• Register
  • Help
Page 1 of 2 12 LastLast
Results 1 to 10 of 16

Topic: Orchestrators: 3 Questions

  1. #1

    Orchestrators: 3 Questions

    I am relatively new to the sampler business (but I am classical music profi). I am currently trying to produce symphonic works using GSt and Advanced Orchestra. I have three questions:

    1) Velocity/Timbre
    Let\'s say I am recording an oboe. Obviously with the different volume the timbre of the instrument changes as well (in the real world). Now with the AO I have got a choice to take the KEY-XFD version and using a modulation wheel to change the timbre parallel to the the change of velocity. Unfortunately the XFD version always sounds as if two instruments are playing at the same time and it is quite unsatisfying especially in a transparent instrumentation.
    The KEY version does not have this problem, but it is impossible to change the timbre some other way than switching between p and f samples which sound extremely different. How can one use this version to sound realisticaly for instance when playing crescendo? I assume I could assign a filter to the velocity in the Giga Editor but it seems to be a major task. How do you use this version?

    2) To the ensembles.
    Where can one use an ensemble? OK, I could compose a piece and let my 4 horns play the same thing all the time. But, if they don\'t? If they play a 4-part chord? I suddenly get 16 horns sounding.
    Do you swap between ensembles and solo when the instruments play/ don\'t play unisono?
    Oh, I am currently working on Bach\'s Orchestral Suite, there are two oboes playing the same thing thruout the whole piece, er no ensemble oboes in the AO....

    3) Timpani tuning.
    Strange thing with timpani, I need an \'A\' and an \'D\' drums. Unfortunately the the \'D\' sounds like slightly higher \'C sharp\' and \'Dis\' is too high for \'D\'. Anybody the same problem? How do you get \'D\'. I guess I will have to fine tune before each note of the timpani. Hm...

    Thanks for your patience.


  2. #2

    Re: Orchestrators: 3 Questions

    All good questions, and I\'d like to hear others\' answers too. Here are my thoughts (I\'m a classical music person too, also new to samplers).

    1) Timbre/velocity - this is a function of how many velocity layers are recorded in the sample set. The more the better of course, but you might have to pay for this by foregoing other desirable things (like stereo). In any case you aren\'t likely to get a really large number of velocity layers - 16 would be huge, more than 32 impossible (in GigaStudio, at least as I understand it). There are subtler problems, too - for instance, vibrato in the real world isn\'t just a matter of changing pitch; timbre and volume change as well. So a simple LFO or mod wheel isn\'t enough to mimic the real thing. I\'ve seen a little bit of stuff on the web about \"physical modeling\" (I think that\'s the name) that tries to address these issues, but I haven\'t run across any demos. Anybody know about this?

    As to your question, how do you use AO (or other sample sets) given these limitations - well, I don\'t know of any easy solutions. It seems to be a matter of painstakingly tweaking each individual note, until you reach your personal point of diminishing returns.

    I\'d be most interested in hearing your Bach Suite. As a learning exercise, I\'ve just finished a version of Falla\'s \"Ritual Fire Dance\" that I think sounds fantastic, but apparently that work is still under copyright so I don\'t think I can make it publicly available. I\'m trying to contact the publisher to see what I can arrange. I\'ll be doing other GSt projects in the near future, though.

    2) Ensembles - so far, I\'ve recorded the wind and brass parts as solos, recording the same line multiple times if necessary. I would probably do that in the Bach as well, even if the oboes play in unison throughout. Even if you had a good oboe ensemble sample, it wouldn\'t sound the same as two oboes. You definitely wouldn\'t want to use an ensemble sound and vary the number of voices - you\'re right that you\'d go from hearing 4 horns to 16 to 8 etc. It doesn\'t seem likely that you could smoothly switch from an ensemble patch on the single notes to solos for the chords - so again, you\'re stuck with recording the individual instruments as solos. I think....

    3) Yes, many of the AO timpani notes are slightly out of tune. It\'s annoying. You could adjust it with a pitch bend each time, or edit the sample itself.

    Sorry for the length of this post - the subject interests me! And I\'m very curious to know what other folks think about these things.

    [This message has been edited by ursatz (edited 07-21-2000).]

  3. #3

    Re: Orchestrators: 3 Questions

    Hi Ondrej,

    Interesting questions and discussion so far.

    1) This may be way off, but for a crescendo I almost always use the MIDI volume control. The sample then sounds consistent throughout the phrase and the crescendo takes place within each individual note, just as it would with a real player.

    2) I swap between ensembles and solo in just the way you say for horns (for example). For ensemble oboes, I use the oboe from another sound module (Proteus 2) in conjunction with GS/AO oboe.

    3) I set up my score-writing/sequencing program to transpose the timpani up (or down? I can\'t remember) a semitone. It is not great but it is OK. You could fine tune with a pitch bend, I guess, though I have not tried it. Oh, I see that Ursatz has already suggested that.

    My replies seem somewhat curt, but it is really late. Please ask me to clarify anything too dense.

    All the best,

    Richard Cox

    [This message has been edited by rpldcox (edited 07-21-2000).]

  4. #4

    Re: Orchestrators: 3 Questions

    1) Velocity/Timbre
    Let\'s say I am recording an oboe. Obviously with the different volume the timbre of the instrument changes as well (in the real world).

    I don\'t use AO woodwinds but I think you have come across a limitation. I know that some of the Miroslav wind samples naturally brighten up allowing you, with the help of a cutoff filter, to control the timbre but this too has its limitation. I think we can expect the future generation of orchestral sample libraries to better take advantage of the cross-fading feature of gstudio.

    2) To the ensembles.
    Where can one use an ensemble? OK, I could compose a piece and let my 4 horns play the same thing all the time. But, if they don\'t? If they play a 4-part chord? I suddenly get 16 horns sounding.

    I\'ve yet to come across an ensemble Horn sample that actually has the force and presense of 4 real Horn Players. I use both AO and Miroslav (have not tried Quantam Leap) and even with triads, the balance, orchestration wise, is not 12 horns but closer to 4. You can always let your ears guide you because you\'re in the sampling world rather than the real one. Mixing solo horn with ensemble works, that is if you keep the combination consistent. I think the AO solo horns are mono (atleast the akai version) so that is something else to keep in mind when mixing.

    3) Timpani tuning.
    Strange thing with timpani, I need an \'A\' and an \'D\' drums. Unfortunately the the \'D\' sounds like slightly higher \'C sharp\' and \'Dis\' is too high for \'D\'.

    If you load up Gseditor and pulldown the dropmenu just above the region window, you will see along the bottom \'Fine Tuning.\' The nice thing about the new editor is that you can now graphically adjust the fine tuning per sample. Audition and listen as you change the tuning. Once you save you won\'t have to worry about midi pitch bends, controllers, etc.. The instrument will be tuned to your liking.


  5. #5

    Re: Orchestrators: 3 Questions

    Hi guys!
    Thanks for your feedback. It is nice to see that other people face the same problems i.e. I know I am not going mad.

    Ursatz (sounds really German!),
    I would also be interested to hear what you achieved (if this become possible). My Bach, to be precise, is just an excerpt of the Gavotte from the Suite No. 3. It is a part of my degree project (featuring also a Mozart, Berlioz and time+health permitting Mahler and Bartok). I will make some of it available in the future provided I manage to do reasonably well and provided the university will not object to it.
    Interesting to hear you mentioning physical modelling. Although I am new to the sample world I am not new to synthesisers and I actually decided to go samplers after several years of experimenting with synths and physical modelling. I owe Yamaha VL and Technics modules and just to make it short (this is a GigaForum after all) I would say that instruments in Yamaha behave like real thing but do not often sound too real whereas those in Technics sound quite close but it is hard to make them behave like real instruments. If you want further infos on this matter please e-mail me at ondra.r@uumail.de .

    Gulliver, you say you use volume control, do you than also change the timbre or do you feel is it not so obvious to hear when it is the same all the time? Your point of the force of ensemble not matching reality is interesting. I also find that the strings of AO sound rather classical (I mean good for Mozart), do you guys make them bigger somehow for the romantic/modern stuff?

    And just to mention some more things which I experience while working with
    AO on some more pieces:
    I find that:
    - the violin detaches are extremely long
    - generally all the detaches do not fit with the normal bow playing, especially when longer note follows a detache, they are of completely different timbre and frequency
    - the violins and celli have slightly higher pitch than the reast of the orchestra, I have to fine tune them approx. -5 cents

    And one more question for the end: do you use the -R version of strings? That would mean you change programs (from the short release to the long one) always just for the last note of any frase and that you have to release a key a bit before the actual note is to be finished. Right? Because I can\'t imagine any other way to use this version.

    Well, enough for today

  6. #6

    Re: Orchestrators: 3 Questions

    Gulliver, you say you use volume control, do you than also change the timbre or do you feel is it not so obvious to hear when it is the same all the time?

    Well, I can control expression/volume but timbre is something that is unfortunetly not so easy to manipulate. I\'ve mentioned previously that some of the Miroslav woodwind samples change timbre or gradually become \'bright\' after the attack. I use these samples in conjunction with filters to control the effect. Sometimes words are inadequate so I have put up a short simple example:

    The mp3 track features an English Horn. You might notice how the instrument changes timbre. I used both expression envelopes and filters to delay or control the speed of the timbre change. Personally, I don\'t think I would ever put that much detail into a single instrument unless it is exposed as a solo or maybe if it is part of a small chamber ensemble.

    I also find that the strings of AO sound rather classical (I mean good for Mozart), do you guys make them bigger somehow for the romantic/modern stuff?

    I\'ve used AO strings and to be honest I stay away from them. I think the miroslav strings are best suited for classical.
    short examples:

    Miroslav solo strings http://members.home.com/vstorage1/bachbr.mp3

    Miroslav flute + Gigaharp + Miro strings http://members.home.com/vstorage1/mozflute.mp3

    As for 20th cent/romantic music, I\'m not a big fan of chorus effect so I usually stack as many different string samples I cam to get a thicker ensemble. My choice is Ultimate Strings with Miroslav ensemble. Again, why talk to explain when you can hear it?
    short examples:

    Rachmaninov\'s Symphony No.2 - Adagio http://members.home.com/vstorage1/rach2.mp3

    Elgar\'s Engima Variations - Nimrod http://members.home.com/vstorage1/nimrod.mp3

    Shostakovich\'s Symphony No.5 - IV http://members.home.com/vstorage1/shost.mp3


  7. #7

    Re: Orchestrators: 3 Questions


    maybe you are interested in this. We have specialized on good quality natural instruments. That means also good sound programming. On more than 1000 presets we tried to give the users many possibilities to form the >users< timbre. Most of the presets are using more than 4 controllers. The controller make very light differences up to very heavy stuff.
    You will find >crescendo controller<, >phrasing controller<, >tuning controller<, >vibrato frequence controller<, >release time controller<, >chorus controller<, >filter type controller<,...
    Our Alt Sax has also a special breath controller, that make it possible to controll the portion of air in the tone.

    For interest have a look at http://www.xsample.de or http://www.bestservice.de

    Georg - Xsample

  8. #8

    Re: Orchestrators: 3 Questions

    Here\'s a couple tips that may be useful.
    The first one, which I suppose may be obvious but which I just figured out recently, is to mix the loudest section of your music first. Of course this assumes that you have already written out the loudest portion. The reason I\'ve started doing this is because I\'ve found that sometimes I run out of headroom for loud sections, and there\'s nothing worse than a long buildup that just whimps out. You can always make things quieter by setting volume/expression close to 0, but you can\'t go the other way, since the limit is 127.

    I use AO and XSample libraries and for AO I created an excel table that lists the controller values I need for various dynamics I want to achieve. I have a page for each dynamic, from fff to ppp. The first column of the table lists the instrument, arranged by section. Then the columns to the right list the various controller values such as velocity, mod, volume, and expression.
    I use Cakewalk, and to enter these values into it I created some studioware panels. For example, I created a panel that groups all the mods for the AO String XFDs, so I can move one fader and change the timbre of the entire string section. I can always go back into the Studioware panel and ungroup the violins for example if I wanted to adjust that by itself. Right next to this group of faders is a group for expression. It\'s grouped so that the levels of each instrument in the section stays roughly even. So once again, raising this fader raises the level for the entire string section. I use volume for overall track volume.
    So what I do is I place markers wherever a different dynamic level is needed, consult the pages of dynamic levels, find the instrument, find the value I need to adjust the fader to, take a \"snapshot\" of this, and then go into the piano roll and draw a line up to this value for each instrument. Cakewalk now has multiple instruments on the piano roll so this is just one window. Or you can record your own fader movements, but I\'m not that good with a mouse, so that might actually take longer and not be as smooth. If you have a midi controller like the control freak that\'s much better.
    Anyway this system of mixing is certainly more efficient than trying to discover what to set the controller values to every single time you mix a new piece. Why waste time doing the same thing several times?
    And mixing instruments by group is WAY faster than one track at a time. Of course you may need to adjust certain tracks a bit, but at least you get very close to what you want, and fast.
    I don\'t always use the exact values in the excel tables because some of them need adjustment, but after experiencing how much faster I\'m working now, I don\'t think I\'ll ever mix without it.

  9. #9

    Re: Orchestrators: 3 Questions

    Hi again!
    So much input!
    Thanks a lot. Very useful tips indeed.

    Gulliver, great work! I particularly liked the Conan and Bach. I could not load Rachmaninov but judging from other examples it must be very good as well.

    Georg, sounds interesting. Are these controllers also documented in the CD booklet?

    Tony, very good idea. I came (with the time) to very similar way of working and now I made similar chart to yours. I find that for the AO, with non-velocity instruments like strings, winds or brass it\'s best to ignore the velocity, play everything in with velocity 127 (or transform afterwards) and to operate the level and timbre just using the controllers.

    And so that we have some new thoughts I put up a new thread regarding concert pianos.

    Keep smiling


  10. #10

    Re: Orchestrators: 3 Questions

    Hello Ondrej,

    we made a separate book for all. But you can reach also all information directly with help files.


Go Back to forum

Tags for this Thread


Posting Permissions

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