Vienna Symphonic Library Forum
Forum Statistics

196,686 users have contributed to 43,023 threads and 258,420 posts.

In the past 24 hours, we have 2 new thread(s), 4 new post(s) and 94 new user(s).

  • No Loops?

    I just received a copy of the Demo CD and I'm quite surprised to find that none of the samples are looped.

    Is this the policy for the entire VSL?

    I understand that there are some long samples of up to 20 seconds, but this does not always provide the kind of usability that I have come to expect (and be used to) with samples.

    Of course this wouldn't apply to staccato/pizzicato and similar short styles of playing, but I'd expect it with held notes for strings, brass, woodwinds and anything that sustains.

    Regards - Colin

  • I'm bumping this up to the top again to see if I get a response.......

  • I think that loops are extremely undesirable in a high quality sample library, because first of all they are artificial - real instruments are not looped. Secondly, you can do your own loops if you want them. If VSL included looped samples it would be a waste of data.

  • William, so you're saying that a string ensemble can't play a note longer than one bowing will last? Not true, they can change direction at different times and thus disguise the bowing so they can sustain 'forever'. Same with most other instruments. Therefore loops would be very realistic. It is not crucial to me, but I certainly wouldnt mind it, especially on strings.

  • William,

    I disagree strongly on these points:

    1 - Loops are *extremely* desirable in the already artificial environment of sequencers and samplers. Nothing is more frustrating than having a sample run out just before you want it to end.

    2 - At the price I'll be paying for VSL, I'll be damned if I'm going to sit for hours creating thousands of loops.

    3 - Loops can hardly be a waste of data. It's just a matter of defining start and end points in the samples that already exist. Probably a few extra k's of data but certainly hours and hours of work.

    I'm really keen to know what VSL's position is on this. I strongly believe that having recorded all these instruments is only the beginning in a vast process of creating a very user-friendly tool. Adding functions like loops , Modwheel switches, Aftertouch parameters (for vibrato etc) should be part and parcel of this whole grand vision.

    Regards - Colin

  • Concerning loops, I appreciate the response from obviously knowledgeable musicians, HOWEVER ---

    Musos,
    In previous sample libraries, looped samples are in fact a waste of data, because the sample has to be re-recorded or re-mastered with the loop as a separate presentation of the original sample. Perhaps that is not necessary with the VSL, but both Miraslav Vitous and Siedlaczek, the previous high standards in sampling, all had separate looped samples that wasted portions of the discs.

    Simon,
    Of course string players within sections can change the direction of their bows to create a sustaining sound that does not obviously end. (Though with a smaller section, as in a chamber orchestra, it is audible when they are doing so.) Any conductor or orchestral player knows that.

    However, I repeat, that is not the same as a looped sample.

    I prefer the purest possible recording of the original instruments, with absolutely the least possible artificial manipulation. And someone deciding upon a loop point is manipulation.

    Concerning the amount of work involved, if you've tried to get actual orchestras to play your music, that's a bit of work too. Years of work in my case! You are getting an incredible amount of timesaving with this technology. Some other people on this site were complaining about actually having to play each line of their compositions. I thought that was the whole purpose of this library!

    With respect and my most sincere apologies for being contrary,

    William

  • Hi William,

    I appreciate your response regarding the loops. I have an old Akai version of the Vitous library and none of the samples are looped. It's a pain in the a**e and I never use it.

    In other sample libraries, there are no new samples, but rather a loop section is defined within the sample itself and when the key is held, there is a very natural-sounding sustain. When the key is released, the sound ends according to the parameters defined for that particular patch.

    In the above case, there is really minimal wasted data. I was simply hoping that VSL would do the donkey work for me, namely finding and defining loop points within the already existing samples. They're there - though sometimes hard to find!

    And by the way, no need to ever apologise - we're all entitled to our opinions, whatever they may be. I love these forums because one gets to interact with the software creators, other users who have similar or differing views and invariably, one learns a lot.

    Still waiting for VSL to respond to this thread........ [*-)]: [*-)]: [*-)]:

    Regards - Colin

  • Hello guys,

    sorry for the delay, Paul asked me to answer this thread, but I was too busy till now...

    First: To say it clear, in the first edition are no loops.
    Our philosophy is, we won't make loops, if they are noticeable.
    That means loops for solobrass or woodwind instruments are not planned with common technologies. Maybe there will be other solutions in the near future.

    To loop ensemble instruments, trills or tremolos should not be a problem, but we decided to wait for the upcoming features of the next sampler generation.
    The looping options in Kontakt (the possibilty to generate multiple loops) are steps in the right direction. So we wait, to see the looping options of Giga 3.0.

    We are managing our resources very carefully, and we don't want to waste time for things which have to be redone a few months later.

    Best wishes
    Herb Tucmandl

  • I'm probably droning on about these topics because I don't yet have the VSL in my hands. Once I do, I will definitely shut up.

    I did actually use loops previously, but simply wanted to make them myself rather than have them forced on me. I did in the thousands for the Siedlaczek and Vitous libraries on the EMU IV. Also, the entire Vitous library for EMU was in fact equipped with extra looped versions of the same sample recordings.

    But on my own loops, I tried to make them almost as long as the original sample itself, in some cases eight or nine seconds, rather than do them the way programmers usually do, which is to find a very short, uniform section of waveform and use it. That is the quicker and easier way, but results in a lifeless sound once the loop point is reached. Of course it is a problem making very long loops if there is a noticeable difference in the performance, either timbral (which happens a lot with woodwinds and brass as they tend even on relatively low and uniform dynamic levels to play with more high frequency overtones at the start of a note) or in volume level. But by making them extremely long, it allows in normal orchestral practice the loop to be almost never heard, or the original volume variations, naturally occuring, to still be heard and not be noticed in context.

    Anyway I realize that the number of loops on this library would be gigantic, however many of these notes don't really have to be looped anyway, given short notes and the fact that the long note samples extend beyond ten seconds. Maybe for new age music all this would be a problem!!!??? Also, if the VSL ever samples a bagpipe, we should definitely insist upon loops.

    Thanks,
    William

  • Herb,

    Thanks for the official response.
    I appreciate your standpoint and obviously I'll wait to see what the first release is really like before I comment any further.

    I'm pleased that VSL hasn't disregarded the idea of loops completely.

    Regards - Colin

  • Hi
    Simon is absolutly right about the sustained strings with subtle changes
    of direction - but I think that instead of looping why not make for a better
    solution: Give us the subtle direction changes , just as in the legato
    perfomances. In a perfect world this would be like one sample for the
    initial sustained note, one sustained including a direction change of the bow to upbow, one sustained including change to downbow direction and
    then a release sample. This way we can alternate forever between up- and down bow and thus sustain forever.

    Bjk

  • String sections stagger their bowing don't they?

    Anyway i don't like that the ensemble instruments aren't looped. The string ensembles should definitely be looped and all this does is force us to loop the samples individually as needed. Hopefully the samples will be long enough that it won't matter though.

  • Hi Peter,

    all sustain strings are longer than 20 seconds.

    best wishes
    Herb

  • Whoa! I certainly wasn't expecting that. In that case maybe the loops won't be necessary at all! Thanks.

  • 20 secs isn't bad at all, but then you're stuck with always using the sustain samples for long notes. Mostly okay, I imagine, but not ideal.

    Maybe the members on this forum will do a Loop-fest one day and post all the data (start and end points etc) in the Files area.

    There IS a Files area isn't there......?

    Regards - Colin

  • hm sorry, but what exactly do you mean with files area.
    another suggestion, why not invent an self learning "auto looping tool"
    -- Peter?

    cheers herb

  • I've been toying with doing loops for the first edition of VSL. Atleast giving loop points. Two fold. Maybe it can help in the pro edition, and it will help VSL users make their own.

    Obviously Xfades will be essential for some, but overall looping is pretty easy with the right tools.

    In most cases you wont need the loops, unless you are the kind of guy who holds a triad/fifths on the violins up high, while playing a melody on the lower range of the violins all at once in realtime.

    Other than that, its no biggie with the long sustains.

    I think one thing to remember is that the guys at VSL have a "plan" and are really considering all aspects. Looping is, and should be IMO, secondary to sampling the rest of the orchestra.

    Herb is right in considering the different looping aspects of future samplers, before compromising the samples themselves. Imagine art file updates for the next version of giga, that takess advantage of new looping features. If they were to give us, already Xfaded samples, they'd be giving us compromised Wave files that may not benefit from features that could be added via art files (multiple loop points, multiple release points from the same file, etc.)

    Not to mention the amount of time it'd take for them to loop things.

    While I say its an easy thing to do withthe right tools. Its not something that can be done very quickly over hundereds of thousands of wave files [:)]

  • "why not invent an self learning "auto looping tool-- Peter? "

    Hm, not sure why you think I'd be able to do that, but theoretically it should be possible.

  • I'd like to express my appreciation to Herb for sharing his expertise in this forum and letting people know what is going on.

    I would add that what king idiot said (sorry, but that's apparently your name) is what I stated previously, that having loop points crammed down your throat is extremely undesirable. I don't care who does them or what clever little system you use - I don't want them. On the other hand, the raw, 24 bit recorded sound of Viennese string players is priceless. And with twenty second sustains ---- there is no longer any point to this discussion. (Except for adding that turning shorter samples into long sustains is a musical grotesquerie that I want no part of.)

  • Herb,

    Usergroups on Yahoo usually have a so-called Files Area where members can upload their files to share with other users. In the case of softsynths like FM7 and Absynth, this files area would mainly consist of user-made patches to be shared with others.

    If it becomes appropriate, maybe VSL could add this.

    Regards - Colin