I intended to post this as a reply to a recent topic outlining future releases of Garritan libraries, but I can't find the original post. What follows are my subjective reactions to GPO, JABB and other sample libraries that I own. Basically, this is just a list of what I have found works for me, what doesn't and what I still need. I hope that developers will take this into account in future releases.
- Low pricing is good. I will buy more sample libraries in the long run, if it doesn't cost me a fortune in the short run.
- Copy protection is bad. The more difficult you make it for me to register and use your software, the less likely I am to buy one of your products again. Dongels are the worst. I have had problems with Windows and Windows apps that have to poll USB ports for authorization. I have had nightmare updates to dongels than have gone wrong and effected other parts to the system. I will never use them again. If that's your method of copy protection, you have lost me as a customer forever. NI's authorization process is the second worst approach. For one thing, their site is in Germany, so its s-l-o-w even on my broadband connection through a fast router and cable modem. Secondly, it is frequently down or malfunctioning. Thirdly, NI employees seem to be extremely dense and unresponsive to requests for assistance when their software or web site is malfunctioning. Given a choice between a library that requires activation and one that does not, I will take the latter every time. My feeling is that the vendor or developer (often the same person or company) knows my name, address and often even my credit card number when I make the purchase. He knows (or should know) that I have a legitimate copy. It shouldn't be my responsibility to constantly prove it to him.
- The GPO approach to building small ensembles (created from solo instrument samples) is very helpful for those of us scoring for smaller orchestras. However, GPO does not go far enough. Where are the solo tremelo samples? Three violins playing a tremelo sound vastly different than a full section of 16 or 24 players doing so. And how about some more pizzicatos, too? A better job was done with JABB where, for example, every trumpet and trombone has a sample for each mute in addition to the open instrument (even the bass trombone). I'm not saying leave out the patches of big string sections, just that if you are going to allow small ensembles to be built, you should do it thoroughly.
- Ditto for keyswitched samples. A lot of people love them, want them, need them. Fine. They should have them. But for those like me, who have physical difficulties playing (I have arthritis and wrist injury that prevent me from using both hands on the keyboard most of the time) or are just not strong players, you should provide separate instruments for each articulation. The best example of this is Kirk Hunter's libraries.
- That said, the one thing I don't like about Kirk Hunter's libraries (I have only the solo strings) is that they appear to have been recorded with some natural ambience or reverb. That's bad. I want the sound to be as clear and flat as possible. Reverb and other processing can be added in the final mix, where I will have much more control.
- Getting back to GPO & JABB ... the dreaded mod wheel. I really hate this approach. Firstly, I don't want to have to record or manually enter a CC1 value on every track just to get the thing to make a sound. I don't want to have to move the mod wheel when auditioning samples in the player. It's a nuicense that I don't have to deal with in Kontakt or other virtual instruments. When I buy a new Garritan library, I want to just add the sounds to a project that I am working on. I don't want to have to go back and re-record every single track, riding the mod wheel (and using the sustain pedal differently than I have been using it on the piano for most of my life). Again, I'm not saying take this feature out -- but give those of us who really dislike it an alternative. The simplest way, I think, is just optimizing the samples for Kontakt 2 (or marketing a special edition for Kontakt users).
- Kontakt scripting is better than any library I've seen with the Kontakt player. It seems to me that you can potentially control articulations better with it than forcing musicians to play differently. I want to be able to play the piano as I normally do (or write what I would normally play, if I had complete control of my wrists, in Sonar's score view) and have the sampler/sample player interpret it in a way that is idiomatic for the isntrument. For example, on a keyboard, to play legato, I would let one note continue to sound as I play the next (maybe using the sustain pedal). If I were using a monophonic wind instrument patch (say, the clarinet) when I play like this, I would like the application to substitute legato samples until I let up on the pedal and strike the next note. Ideally, I think what developers should be moving towards are several virtual instruments (one for each of the ways that sound is physically produced: wind, bowed, plucked, or struck) so that natural sounding tracks can be produced independent of the input method (keyboard, guitar, wind controller or electronic drum pad.) The best example of this so far is Real Guitar. It allows me to play a monophonic melody on my MIDI keyboard or (in a different mode) "strum" chords. Unlike other samples which would assign a strummed chord to each key, this virtual instrument redistributes the voicing of the chord I actually play to its nearest equivalent on the guitar -- so that I am not playing inversions that are easy on the piano but impossible to finger on a fretboard.
- Finally, I wish sample makers would stop trying to turn out the upteenth string or percussion library and get to work on some of the samples that no one has done yet. For example, where is the equivalent of Bela D's "Diva" for male voices? I mean, come on, guys ... there are only two sexes. What's the problem here? Bel D recently released a collection of children's choirs, but it's still all treble. No men's voices. And every vocal sample library seems to be a classical sounding choir (most very thick legato pads with slow attacks). What about jazz, pop, etc? You'd think there were no other kind of music. Although some libraries contain sampled words or phrases, these are not instruments in the way that string, woodwind, etc. sampled instruments are -- where I can simply assign them to the melody lines that I have recorded. And forget about typing in words -- a la EWSC -- the technology simply isn't there yet. It takes so much work and it still never sounds natural. All the time wasted on this could be better applied to making really nice lead vocal samples (solos of multiple singers, so that they could be combined into duets, trios and smaller choirs) to serve as guides and placeholders for live recorded vocals.
There are lots more examples of developer's short sightedness. For example, there are relatively few accoustic bass samples in bass sample libraries. And there are relatively few arco bass samples in accoustic bass libraries. There are people who make their careers playing this instrument -- and a wide variety of sounds that can be gotten from it. Also, as a lot of contemporary arranging manuals encourage you to not write a bass line -- just let the player improvise from one of the bass clefs or chord notations -- it would be helpful to have an application that generates basic, alterable, patterns in the manner of Real Guitar and Steinberg's virtual bassplayer -- minus the Steinberg dongel, of course -- for when I am writing something contemporary that the bass player is just going to change anyway. Something that just gives me some idea of how it will sound.
Quite the lengthy post there, but I cannot disagree with anything you mentioned. I think with GPOA, Most of your concerns are addressed, since its all going to be made K2-compatible, but I'm unsure of if the CC1/CC64 playing changes will be made.
I guess if you really wanted to, and knew a certain amount about the K2 Script, you could take GPO and create your own scripts to do what you say, but i think this is best left up to gary & team, don't you?
P.S., about the typing and choirs, even with Garritan Choirs (TBA), if/should it inculde a word feature, we will still need to type out what we want the orchestra to say, but I totally agree, EWQLSC is a PITA and, IMO, went the wrong way with trying to realize their dream. Perhaps Sonic Morphing will be the Garritan answer to this decade long problem?