String arranging

[Phil]

Hi,
just wondering if anyone\'s got any tips on arranging strings so that they sound realistic. I\'m using the AO strings sets, but it\'s all sounding a bit too much like an organ (?). I\'ve tried various panning, slight shifting of timings, etc, but it\'s not sounding too great. It doesn\'t help that GS keeps changing the instrument files that I\'ve loaded up, but that\'s a whole different post.
Thanks
Phil

[clueless]

I find that the AO strings are very sensitive. They can\'t be played straight from the keyboard. I go into the sequencer and use the expression control almost continuously. All aspects of a string part must be manipulated -- attack, sustain, decay, note transition, ... I also regularly switch between the various articulations available.

Many times I layer with other string libraries. However, the layering must be dynamic. Ultimate strings complements the AO sound very well. I am a perfectionist and, frankly, even after all the time-consuming tweeking, there is still much room for improvement.

Sometimes a blend of string section and solo string (at a lower volume and panned) makes for a satisfying sound.

Clueless

[Ed Lima]

Everything clueless said is true. You may also want to look into a MIDI wind controller like the Yamaha WX5. It does a lot to improve phrasing and dynamics in string parts.

------------------
Ed Lima
ELM&M
info@edlima.com
www.edlima.com

[jdeborde]

Make sure you\'re using midi controller #7, Volume, to almost always be either increasing or decreasing the volume throughout a phrase or note, as a real player would. Also, if it\'s a chordal passage, go through and play each line (voice) one note a a time, rather than grabbing the chords. Think like a string player rather than a keyboardist, bearing in mind that it\'s virtually (pun intended) impossible to duplicate a real string section, your efforts can only bring you closer to the real thing.

good luck!

[Phil]

Thanks, I\'ll give it a go and see what happens.
Do you think the best results come from using the XFD instruments or the ensemble instrument sets (instr KEY)in the AO library?

[Jamieh]

Ed, are you using the WX5 with GigaStudio? I have been debating getting one for a while, and I wondered what your experience with it has been?

[Ewen]

I have the AO strings and Kirk Hunter\'s Virtuoso Strings. After I got the Kirk Hunter Strings, I haven\'t used the AO Strings anymore. (It\'s a huge difference between those libraries)

You should consider \"upgrading\" to Kirk Hunter\'s Virtuoso Series String Orchestra, if you want that \"Hollywood\" sound.

I\'m very satisfied about it...

Ewen

[This message has been edited by Ewen (edited 04-12-2001).]

[Ed Lima]

Jamieh, yeah I use the WX5 to control AO samples in Gigastudio. It works extremely well on expressive solo parts. I\'ll use it when layering a solo violin part that is identical to the first violins part, but the solo sound adds some \"rosin\", some spirit to the phrase. One small thing, though: on my unit at least, it seems that the pitch bend wheel doesn\'t reset all the way back to zero when you let go, leaving the sample a bit detuned. Sending a simple reset from the sequencer clears it up, but you should know that that can happen.

Through use of the WX5, I\'ve discovered a way to mix while sequencing that expands on John DeBorde\'s idea of using CC7 to swell and decay individual notes. The thing is that AO samples (if not all of Giga) respond to CC11 (expression) exactly as it does to CC7 (volume). I can set the WX5 to translate breath strength into either controller, but by setting it to transmit CC11, I can then use CC7 in my sequencer to mix relative instrument volumes, and use CC11 strictly to swell or decay instrument sounds. And by simply dropping in one CC11=127 message, you reset the individual instrument back to full strength without fudging up the mix. In other words, you can basically swell while performing the part without messing up any mix settings in your sequencer.

Also, just to emphasize John\'s earlier point:

>Also, if it\'s a chordal passage, go through
>and play each line (voice) one note a a
>time, rather than grabbing the chords

That is SOOO true. I won\'t lie to you. I\'ll sit down and sketch a six-staff string arrangement for a piece, just to make sure that I\'ve got the spacings and doublings I want. Then I\'ll read each part in on the WX5, one part at a time. I\'ve never been able to \"fly\" those kinds of passages straight into the computer off the top of my head. They sound really poor and amateurish when I do that. Of course, I\'m sure that there are as many opinions as there are users on this forum, but I think everyone would agree with John in that you should definitely get your string parts in one at a time.

------------------
Ed Lima
ELM&M
info@edlima.com www.edlima.com (\"http://www.edlima.com\")

[This message has been edited by Ed Lima (edited 04-13-2001).]

[jdeborde]

In my relatively limited use of AO (2 months), I have found the XFD instruments pretty fake sounding when you go from p to f, so I usually go with the KEY option. That being said, I don\'t find myself using the AO strings much, except maybe layered with other libraries. I\'m still experimenting a lot though, trying to find an approach that works best for me.

my $0.02

[Bill]

Ed,

What value do you start off with for cc11? GSt’s output is so low to start off with. Setting cc11 down at the start really makes it feeble.

[Ed Lima]

Yeah, GSt is kinda wimpy when outputing signal, so I make CC11=127 the loudest, if it\'s a forte passage, anyway, so I\'ll usually start there. So if you\'re fading a note out, ramp CC11 down. To fade a note in, ramp CC11 up from zero to whatever value you want. Again, we\'re talking about NOT using CC11 to mix instrument volumes, only to emulate what happens at the mouthpiece or bow. If I want first violins louder overall, then okay, I\'ll boost CC7 so that everything is louder, but for \"relative\" changes, like swelling into this phrase, decrescendoing that phrase, etc., I\'ll use CC11.

The confusing part is this: perceptually, the two controllers are identical. I\'m only making a point of distinguishing the two in my projects because by having two controllers alter my data the same way, I can use volume solely like an engineer would, while using expression solely like the performer would. I hope I\'m explaining this well.. if I\'m being confusing or whatever, by all means flame away.

------------------
Ed Lima
ELM&M
info@edlima.com www.edlima.com (\"http://www.edlima.com\")

[This message has been edited by Ed Lima (edited 04-13-2001).]

[Jamesmcwilliams]

I\'ve got AO strings, and the instruments that i use the most are the Soft string sustains, contrabass, and KEY.

Many of AO strings sound organ like, but it really does depend on how you use them, I don\'t think that AO is very well suited for fast passages.

Also i agree that you should write string arrangements one note at a time, rather than using chords on the keyboard, i have got much better results this way.

[Lucas]

One thing that made AO strings usable for me is to play with the attack and release. I can
get some very nice sounding slow passages. Still working on the fast stuff, though.

[clueless]

This comment is regarding the discussions about CC7 and CC11. I use MOTU Performer and find volume control (CC7) useful to set a general instrument level for sections of a piece or for the entire work. Whereas the expression control, CC11, is good for individual note-by-note variations. I display the volume levels and pan positions on a second computer screen as a virtual mixer. But expression levels are not displayed (except in the normal piano-roll windows).

Maybe someone can enlighten me as to whether there is fundamentally a difference between CC7 and CC11. For example, do some instruments change volume and filter values simultaneously with expression value?

I may not be using them in the most efficient manner. I would be interested to hear how others use them.

Clueless

[Gulliver]

Phil,

Working with AO Strings set can be very frustrating because you\'re essentially stuck with an ff, pp, and a detache sample. For legato lines you can mix both ff and pp layers on two tracks in your sequencer and continuously adjust the balance between them. If your string line is playing a more aggressive legato passage, push the ff expression envelope higher and cut down on the pp counterpart. Switching between the two layers midstream can be distracting and this way you have better control than you would if you tried GStudio\'s crossfading features. For staccato and other short aggressive lines you can use both detache and ff samples. You just have to watch note duration and you can slightly shift starting notes back and forth a few ticks to simulate random timing variations like in real strings sections. Also you can utilize slight pitch bends between different string tracks to further separate the lines. Strings don\'t exhibit the same massive shifts in timbre as is present in brass or percussion so you can get away with pitch bends. That is as long as they\'re not too dramatic. http://www.northernsounds.com/ubb/NonCGI/images/icons/smile.gif
I really suggest though you look into something like Ultimate Strings to add to your sample collection.


clueless: i think the way you use both cc7 and cc11 is consistent with the way i\'ve heard how both controllers should be used. I know that with certain synth manufacturers the expression controller is considered the true \'continuous\' controller, meaning that you should use cc7 to set overall levels and use cc11 for envelopes even though both controllers have the moniker CC (continuous controller). I remember with some older sound modules you would hear random \'pop\' or \'crack\' noises if you tried to use cc7 on envelopes. Doesn\'t seem to be the case with newer stuff including GS.

Regards,
F. Juma

[This message has been edited by Gulliver (edited 04-15-2001).]

[Gulliver]

.......

[This message has been edited by Gulliver (edited 04-15-2001).]

[Simon Ravn]

Funny. I never used anything but CC7 for continous volume changes. I gotta try out this expression thing I guess... http://www.northernsounds.com/ubb/NonCGI/images/icons/smile.gif

[JWink]

I also use CC7 to set overall volume (and occasional boosts as needed), and CC11 for crescendi/descrescendi etc. I do own one synth that won\'t let me do this: The Roland M-SE1 (String Ensemble module) uses expression for filter cutoff. I find this highly annoying.