy’ trumpet with more presence and attack, than if I did not enter that expression in the score. This as well, affects the result of the dynamics.
Because everything is done via notation, if you looked at a score, you'd see hundreds and likely thousands (in larger pieces) of dynamic profiles in a piece. Each instrument has its own part, and each dynamic profile is individually notated, one dynamic at a time. Sounds time consuming, and it is. However, there are keyboard shortcuts that allows for quicker entry.
When the score is done, and as I am composing it, (I use no keyboard; I sit at the computer and create a virtual score using only the software) I tweak the overall sound using the mixer in the software. Each instrument and each part can be manipulated in the mixer. So, if a part overall is too quiet or too loud, I can adjust the dynamic profile of the entire part as I compose. I spend a lot of time doing this.
I think I am likely the only person on this site that does it this way. There are certainly quicker ways to do this. However, I am an admitted control freak and I’m not happy until the overall result sounds the best that the software and I can make it. |
|
|
|
Aug 05, 2012 | 8:30 am |
|
|
Artist
 |
Kenneth, you are good at articulating what's going wrong. I always just knew my sound just wasn't right. You seem to have tried all the things I've tried. At first I used volume exclusively; I'd pick a velocity that gave me the tone I liked for the entire piece and I'd dial the volume to where I wanted it for the swells and dynamics. Got the job done, but not very realistically, in retrospect.
Then I tried adjusting velocities. I got to the point where I didn't have to touch volume at all -- reserve that for the mix -- but this technique does better with fast moving pieces where there aren't many if any long duration notes.
Now I've discovered the expression editor in the Piano roll and it gives me what I need, for now. I'm writing an adagio for strings and I'm pleased. I edit the dynamics with the mouse, both velocity and expression, once I know what notes I want. I'm wringing some emotion out of this machine.
That said, it's clear there are better string libraries out there than Logic has out of the box. I'm shopping for one. Some libraries appear to have dozens of parameters! But whatever library, I am resigned to having to edit for hours, to get the strings just the way I like them. There are simply too many moving parts for each instrument, to grasp them on one take, short of hiring an orchestra. |
|
|
|
Oct 18, 2012 | 11:40 pm |
|
|
Artist
 |
I go in by hand after writing the notes I want (also usually by hand) and change the tones and volumes. A lot of times I layer layer layer (There will be three layers of viola playing one thing to be loud enough and get the right tone for example)....
I dunno really, Soulima probably covered it all. There are good string libraries out there. I have Garageband's expansion packs and have built a couple of my own kits to reflect my efforts to get the sounds I like to hear in my pieces. |
|
|
|
Oct 18, 2012 | 11:57 pm |
|
|
Artist
 |
@olympus - who is "Shadow"? Is that post in another entry? I liked to hear that you layered your voices. I didn't use to do this, but have in later pieces realised that it can add a real "body" to the sound. I've never done more than two layers, but I'll try three to see what extra emphasis that gives. To be fair, my pieces are rarely loud, so often I don't need the volume, it's more about the overtones sounding expressive.
I wish that the Garageband strings had a bit more reverb on them, to make them more realistic. I use them, but tend to do a lot of gentle playing to try to keep to a minimum anything that to my ear sounds synthetic. Some of the strings take attack much better, and with those I'm free to play as loud as I like with them, and often I'm pleased with the result, as it adds drama that i hadn't heard in my inner ear. I also like the free Violin and Cello samples from Boldt. |
|
|
|
Oct 19, 2012 | 6:52 am |
|
|
Artist
|
@ihussain oops, sorry, meant Soulima. I confused him with ShadownofNine when I wrote my comment. Edited to fix that. |
|
|
|
Oct 21, 2012 | 8:40 pm |
|
|
Artist
 |
Coming from someone whose only focus has been GarageBand, layering and especially reverb definitely help with making a more realistic atmosphere. Though, I think you already have a pretty good understanding of GarageBand and how to be convincing with it. My question is, are you really going towards pieces that will sound real in the real world or sound real the second you hit play? Or more simply, are you into making music digitally or do you prefer keeping the realism of practical melodies that could be played by live musicians?
In terms of actually creating the perfect swells and dynamics with GarageBand, that's the toil of note by note, and even then sometimes it's still wrong. Fixing that requires purchasing better instruments and platforms. But in terms of the actual sound, you can still fool people a little, or at least have your melodies sound full and effective on the surface. And like I said, that requires layering and especially reverb. |
|
|
|
Oct 21, 2012 | 10:54 pm |
|
|
|
Join Now
Login
