In listening to my own compositions as well as several that are posted in the Listening Room, I notice that GPO string ensembles tend to have a tubular sound (for lack of a better description), sometimes even an organ or calliope quality about them (particularly as an mp3 file). I'm experimenting with ways to minimize this and have found that if I eq out the following frequencies I get a more natural response:
Cello 421 Hz and 2.71 kHz
Viola 607.6 Hz and 1.17 kHz
Violin 523.6 Hz
Has anyone else discovered a better way to eq them? If you try my settings, please let me know what you think.
You'll have better luck with strings if you build an ensemble, as opposed to just loading a string section. There are tutorials on the Garritan website. I've found that using solo strings doubling a string section (that is at a lower volume level) can give some nice results. It takes a lot of time, but can really improve the realization of your score.
I always use the KS strings. Whenever the strings should play legato quietly, I use the C# keyswitch, when they are supposed to be aggressive, crank up the velocity of the first note in the phrase, and use the C keyswitch along with high degrees of modulation.
This is what I do all the time, then I layer two solo violins. I copy the 1st vlns track, and paste it into the two separate empty midi tracks for the two solo vlns. I humanize the two solo vln midi tracks so that the onset and duration, as well as the velocity are random percentages of the current data.
Use fader automation to allow the solo instruments to be heard in delicate and aggressive phrases, just as you would in real life.
This is of course just what I do, but there are so many different ways to get realistic results.
For violins, I always cut 2-3 db @ 2.5khz with a medium bandwidth setting. I don't EQ the other sections except only in certain phrases.
The EQ of most all of the instruments in GPO are very natural. Straying too far from the natural EQ is dangerous, and if the strings sound poor, it is most likely a serious lack of modwheel emotion and balance between other instruments.
As Dan alludes to, EQ is only part of the story. It really depends on what other instruments you're using (a string arrangement for a rock tune is quite different from a full orchestral mockup) and the effect you're trying to achieve. Just as no one string sound is useful in all occasions, so to are EQ and level settings highly situational. Even setting aside the "every mix is different" reasoning, the type of reverb that you put a string mix through will have a significant impact on the overall perception of light, dark, and balance. So, if we're going to talk about making strings more realistic - you'll need to describe context, more settings in detail (like how much of a dB cut and at what Q factor for each EQ setting) as well as the reverb and intended end result.
That said - I don't think I've ever produced a mix that didn't have EQ settings on each section's group channel (I use the technique of layering in a few solo instruments with sections, too - but bring each section into its own submix) and then further tweaks on high damping in the reverb/convolution area - and sometimes even more EQ, a multi-band compressor and harmonic exciter in some cases (using iZotope's Ozone 3). That said - the amount of processing I'd do for a spec film cue to "Matrix Revolutions" would be quite different than a mockup of Barber's "Adagio for Strings".
RPearl, I laughed outloud at your last post thanks for the confidence
I feel like a goofball. Houston brought to the discussion the point that the strings will sound totally different based on the reverb settings, and I totally agree. I should have mentioned that in my earlier post.
Good point too Houston! This is why multiple different opinions in a topic is a good thing.
I too use groups for all the separate sections, and as Houston said, this is where it is very convenient to make any corrections to the overall section sound.
Just to add another point, it is not advisable to insert a reverb into the individual section group master/s. You will always want a separate "AUX" channel for each section's reverb, and use whatever means to send the group master (in stereo) to the reverb channel. Perhaps a stereo aux send. This aux send will allow you the freedom to automate the group master volume, and not have the reverb (tail) controlled by the same fader. I know this may sound confusing, but you will always want the reverb to breath on it's own,... independantly of the group master/s.