Vienna Symphonic Library Forum
Forum Statistics

180,768 users have contributed to 42,140 threads and 254,362 posts.

In the past 24 hours, we have 3 new thread(s), 15 new post(s) and 57 new user(s).

  • More Dimensions Pre-Sale Questions: FAQ?

    orry, sorry, sorry. I'm sure this is posted somewhere among the dozens of posts, but I'm hoping to get a concise ruling from the refs. I think D/S is such a different product that it should have it's own FAQ.

    1. What are the -real world- requirements for D/S? I keep seeing anecdotes that would indicate something like 24gb -and- an SSD are necessary for a complete 'orchestra'... while the 'minimum' requirements are '4gb' RAM. I can't afford D/S -and- a new PC right now so I don't wanna get it and then find out I can't actually -use- it with what I have now... or at least I wanna know what I need to buy in order to be productive.

    2. I see in the video that one can load patches for individual 'players' 1-8. But elsewhere the talk is of 2 players per 'desk'.So, can one isolate individual 'soloists' or is this a matter of 'balancing' the level of each of the 8 players... or determining auto-divisi?

    3. On a related topic, I currently use Chamber Strings. If I want a -smaller- sound, can I somehow 'mute' some of the players?

    4. I assume that a big part of the large RAM requirement is for loading separate patches for each 'player' or desk. Is that the basic idea?

    5. Are there 'RAM saving' lower-horsepower patches for when one -doesn't- need divisis or lots of detail... just all 8 guys unisono on the same articulation? Perhaps with fewer layers?

    6. Since there are no mute... er 'con sordino's... is this to be expected in a future update or will it be a separate 'extended' ($$$) lib?

    As always: THANKS!

    ---JC


  • last edited
    last edited

     Hi Suntower,

    @Another User said:

    6. Since there are no mute... er 'con sordino's... is this to be expected in a future update or will it be a separate 'extended' ($$) lib?

     First things first, we have to finalise the other DImension sections, cellos, violas, basses. So it's way to early to talk about a sordino release at the moment.

    best

    Herb


  • last edited
    last edited

    @herb said:

    You can easely compare with any other instruments, performing 8 Dimension Violins needs same CPU ressources as you would perform with 8  instruments you already have. For example stack all instruments of Chamber strings I and II and perform them together, than you will see how your system is managing the CPU ressources.

    Thanks for taking the time to reply. Two follow-ups.

    1. Let me see if I have this straight: 8 Dimension violins playing unison on one instance of VI requires -approximately- the same RAM/CPU as running -8- VI instruments of the same articulation of Chamber String Violins.

    So, if you have an ensemble in Dimension Strings that is similar to Chamber Strings... 6  violins, 4 viola, 3 celli, 2 cb... In simplest terms, that's RAM/CPU for 4 instances of VI. The Dimension Strings version would require the equivalent RAM/CPU of -15- instances of virtual instruments, right?

    2. There seems to be a LOT of unwritten emphasis on using SSDs and -not- 'streaming' samples from HD so much. How realistic will it be to stream Dimension from hd. It seems as though you're -strongly- hinting that a 7200RPM drive will not be adequate. So I guess I'm back to my original question: what -will- be a reasonable configuration for, say a Dimension 'chamber strings'... 8 firsts, 8 seconds, 6 vla, 6 vc, 4cb? Can you predict whether or not my current machine (i7 Sandybridge, 16gb RAM, 7200RPM HD, no SSD for samples).

    Suggestion: perhaps a chart with 'instance counts' for various instruments on a base-line machine as Universal Audio does with their UAD-2 cards?

    Thanks again,

    ---JC


  • to 1) I would not expect that. it should be the same in terms of CPU but RAM is totally different. If you have the staccato articulation of chamber string loaded into 8 VIs it gets effectively only loaded once. if you do that for DS, as no samples are shared or the same between the eight different players I guess, it will load 8 sample sets, msking it use 8 times the RAM of one player. not so with chamber strings with one articulation loaded into 8 VIs.

    besr


  • Herb/Paul/et al... Can I get a response to my 2 follow ups when you have a chance?

    Thanks,

    ---JC


  • Hi Suntower, 

    It´s really hard to make predictions and promises. It depends on your arrangement, the number of voices you will be streaming, if you are using Velocity Xfade or not....

    If you expect a great performance, SSD´s are the way to go. 

    Best, 

    Paul


    Paul Kopf Product Manager VSL
  • Hi Paul,

    I often feel like a guy in a Rolls Royce dealership asking how much for the optional moon roof. (My dear fellow, if you have to ask, you probably don't belong here!)

    I'm not asking for -promises-. I do think it's reasonable to publish -some- guidance/case studies... as UA does for it's UAD-2 cards. They're not -guarantees- but at least you get some general way to predict so one can plan purchases.

    I think it's reasonable to ask that you publish -some- sort of table. For example: take the current demos up @ the Ilio site. Just publish the machine specs used for each demo, with some idea of 'max voices' or '# of instances of each VI'. -Something-. That shouldn't take ages to come up with.

    Steinberg argued against this for -years-, saying there were 'too many variables'. But when they finally started publishing some guidelines they were -applauded- for doing so. They were scared of 'over-promising' or getting into being 'techies'. Users (especially VSL users) aren't that petty, I'd wager. I think most people just want some realistic expectations. Most of us aren't silly enough to complain if their machine can't render Mahler's 8th on a Celeron.

    I know one thing for -sure-... in-house, you guys -do- know what it takes to render various pieces of standard repertoire with VI, MIR and the requisite libs. I hope you'll consider making your results public.

    Best,

    ---JC


  • last edited
    last edited

    @Suntower said:

    I think it's reasonable to ask that you publish -some- sort of table. For example: take the current demos up @ the Ilio site. Just publish the machine specs used for each demo, with some idea of 'max voices' or '# of instances of each VI'. -Something-. That shouldn't take ages to come up with.

    Agreed! It would be such a help if right beside every audio demo VSL would list the hardware specs, including number of computers used, additional software used, and a .pdf of the score we could download to follow along with. We really need this information to make sure that we buy the right hardware and have realistic expectatoins of what our existing computers can do.

    I'm really excited about the Dimension instruments and MIR and plan on building a few very powerful PCs, with the goal of being able to have a full orchestra playing all sections in real time with no glitches or slowdowns. In order to do this I'm going to need very specific details as to what the optimal hardware specs would be, and I'd need to hear some really complex demo scores being played effortlessly on that hardware before I'd be willing to buy the Dimension libraries.

    Right now I have most of the SE that I'm running on a MacBookPro, so I don't have a high end workstation to demo the new instruments and MIR on in order to get an idea of how powerfull a system I'll need. That's why listing the specs of the systems used to make the demos is so important.


  • last edited
    last edited

    @Suntower said:

    I think it's reasonable to ask that you publish -some- sort of table. For example: take the current demos up @ the Ilio site. Just publish the machine specs used for each demo, with some idea of 'max voices' or '# of instances of each VI'. -Something-. That shouldn't take ages to come up with.

     Just my viewpoint, not intended to dilute your question in any way.

    This is not an exact science, although intuitively you'd think otherwise. The fact that on the one hand, demos are mostly done by non-VSL third-parties, and on the other that each composer could have a particular way to program instrument matrixes, to perform/use these virtual instruments (articulations used, x-fades, releases, use of RAM or disk streaming, etc), to use other mixing resources or not (MIR, VEP, addtl' reverb, EQ, Comp, etc...). Asking for details associated to a specific performance will not necessarily have any bearing on how you may use the very same library at your end. So I can see how a certain representation for a given performance may not accurately represent the way the end-user could attempt to reproduce such performance. Technically.

    I think the suggestion to, as an end-user, utilize how you would perform 8 String Chamber instruments simultaneously as a reasonable representation of how you would expect to use DS, makes reasonable sense. I'd actually even caution that as the number of articulations in Chamber (that somebody could in fact use in their matrixes) is larger than in DS, so 8 Chamber instruments may be a bit more demanding than 8 DS instruments- RAM-wise, for instance. But at least the one fair assumption in my mind is that your style to program and use matrixes would be particular to you, and similar between the instruments, so better to rely on comparing that to comparing how somebody else would use these libraries.