Symphonic Cube : More velocity layers?

[charl]

Dear VSL team, there's no doubt in my mind that you've developed a great product. I am however wondering if you have plans for more velocity layers for instruments in the Symphonic Cube.

Frankly I think 4 velocity layers for an important instrument like 'solo violin' is problematic because you can really identify the 'jump' when velocity hits the next sample zone....

I've had this feeling a few times already, and am begining to wonder what your plans are for future releases? I would be quite troubled to hear that the Symphonic Cube would not extend the number of velocity layers contained in important articulations of key instruments like solo violin, cello etc...

If decisions have not yet been taken, please take into account this VERY IMPORTANT issue which I'm sure is a concern for other users.


Please feel free to point me to a thread on this topic if answers have already been given on the forum. (The search doesn't work properly)

Many thanks in advance
Charl [:)]

[evan evans]

VSL,

yes this has been one of the lackluster features of the library. Most of the other large libraries have 8 velocity layers. I actually expected more when i bought VSL because everyone touted it as having all the nuances of the instruments. But what good is it if all the nuances are at "pp" and "f" only ? Good for baroque music perhaps (a musical form that I've grown to dislike).

Although there is some instruments with 4 and on the rare occasion 5 layers, I would have to say that the "median" is 2, a strong formidable 2.

If there's anything that could threaten the #1 spot that VSL is enjoying it's that VSL would continue to not provide a full range of velocity layers. I can always EQ another massive library to be well within competitive range of VSL, especially one that might "wish to" compete with VSL in the future.

As an adoring user and someone who does not let geniuses off the hook easily, it is my, probably irritating, recommendation that the incredible VSL consider:

1) releasing their instruments with a larger sampled velocity mapping
2) providing more alternate notes than 2 (for instance it would be nice to program the alternation tool with a set of 9 different sampled notes instead of the repetition tool with a prefixed performance of repeated note; this would also allow for realtime repetition instead of programmatic repetition)

Evan Evans

[jrm1]

I would have to agree with all of the above. Reconfiguring existing sample data contained in the rep patches so that alternation instruments with 9 different hits would be a far more flexible and usable approach to fast (and not necessarily repeated) lines. For me at at any rate. And more layers are promised in GS3, so perhaps the VSL guys are planning more layers for existing instruments in any case.

Breadth of choice of instrument is far less important to me than the range of expression contained in each one.

That said, I'm amazed at the attention to quality. Thank you.

[evan evans]

Even 5 different hits could be matrixed to the following long unnoticeably repeated configuration:

12345342134254231524312513245432

That's 32 distinctly DIFFERENT adjacencies (notes next to each other), and the next set of 32 could be the same in reverse, giving 64 before a single loop of the pattern gives it away.

it would circle around and not sound repeated in any way with just 5 different notes for each velocity. This, and now this is my opinion, and those who respect me can value it if they wish, ... it would have been wiser, based on my side by side comparisons, to have recorded 5 to 9 individual instances of notes, rather than the performed repetitions which take up exactly 25% of the libraries. It would have allowed for full realtime flexibility of tempo, phrasing, and eliminate the need to anticipate a future phrase of any number of repeated notes before it occured.

in my tests I found no quality difference in repeated notes using the alternation tool and performed reptitions. I have ears, and it simply was not worth the extra 25% of disk space and confined programmatic architecture. The other way worked just as well if not 99% perfect, and gave the user 500% more power.

Evan Evans - Working in the real world

[charl]

Evan I agree with you on the multiple alternation issue.
I do think this method would very much enhance flexibility & I believe I read an interesting post of yours in which you were describing how you cut out some samples from a repetition sample and you reprogrammed the stuff in an alternation tool (or something like that)...

I do find that it would be a huge bonus -in terms of user productivity- if VSL would offer the confort of having "single articulation multi alternations programs" with multiple velocities..
Of course, "Multiple articulations + multiple alternations + multiple velocity" instruments is where VSL should ultimately go, and I clearly believe they have this in the back of their head.

Two points on this:

1.Sampler & computer specs: The power is there.(Or nearly there : I'm hoping Exs will be more powerful in Logic 7) I also believe the freeze function & sample 'memory' management will evolve towards much greater flexibility.

2.Which instruments?: I would say 'very expressive' lead instruments like solo violin & cello would be great starting points for "Very extensive programs".

The question is 'when' & at what price will this be available for us VSL users.

[charl]

Just to re-focus on the velocity issue...

I would be so grateful if someone from VSL could give us 'some' kind of comment on the subject.

I'm begining to wonder if this is a touchy issue and am surprised you guys haven't expressed your thoughts yet in response to my original post.

Has there been other threads on this? [*-)]

If only the 'search function' would work apropriately. Conducting searches before posting does help to avoid waisting anyone's time.

Regards

Charl

[michi]

Charl, we recorded as much velocities which made sense to us, that means up to 6 different velocity layers in one certain patch. the whole range of dynamic and sonic possibilities of one instrument is covered by the pure amount of the different patches and performances. we don't believe in doing 100 different staccato velocities, although we know about their existence, because the difference from one layer to an other needs to be relevant, otherwise it's a tough job to sequence let's say for example, a pianissimo line without jumping between the different velocities all the time. if you need a transition from one dynamic step to the next smoother or whatever, you can choose our crescendo- and diminuendo layers or -repetitions. of course, that means more editing at the moment, but we are still at the beginning of a long development road.

cheers, michael hula/VSL

[charl]

Thanx michi for providing some thoughts...

I understand most of your points quite well..
It hasn't been VSL's philosophy to create patches with a great many velocity layers.. simply because with all the different articulations provided by the library the user should be able to find what he needs...

Well most of the demos show that in an "orchestral" context this works pretty well - setting aside the fact that getting great results with many many different articulations (à la VSL) probably requires "more work" as you say yourself.

@Another User said:

we don't believe in doing 100 different staccato velocities

Neither do I [[:)]]

But I recently put my hands (literal french expression) on an exs 'solo violin' sample patch with 16 velocities & 4 articulations (via keyswitches). Frankly it's a wonderfull experience.
Expressiveness is beautiful. And basically I think it's gonna be for me a real quality (+ time saving) tool for making music. Of course this is no answer to every musical situation I'm going to encounter in the solo violin department. This is why my first choice was the Horizon solo strings.

You may not have believed in 'recording more velocity layers', or 'integrating more velocity layers'.But I have trouble understanding why you would want to stick to this philosophy because my experience (& I can't be the only one around) is telling me the following :

Not only do these extra velocities make a real audible difference, but also bring a greater expressive response with 'no' extra effort from the user who's playing the instrument. It is a fact that 'dynamic layers' on some solo instruments just doesn't cut it at all - simply doesn't exist.
That's why I believe many users would largely benefit from extra patches and might actually be prepared to pay for them. There are many occasions where they'd be a real time savers...

I see your point on 'sequencing difficulties' with too many layers, but let there be 'lighter' patches for those who prefer them in certain occasions. I wouldn't agree that this is an argument for not having more extensive patches...

Finally, I'm aware that in a 'global orchestral context' having patches with fewer v-layers is probably not much of a problem. But I still believe that leading solo instruments REALLY should have more velocity layers, especially if they might be used in a different context than an orchestral one, and in which there aren't as many instruments playing simultaneously.

My point is: Let me share with you 'developers' my very sincere enthousiasm 'as a user' for a feature that I believe you may have slightly overlooked, and that I'd be very keen to find in further developments of your library.

sorry for the long post
Charl [[:)]]

[jrm1]

Live players produce a different nuance in every musical event they make even when attempting to stay in the same dynamic. That's what makes it sound real. There is no doubt that each sample in the VSL sounds real and live - it is after all a recording of a human being making a musical action each time - but as soon as the ear hears the same sample again it all sounds less real.

Michi and the VSL team seem to be suggesting that this is not all that important - after all, when the user brings in more instruments the degree of artificiality (repetition) is masked sufficiently. I would strongly disagree, if that is the position, that this is a small issue.

A 'for example': you want to start a piece with 14 violins playing staccato with a slow build in dynamics over 64 strokes of the bow. The sample 'signature' becomes exposed: even with the repetition tool. It just sounds FALSE. Brilliantly recorded samples, but not enough of them.

WE NEED MORE VARIATIONS OF THE SAME. AND I'M SURE MANY USERS WOULD PAY FOR THE EXTRA.

[KingIdiot]

Most of the libraries have 8 velocities? thats BS. you're lucky if you get four. And not all of them have alternate samples. There's only one developer, I know of, that does this.

Anyhow.

Use the repetition samples for more dynamic variations and inconsistancies between note to note performance..

They offer u p to 8/9 variations of the same dynamic and the crecendo/decrecendo repetitions off multiple velocities.

When it comes to slower notes use slower repetitions, if possible. Try to layer them on top of or crossfade into other sustain samples if you need longer notes. If not, use various different articulations. Using dynamic Xfade patches will allow for alot of variations on note to note performance.

VSL is the only library out there with THE MOST OPTIONS for variations in note to note performance. It excels at this, and will probably be even better as time goes on and more features are implimented and more sampler options become available.

The problem is, you cant think of VSL patches as typical "one patch does all". Its best used as a construction kit IMO (not that it cant be used as standard "simple" playback).

Once AMP comes around I think things will be alot easier. (tho it will take time to get all scripts to work at optimum levels)

64 velocities is never gonna happen. Its incredibly difficult to get a player to even be able to differentiate that much, add that it would take forever,...plus all the inconsistancies from note to note would be MORE obvious since it would never sound like a real player playing each consecutive notes (this is one of the real problems with samples if you ask me)

a 32 note repetition crecendo and 32 note decrecendo repetition might be a way of achieving it tho. Then chop up the repetition or use release triggers.

[michi]

@jrm1 said:

Live players produce a different nuance in every musical event they make even when attempting to stay in the same dynamic. That's what makes it sound real......

Michi and the VSL team seem to be suggesting that this is not all that important.....

no, jrm1. this is not our suggestion. but we had to define targets, which are possible to achieve. four years ago, people thought we are crazy, when we talked about our project, today we are facing a product which makes everybody proud being part of the team. and we are still pushing! read the second part of the last sentence from my earlier post again, you can trust in our graveness.

thank you for your input.

[Ren_]

I agree with King's point of view. I don't think "most of the other large libraries" have eight velocity layers. Not at all.

Also I think there's some confusion on what a 'velocity layer' must do. I read about uses of the velocity layers to avoid machinization, and I fully agree. I hope this can be done somehow.

However, from here to 'Barroque oriented'... my god.

-René

[Dietz]

The main misunderstanding is that many people think old-school: One patch = everything you have.

The crucial point is that there is actually little need to do those "fake"-dynamics on behalf of layering when there are actually _played_ variants of them. Yes, this is a new way of doing things, and no, this is not what one would be used to make with the common "String Pad", for example.

Of course, with the means and tools we have up to now, this is sometimes more demanding than the mentioned "Pad"-approach. But what you get is _real_ performances instead of faked ones. And rest assured thet we know pretty good that there will have to be new ways to play with these possibilities even more easily than right now.

/Dietz - Vienna Symphonic Library

[charl]

King,

@Another User said:

The problem is, you cant think of VSL patches as typical "one patch does all". Its best used as a construction kit IMO

This is a good way of putting it.

And that's exactly the problem..

I actually happen to believe that you should be able to ALSO think of 'SOME' VSL patches as being typical "one patch does alot" patches. Thus opening the library power to different approaches for different applications, jobs, deadlines.

The samples sound great, the angle from which they can be integrated in our workflow could be multplied..

What's this "AMP" thing? Looks interesting but I have no idea what it is?

Also who's been talking about 8 velocity layers as standart practice? I missed that - not saying it isn't there just can't seem to find it easily.

Anyway, don't we all agree that '10' should be minimum requirements for any 'top notch' sample library. [8-)]

Michi what are you guys up to? Revolutionnary concepts again.. [:)]
If you plan on developing multi threaded-realtime-frozen-velocity legato, I'm afraid I've already copyrighted the idea + domain name.....

Charl

[Dietz]

@charl said:

[...] Anyway, don't we all agree that '10' should be minimum requirements for any 'top notch' sample library. [8-)]
[...] Charl

-> http://vsl.co.at/english/pages/profile/news/sample_counter.htm - and still counting ... is anybody in for 6.000.000 samples ....? [+o(]

/Dietz

[evan evans]

@Another User said:

if you need a transition from one dynamic step to the next smoother or whatever, you can choose our crescendo- and diminuendo layers or -repetitions.

are so typical. TYPICAL. I'd like for once to hear them say, "Yeah, we should have done that. That's a great idea."

YOU WILL NEVER HEAR THEM SAY THAT.

They will not admit to being wrong , that their library has areas needing improvement. i don't really understand how they could all be like this. All I can think is it is truly some Viennese thing, as insulting as that sounds. It just doesn't make sense, this mass egoism.

I think we should all give VSL $1 each, the day they actually admit that the library in it's current condition is in need of improvement. NOT in a NEW library with more articulations, which is AGAIN just another way of avoiding the issue and giving a work around, ... but in a genuine apology for not having delivered to many customers needs, and admitting that the library is missing things that should have been considered, and that the engineering of their software did indeed miss some very valid strengths if not an entirely different approach to a more practical library.

I mean, this library is not exactly practical. I love seeing VSL tell you "it's practical" and in the same sentence tell you that all you need to do to get repetitions to work is go to this particular thread and read the 7 pages of 40 posts on how to do it. If that's engineering than I guess I'm not a rocket scientist.

Sorry if it's hurts fellow VSL users, but I will not stand by and let them get off the hook so easily. Playing devil's advocate might not be nice for the opposing party, but it can actually be a very effective means towards innovation and improvement. i think listening to customers criticisms, and providing reassurance that the customer is onto something that VSL would consider and is something that VSL had not thought of is a powerful business tactic that the VSL crew do not seem to like very much. Instead, they like to say, "the library is amazing","this is the ultimate library","a great deal of money was spent on this library","the musicianship of this library far exceeds that by others in comparisons".

You know I remember this attitude from when I used to deal directly with the developers of the notation software, SIBELIUS. i would say, could you please implement harp pedalings? The twin brothers in private meetings with me and other Hollywood heavyweights, would answer, "you just draw them with the line tool." Can you believe that? I have multiple instances of them trying to convince us that a work-around was gonna make us happy. That kind of customer relation is useless. What would have made me happy is them saying, "That's a great idea .. we will add that to our suggestion list, and perhaps implement it in the near future." Workarounds are silly excuses.

A real leader admits fault. Every CEO knows that (except one in Vienna I guess).

Evan Evans

[evan evans]

@jrm1 said:

Michi and the VSL team seem to be suggesting that this is not all that important - after all, when the user brings in more instruments the degree of artificiality (repetition) is masked sufficiently. I would strongly disagree, if that is the position, that this is a small issue. ... WE NEED MORE VARIATIONS OF THE SAME. AND I'M SURE MANY USERS WOULD PAY FOR THE EXTRA.

I, OF COURSE, agree with this and second it, including the bold yelling.

[:)]

Evan Evans

[evan evans]

@Another User said:

This is a good way of putting it.

And that's exactly the problem..

I actually happen to believe that you should be able to ALSO think of 'SOME' VSL patches as being typical "one patch does alot" patches. Thus opening the library power to different approaches for different applications, jobs, deadlines.

Are you getting the same clues that I am Charl about how they treat their customers with work arounds instead of listening and saying thank you?

8 Velocity layers is the start. 10 would be the mark of a great library.

They've got the room, the business plan, the musicians. What's the big deal? It;s not like they have to spend much more money. and as mentioned above, I would certainly pay MORE for MORE. unfortunately I was dissappointed that the $8K to $12K I spent was for as meager of a library as it was. But I suppose if it's worth it, we pay for it. That doesn't mean we are happy about it.

Evan Evans

[evan evans]

But let's get back to the actual title of this thread shall we?

WILL THERE BE MORE VELOCITY LAYERS IN THE SYMPHONIC CUBE?

I understand if you can't comment on that at this time. That is a sound business tactic that I agree with. But who knows, the VSL team has shown incredible graciousness in the past with giving information to their customers/users.

So what's the word VSL?

[:)]

Evan Evans

[cm]

evan, i'd ask you kindly not to post anything for - hmm, let's say three months.
thanks for your contribution so far, others might follow.
kind regards, christian

btw: i'm sure you understand the inscription *i am root* - edit: and please don't ask me to explain it!