• Register
  • Help
Page 1 of 3 123 LastLast
Results 1 to 10 of 26

Topic: The Garritan Library Poll

Share/Bookmark
  1. #1
    Senior Member fastlane's Avatar
    Join Date
    Nov 2004
    Location
    Shelton, Washington State
    Posts
    3,024

    The Garritan Library Poll

    There is a library poll at the Garritan site. I see that a choir library is way out ahead in votes.

    I would like to see a Garritan choir library that would be affordable like the other Garritan libraries but I'm not sure what that poll means now.

    There are several "professional" choir libraries on the market but I can't justify the cost.





    Phil

  2. #2

    Re: The Garritan Library Poll

    My thoughts exactly. This is YEARS overdue. Garritan waited so long to release the organ that I ended up going with one released by another publisher and couldn't justify the cost of getting the Garritan version when it finally came out. I know their reasons, but that doesn't help me finish my project.

    It always mystifies me why companies bother to do market research and then ignore the results (especially when consumer feedback is provided free.) It doesn't help anybody and discourages your customers from giving feedback again (which means there will be fewer responses, and therefore a less reliable feedback, in the future.)

    As for the existing choir libraries, I have to say that I am still not impressed with what you get for the price. None have the type of versitility that would make them genuninely useful to me. For example, like most virtual instruments, they seem to limit you to either massive choirs or solo voices, where as most of the work I do is solos + smaller groupings. And I am afraid all of them make it too easy for the voices to sound like a robotic answering machine from the 1970s. Not very helpful, unless you want it to sound like Stephen Hawking's greatest hits.

    Allegro Data Solutions

  3. #3
    Senior Member
    Join Date
    Oct 2009
    Location
    Canyon, Texas, USA
    Posts
    135

    Re: The Garritan Library Poll

    Quote Originally Posted by fastlane View Post
    There is a library poll at the Garritan site. I see that a choir library is way out ahead in votes.

    I would like to see a Garritan choir library that would be affordable like the other Garritan libraries but I'm not sure what that poll means now.

    There are several "professional" choir libraries on the market but I can't justify the cost.

    Phil
    I would quickly buy an affordable choir library if it contained tutti, 1st, 2nd, and solo patches for sopranos, altos, tenors, and basses in all of the pure vowels. Each needs an aggressive, normal, and delicate attack version. Then if each can handle cresc, dim, legato, etc. as well as the Garritan woodwinds, I would be quite happy. There are many other things that would be nice, like humming sounds, hand claps, foot stomps, etc., but, these are not vital.

    Norman

  4. #4

    Re: The Garritan Library Poll

    I am sure Garritan is NOT ignoring our wishes for the choir...but programming the voices to be able to sing words is very complicated, esp if you want to avoid the robot syndrome, or be as complicated as EWQL.
    gpo, jabb, symphonic choirs (only until Garritan's is out!), cubase, sibelius, reason

    Nord Stage 2 88, Fender guitar and bass, Roland td9 drumkit, Norman acoustic guitar

    three cats

  5. #5

    Re: The Garritan Library Poll

    Quote Originally Posted by sakari.lindhen View Post
    I am sure Garritan is NOT ignoring our wishes for the choir...but programming the voices to be able to sing words is very complicated, esp if you want to avoid the robot syndrome, or be as complicated as EWQL.
    You're absolutely right, Sakari, and you've said in your post what I wanted to add to this discussion this morning.

    You and pretty much all of us sympathize with ejr's sentiments:

    Quote Originally Posted by ejr View Post
    ...As for the existing choir libraries, I have to say that I am still not impressed with what you get for the price...I am afraid all of them make it too easy for the voices to sound like a robotic answering machine...
    Despite all the technology we have now, it seems that the ability to create a useful, easy-to-use, good sounding choir is still just not possible. The human voice is probably the most complex, nuanced musical instrument. It seems logical that harnessing all those variables in software would be so difficult as to be science fiction. We're all aware of the limitations in even the most advanced, expensive string libraries, brass libraries et al - No matter how great the libraries are, we have all learned by now that it's simply impossible to completely duplicate the incredibly complex sound of an orchestra. And I believe the same is true of a choir program.

    A couple of years ago, Gary Garritan told me a bit about the challenges of developing a software choir. Recording and assembling good samples is straight forward enough, even though that alone is a huge, time-consuming and very expensive project. It's the word-building part which, not surprisingly, is the stumbling block.

    Some kind of morphing technology has to be used, and theoretically that's what you use to put together a choir library that can sing any lyrics the user wants. EWQL's complicated choir program uses morphing, and any such library would have to. One audio signal is used to shape the sound of another one. I've experimented on my own with the process, using tools available online. I included a modest amount of a morphed choir in my 2010 Garritan Christmas CD contribution, "O Little Town." I used the choir in GPO with its sound being modified by an audio track of me speaking the words "O litte town." Here's info from online from "Waves" about the process, explaining how it's actually classic Vocoding which was developed way back in the 1930's:

    "...The vocoding engine allows two audio signals to be combined using one source as a modulator and another source as the carrier. For example, using voice as the modulator input and a synthesizer as the carrier, the result would be that the synthesizer will “talk” and actually say the words spoken by the voice. In the same way you can use a drum track as your modulator and the result will be a rhythmic keyboard track in perfect sync with the drum tracks events..."

    And so the Waves manual explains:

    "...If you insert this plug-in over any vocal track, all you have to do is press play and play the morphoder’s musical keyboard and it talks..."


    The theory of how to get a sampled choir to sing is clear. But software vocoders/morphers are not created equal, some better than others. The sticky point is what the results sound like when the two signals are combined. The "robotic," slightly metallic sound being complained about in this thread is an artifact which so far seems impossible to get rid of. Garritan, or any company wanting to develop a singing choir would need to first developer their own vocoder, probably based on another company's existing algorithms, and then improve it as much as possible.

    If you've followed this post, you can see what another huge challenge is - I've talked about using an audio track as the modulator, but the user of a word-forming choir library isn't going to be expected to set up a mike, record a clean track to work from, and then do all the tweaking needed for the best results - People can already do that on their own, and the results are maybe adequate, probably not as good as what EWQL has. When I used the choir in my Christmas project, I justified the odd, artificial sound as being part of the dreamy atmosphere of my arrangement. But I knew then, and I know now, that I could never expect that kind of choir to stand on its own, to be offered as a supposedly natural sounding choir.

    Putting together a folder of audio clips to be used as modulators, and then programming an interface that makes using those samples practical - that's a mind-boggling challenge. Obviously the basic set of audio clip modulators would be the vowels, then you would need all the consonants. And having those two basic sets wouldn't mean they'd automatically work together seamlessly. I can't imagine how many samples would be needed, how complex the programming would need to be, to make smoothly operating modulators. Obviously the better, easier, more reliable way of developing a language track would be simply to record it - But, you can't sell that. "Here's your choir samples - OK, now you need a really good microphone, you need to record really clean tracks, you need to know how to mix everything together precisely in your recording program"--- And yet, for a user to work that way would obviously be the more reliable method, rather than cobbling together little bits of sound and expecting them to sound natural. We've all heard the voices on computers and automated phone calls - Those pieced together messages sound better than they used to - So maybe that's encouraging at least.

    The two sample sets - detailed choir and solos, and the spoken word/vowels/syllables set - PLUS, and it's still probably the biggest problem - the vocoder which can blend them together in a convincing way. Those are the challenges.

    All good in theory - but, EWQL's choir is as good as it gets, and is likely to be as good as it gets for a long time to come - maybe forever.

    So I can easily hypothesize that the Garritan team started tackling the project, but came up against all these difficulties. It seems very possible to me they couldn't come up with anything better than what already exists. It could be a case of the old adage "Where there's a will, there's a way" not being true.

    Randy

  6. #6
    Senior Member Tom_Davis's Avatar
    Join Date
    Aug 2005
    Location
    Ellendale, ND
    Posts
    1,486

    Re: The Garritan Library Poll

    Perhaps a program where the words would be spoken into a microphone and recorded to a track then tones assigned to individual syllables.

  7. #7

    Re: The Garritan Library Poll

    I'm disappointed by how far in 2nd place a good advanced string library is on that poll.

    a choral library will probably not be realistic for a great many years.

    so why not work on something, and release something, that is realizable and in demand: an advanced string library.

  8. #8

    Re: The Garritan Library Poll

    Quote Originally Posted by Tom_Davis View Post
    Perhaps a program where the words would be spoken into a microphone...
    Right, Tom - In that text-dense post I put up earlier, I said that what people can do now is record a spoken track to be the modulator on choir samples, given they have a vocoder/morphing plug-in to use. But a choir library couldn't be released like that - as a do-it-yourself tool kit. The library of syllables/vowels/consonants etc would have to be part of the software.

    Randy

  9. #9
    Senior Member tedvanya's Avatar
    Join Date
    Jun 2009
    Location
    Canada,winter Mexico
    Posts
    1,005

    Re: The Garritan Library Poll

    For me it is impossible to even think to buy a choir which could be programed to announce words, and what I hear, they are not that great anyhow.
    As we understand that we could never equate human playing in any instrument, including of course human voices, we have to come to some acceptable compromise which allows us to imagine how it would sound what we create in a real performance.
    The choir sample came with GPO is so bad, it should not been in the library.
    In the GVI factory samples, Bestservice has a sample of a classical choir, Ahs and Ohs only, which is, to my taste, sounds good enough to enjoy. The Papelmedia choirs, in Giga-format, are also good. Of course Aria will not play them, but I am happily using G-player by Soundlib, which plays all Giga samples.
    I have used the Bestservice sample, breaking them in Sonar X1 into SATB, adjusting for attack and release, and the results are good (IMHO).
    I do not know, if it would be unlawful or not, but if it is not punishable by death and torture, I would be happy to share them with who ever is interested. (Please advise). G-player is working fine for me, they just came up with some nice additions and the player is available at a very reasonable cost.(As I said in an earlier post, I am not associated with Soundlib in any other way than being a happy customer.)
    An other poor sample in GPO is the timpani,(horrible), SAM timpani (Giga) is much better, I have a free sample of it also if anyone is interested.

    Ted

  10. #10
    Moderator
    Join Date
    Jun 2000
    Location
    Chandler, Arizona
    Posts
    4,045

    Re: The Garritan Library Poll

    I've always liked the sound of the GPO timpani considering it is just one velocity layer. There are quite a few samples out there that are much worse, trust me!

    Jim

Go Back to forum

Tags for this Thread

Bookmarks

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •