OK, can someone explain to me now how Vienna VI sees into the future?

[Ouch that hurts]

I asked this question a few times when the Vienna Instruments first came out, and never really got an answer, probably because not many people had used them at the time. I'm wondering now if more people might be in a position to comment:

My question regards the facility whereby the player adjusts to the speed of your playing an automatically uses shorter samples for shorter notes. This seems to me to be an absolute physical impossibility.

A virtual instrument, or anything/anyone else without extra-terrestrial psychic powers, can only know how long a note is after the note has finished. And after the note has finished, it's too late to decide what sample to use when that note starts!

I've been trying to get my head around this but I still don't get it.

As far as I can see, the best the player could do is guestimate how short the next note is going to be on the basis of the previous note(s). If I'm right here, how useful is this really in practise? It seems to me it could only really help when the music follows very limited and predictable patterns, and most orchestral music doesn't.

Take for example the following rhythm:

Quarter > Quarter > eighth eighth > sixteenth sixteenth > thirty-second thirty-second > whole note pause.....

Presumeably the player can sense that you're playing gradually faster note values, and trigger shorter samples accordingly. All well and good, but when you get to the whole note pause, won't it think you're still getting faster, and play a really short sample, cutting off your whole note way too short? Or does it have some magic way of avoiding this?

I can see various scenarios like this that would make the system unworkable. So I suppose I basically have two questions:

1. Am I right that the software can't actually make intelligent predictions about the length of the NEXT note, or am I missing something?

2. If I am right, how does this pan out in practise? Do you need to go through afterwards and tweak a lot of the deafault note-values that it has chosen? How easy or difficult is this? Can it read your MIDI file ahead of time on playback like Synful (that would be cool)?

Thanks for any insight.

[jovan]

I can see various scenarios like this that would make the system unworkable. So I suppose I basically have two questions:

1. Am I right that the software can't actually make intelligent predictions about the length of the NEXT note, or am I missing something?

2. If I am right, how does this pan out in practise? Do you need to go through afterwards and tweak a lot of the deafault note-values that it has chosen? How easy or difficult is this? Can it read your MIDI file ahead of time on playback like Synful (that would be cool)?


ad 1) Using a keyboard, VI is allways one note behind the reallity. Using a sequencer or notation program you can VI force to play allways the actual note length.
ad 2) VI cannot look ahead!

I am using VI now more than a month. And I must say the problems you see have not happened.
VI is so much easier to handle than another sampler (K2, GS3 etc).

:-) Jovan

[FredProgGH]

Well, I think Satan is involved here somewhere and I'm getting to the bottom of it.


















:D

[SyQuEsT]

10 minutes with this VSTi (specially if you look the tutorials) and you're on ! it's easy, very simple BUT greaaaaaat :)

[Marcussen]

Hi Ouch.

Some of my previous comments on this feature can be found in this early review here at Northernsounds:


"Then there is speed control. While a nice thing in theory, I doubt I'll be using it except maybe for changing between fast and slow legato. So those who are hoping that this will be the solution to end all might be disapointed." (http://www.northernsounds.com/forum/showthread.php?t=42898)

And some comments I made elsewhere


It will do that just fine, but it might not do it as you want it to, and thats the problem.

Example. Imagine you play four notes. The first and last you want as a detache, the two in the middle as staccato. However the last note you want to come right after the staccatto. Problem is it will play a staccato rather than detache, becuase how on earth should it be able to guess that of the three notes repeated closely you want the last to be a detache.

www.marcussenmusic.com/speedtest.mp3 Heres an example of what I mean. The first part is how I want the phrase to sound (using keyswitch) and the second time its using speed control.

So to me atleast it seems you would use it for things where the exact change of articulation isent pivotal. This could be going from fast to slow legato playing. So I'm not saying its useless at all - I'm mearly saying that if you hoped that you could load up the articualtions and just have it play the right ones, you may be dissapointed. Keyswitching is still the best method overall for this.

Did I explain myself clearly?

In all fairness here are some comments made by Maya from VSL in reply to my comments:


Hi Christian,

hope I can throw light on this question.
In the particular case described here the use of keyswitching would be an overkill (it is a way too powerful feature for switching between two states). There are several ways to change to a detache note again after playing the staccato notes detected by speed control.
A cell, where an articulation can be loaded (in this case detache and staccato) always consists of two areas, which can be triggered and/or crossfaded or toggled. So in case you like to have the last note of this particular phrase to end with a short note, you should use the speed control like you explained it in your 2nd example. In case you want the last note to be a detache again, you can drag the detache patch into the second slot of the staccato cell and use the cell crossfade option to switch to the detache notes during playing whenever you want.

But there is another way to do the same:
Load the detache samples into the 2nd line of the matrix. Then use a self defined controller to switch between the two lines of cells vertically. That´s a bit easier
...

Best,

Maya

So bottomline is that Speed Control is a nice feature, but implying that it can think ahead is confusing and misleading (as this thread shows). However speed control definately has its uses, but its not the solution to end keyswitching.

Hope the answers you question.

[kotori]

sorry if I sound noobish and stupid, but... VI only comes with the expensive libraries? It is impossible to get VI for the Opus libraries thus I am stuck with using Kontakt? VI as in the vsti controller for the sample libraries. And one tiny eencee (spelling) weencee (spelling) noob questions: is 2GB ram good enough for VSL Opus 1 + 2? How much is good enough? I apologize for off-topicness, but creating another thread for a a few answers is a waste of space.
Hi,
Yes, there's no VI Opus and sometime around last christmas VSL said they had no plans for it in the near future. I'm a Opus1+2 owner so at first I was a little put off by this fact, but then again I don't know how much Opus would benefit from VI since it doesn't have a huge amount of articulations to switch between and no repetition samples. It's possible to emulate quite much in Kontakt using scripts. Eg. you can set it up to play legato on pedal down and speed controlled articulation switching on pedal up using my MIDI Filter and Matrix scripts (http://nilsliberg.se/ksp/scripts/scripts.html). That said, I think a VI Opus will come at some point (this is me guessing), but that VSL wishes to wait some time to get many buyers of their current VI products first. I think you'll be ok with 2GB but it of course depends on the size of your arrangements (but you can always use the RAM purge function in Kontakt).

Nils

[Ouch that hurts]

Thanks Marcussen you've put your finger on exactly what I meant. It's reassuring to my (fragile) sanity to hear from an actual user that VSL haven't in fact altered the laws of time and physics, which is what the hype about this thing seems to suggest!

Your experience using it is exactly what would seem the logical outcome to me. I wasn't trying to imply that it was useless, only that it could only ever be a kind of rough-and-ready mockup tool, and that using it for any kind of finished version, in music of even moderate complexity, would require a lot of editing. This casts a somewhat different shade on the way it's been spun, as a massive time-saving tool that will make everything smoother and easier. And I kind of wonder whether there's a massive advantage in mocking up with it or just doing it the "old" way, with one articulation or a few used broadly with keyswitches.

That's all I meant. As one tool among many, used within the appropriate circumstances, I'm sure it has its uses.

[Marcussen]

Well. There is no doubt in my mind that you will get a better result, faster. But if your just looking for a tool to create fast scetches then you might as well save $10750 and get GPO.

Saving time is only a fraction of all the positive things that can be said about the VI.

[jovan]

Well. There is no doubt in my mind that you will get a better result, faster. But if your just looking for a tool to create fast scetches then you might as well save $10750 and get GPO.

Saving time is only a fraction of all the positive things that can be said about the VI.

O yes, and if you accept that the music sometimes sound a little bit thin...
A Mercedes & a VW - you may drive both cars...
And you can save a lot of money if you buy the Volkswagen!

:-) Jovan

[Alexcremers]

The future, Conan?

[Ouch that hurts]

Well. There is no doubt in my mind that you will get a better result, faster. But if your just looking for a tool to create fast scetches then you might as well save $10750 and get GPO.

I don't think anyone would consider VSL just as "a tool to create fast sketches". It's obviously positioned and priced very differently from that, as a solution capable of serious quality results. If I were to buy it, that's what I'd be buying it for. I already have a pretty good palette of orchestral sounds.

I was just interested to work out exactly where the speed control tool fits within all this, as the basic concept of it did seem flawed to me - or at least limited - in exactly the way you described.


[Saving time is only a fraction of all the positive things that can be said about the VI.

Certainly, I'm sure that's the case. I didn't mean to rain on anybody's parade who's loving the new VI. I just wanted to clarify things regarding this one particular feature.

[Marcussen]

hehe - no raining here. :D

I just needed to say that working faster is only partialy what makes the VI great. So if someone was reading this thread and was turned off by the 'speed control' thing I dont think it would be fair :)

Vienna Instruments is truly a leap forward in terms of computer orchestras.

Jovan. When you say "accept music will sound thin"... are you refering to VSL sounding thin?

[jovan]

1) Vienna Instruments is truly a leap forward in terms of computer orchestras.

2) Jovan. When you say "accept music will sound thin"... are you refering to VSL sounding thin?


ad 1) I totally agree.
ad 2) I don't refer to VSL sound. It is the other one. I own it and tried it several times but I was never satisfied by this kind of sound.

:-) Jovan

[Bruce A. Richardson]

The more common part of the "seeing into the future" aspect of Vienna is the creation of a small latency, within which it is possible to analyze what you've played and to insert the appropriate keyswitch before passing the MIDI note on to the playback engine.

As processors continue to get faster, this amount of latency we musicians can tolerate yeilds ever more thinking time.

All of this has to be considered "time savers," since one of music's primary attributes is the establishment of pattern for the express purpose of breaking it for musical effect. So, any kind of predictive technology will always eventually get it wrong. There is always a need to surgically fix those things, but the idea is to get you through the bulk of the easy stuff, so you have more time to deal with the less easily predicted parts.

[Nick Batzdorf]

VSTi

I'm about to go on a 1-man crusade against calling virtual instruments VSTi-s. :)

(Even though VST is one of the Vienna Instruments formats.)

[Crystal]

Very interesting thread.

I don’t own VIs but watching their tutorials, I see a incomparable more productive tool than any other samplers. The Universal Mode example in the video is quite impressive.

But Marcussen’s remark shows that it doesn’t fit to all situations, and needs note by note tunings sometimes.


It’s another proof for me that the VSL team should seriously plane to embed their own sequencer in their library.

Imagine the power of the tool : you record a line with the Universal Mode to do the biggest part of the job, and then chose any sample you want for any note in the editor, step by step.

No more key switches.

The ability to get a real playback. (My dream...)

And even the ability to ‘think ahead’ on playback.

[Mike Greene]

I'm about to go on a 1-man crusade against calling virtual instruments VSTi-s. :) I just went into march formation right behind you.

- Mike Greene

[mirage]

Oh..well...gonna be snooty about the word "vsti" eh?. Consider this my apology in advance because I use the word all the time and it's probably not going to be easy to change.

I also still say, "hey let's tape that movie on tv tonight" even though I've been on dvd burners and tivos etc for five or more years.

[Nick Batzdorf]

The only reason I don't like "VSTi" is that it's confusing to people who are new to this. I chose Virtual Instruments - after much agonizing - for the magazine simply because everyone gets the concept instantly, but none of the other nonspecific names bother me in the least - software instruments, instrument plug-ins, softsynths/softsamplers, or whatever.

[howardv]

Don't know how VSL does it but selecting alternate release samples based on elapsed note duration might do the trick. One set of releases for long notes and another for very short ones. I think this might work well for certain instruments like pianos where notes tend to lose energy the longer they are held making for releases that are somewhat different sounding than those in rapid-fire staccato playing. Scripting should be able to pull it off too.

Howard