@evanevans said:

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

Evan Evans

Evan,

I don't want to add fuel to the fire, but this just isn't necessary. You often mention that you're working in the real world - well in the real world the dynamic range you can ask from a player is normally from ppp - fff. I make that 8 velocities. The truth is (and I know you must know this because you work with live muscians) that there is very little audible difference between ppp & pp, and ff & fff. So that would require even less.
The Debussy La Mer mock-up was created using VSL samples that have no more than 2 velocities per instrument. Not once did I receive the criticism that the dynamic range was lacking (maybe you disagree?). When you consider how many samples have been/are being recorded, it's a little naive to think that VSL would listen to every suggestion and respond accordingly.

I think constructive criticism is healthy, but we also need to recognise that VSL have achieved a massive leap in realism, and in the most part the technolgy isn't sufficiently up to supporting it fully. I'm happy to have the Perf-leg & RepetitionTools with only limited velocities as they're really working for me and (more importantly) helping to feed my family.

Andy.

p.s. I understand if you feel unable to reply.

Wow,

another Evan mega-post (you can't resist can you?) [[:)]]


I just want to say a couple of things..

My idea in this thread is to create à discussion, à friendly constructive discussion on certan aspects of the Library. I frankly apreciate the fact you express your opinion & insight on the velocity issue but I don't feel the tone of your post, Evan, is helping me out.
I really apreciate having answers from the VSL folks (even if they don't always match my expectations - but this is just fine with me) and I just cannot imagine how they'll enjoy participating in a thread that's blatently becoming a personnal rant (and not particuliarly kind) about the way they handle their communication.

I'm pretty gratefull that these guys take some of their time to post, and I also understand how & why VSL has gradually become such a passionate thing for them. I like that. It shows how much of themselves they've given to the project...

I personnaly try to express my thoughts & ideas in the most positive way I can because my 'humble' experience in life has shown me that this is how you get the best of people & the best out of yourself.

I take quite a bit of time when it comes to writing my posts. English isn't my native language & I want my points to come across with max transparency.
I understand VSL folk's replies can sometimes be disappointing. I personnaly am disapointed by some parts of their replies because they decide not to comment in detail on what I consider to bethe important points. However even if I feel they're not 'textually' adressing the issues I'm raising, I respect that it's their choice not to.

Further more, the reality is that they (as insiders) have a broader picture than I on the future of the library. The best I can do is tell them "Hey listen, this is how I, as a user, feel about what could make your product better & more flexible." They'd be idiots not to listen. And I think they are actually listening.

I actually suppose (& trust) that they might have an "undisclosed" plan on how to give us the flexibility that's missing in the current VSL library.

Well that's how I understand Michi's posts anyway.
And it's my only explanation as to the fact that they're treating some of our thoughts and proposals as simply irrelevant... I accept that.

To me, my ideas & suggestions seem 99.9% relevant untill someone (or something) opens my eyes on why they're only 99.9% (This is why I posted on the forum - I really don't mind changing my mind if it's for a better understanding of a situation )

I still haven't changed my mind because they haven't convinced me yet that I was wrong. Also, they don't seem to need to convince me right now. That's no biggy, I can live with that, so can they, and we can continue having a whole bunch of other pleasurable discussions on music, vsl or whatever...

I for sure am really looking forward to discover what our VSL friends have in 'la grange'.
And if the stuff they come up with doesn't match our X-treeeeeemly high (near godlike) expectations, we will be once again confronted to nothing else but... our expecations. [8-)]

More seriously, I for one will surely continue to sypatheticly tell them what I think could still make the product better... But for a start why the hell don't they fix the 'search function' once & for all. So we don't all endlessly repeat ourselves. [[:)]]

Charl

Evan, if you have any further thoughts you'd like to share feel free to PM me.

Evan Evans' posts on this are way out of line. I want to say I totally disagree with what he is saying. specifically -

VSL apologize to users? Are you kidding? - for giving them the most revolutionary advance in the art and science of sampling yet created?

8 velocity layers are necessary - that is absurd. Of course it is always nice to have more samples, but you could create a recording with three and one with eight, and nobody could tell the difference - if you know what you are doing.

"Pompous business atittude" "Any criticism from customers becomes a "touchy" issue wih VSL. Again, it is their motus operandi."

- that is total bullshit. The VSL has showed an incredibly open and friendly attitude that responds to their customers as much as any business I ever dealt with.

"Some Viennese thing... this mass egoism" -
Not only is this prejudice, but it is total crap - there is no egoism that has been shown by the VSL at any point, only fully justified pride in having created a great product. This is a real insult and I don't feel the person who wrote this has any damned right to do this on the website paid for by the company he is attacking. Evan, you are goddammed lucky to be using what you are using, but instead whine and bitch because it isn't flawless. Guess what - nothing is. Do something better and get back to everyone here. Don't respond to me Evan, I am sick of this crap you write and I'm out of here.

make love , not war

[H]

@Carter said:

make love , not war

[H]

You want Evans to make love to CM? I'm not sure Christian would agree. [:D]

I'm glad he's "on a break" for a while. I think he ought to spend less time typing, less time complaining, and more time actually writing music. I'm hoping that other people will feel freer to post in the forum now, without Evan's huffing and puffing going on.

Charl,

Yes I agree that there should be some "do it all patches". the problem is the limitations are eing reached with alot of VSL stuf. Dimension splits are abound, and there are programming limitations. Their Design isn't based on a sound design/synth approach. Its more methodical. Which IMO is what is needed for some things, and only VSL is going to offer a solution.

They've also got grand concepts that go beyond whats already released (outside of more samples), and they are working to that final solution. Spending time recording and editing and sound designing for whats not their ultimate goal, might jsut be a waste of time to them (on a scale of weighing it against their ultimate goal)

I'd love to see some more "playable" patches, and have made some for myself (that are mixed with other libraries to save on poly and arent too consistant, because I threw them together) in Kontakt. It will be possible to see some more playable options in giga 3, and I'm sure you'll see them pop up.

One problem with "playable" patches is that people will ONLY use those. I mean look at the people already not willing to load up other patches. Give them even MORE reason to fall away from sequencing with the approach that will give them the best results, and people....lazy as we all are, will take it.... Thus making them wish for more "easy" stuff...that wont ever reach the best results.

AMP will allow for scripts to be written that affect the MIDI stream based on lengths of notes, you could load up all string patches and have it automatically ick the best patches and make adjustments to have them play at the correct length. (actually combined with a few other scripting programs I think it could do some REAL damage with autoloading and unloading, but tahts WAY later when I start harping on MAx and Ian to include some crazy ideas)

About the 8 velocities, that was Evan who made the comment about "most" other libraries having 8 velocities.

I personally think that 8 velocities is overkill in alot of instruments. Not to mention it can really make for samples to stick out. I'd rather have sound design options and morphing textures mixed with a sort of modeling. Or frequency domain control (ala GPO), with alternate takes of the same velocities. Attack inconsistancies are a real problem and its hard to keep a uniformed attack between each velocity (which is why I'd like to see morphing, and ADSR stuff mixed together...but hey thats me...and i'm weird...)

Actually to be honest, I have a lot of different concepts now on what will make an ultimate sample playback set up...but some of it involves a total different approach to sampling anyway...others involve VSL's approach, others involve design stuff that was an offshoot GPO and is pretty cool (wait till gary releases futur stuff)...and blah blah blah....it'll never end, and we should all give up and use what we got and whats comming out, because we'll always want to hear a real orchestra no matter what....if not go sample your own orchestra [:)]

yup this is why I dont do much orchetral work right now...and have found that rawk star drive in me to pick up the guitar and scream into a mic, drink lots of alcohol and get the chicks.....Its so much more fun.

King,


as usual some interesting points..

You're right about technical limits.(ex : Legato instruments) Vsl are clearly pushing the borders and trying to anticipate what will be technically available. That's really good...

@Another User said:

...that rawk star drive in me to pick up the guitar and scream into a mic, drink lots of alcohol and get the chicks

I use to do that 10 years back (the guitar & mic stuff), It was such a fun experience we decided to do a 'remember' one shot gig last year.. Oh man this was like a 'huge' flash back on my twenties... eh eh eh ... just wonderfull...

The chick stuff, I also use to do 10 years back but I haven't found a good reason to quit yet... [[:|]]


regards
Charl

Oops.. lenth limit..
____________________________________________________

And finally the 'essence' of your post :

@Another User said:

...that rawk star drive in me to pick up the guitar and scream into a mic, drink lots of alcohol and get the chicks

I use to do that 10 years back (the guitar & mic stuff), It was such a fun experience we decided to do a 'remember' one shot gig last year.. Oh man this was like a 'huge' flash back on my twenties... eh eh eh ... just wonderfull...

The chick stuff, I also use to do 10 years back but I haven't found a good reason to quit yet... [[:|]]


regards
Charl

Some more 'playable' patches would be great when GS3 comes out... I woudl really appreciate that... or to get an art file from someone editing a super instrument in GSedit if nothing else...

I think the Guitar was the effort of only a couple people at VSL, and done "on the side". Not directly involved with the schedule, with regards to VSL's other instruments...


but anyway, thats not what I'm talking about. What I'm talking about is working towards a bigger concept, in terms of how VSL samples will be used. Do they spend time working on more instruments and working towards an ultimate playback solution, or do they work on programming techniques, developing "other" sound design routines, that they take time out of their schedule for options that dont fully really go with their bigger plan. (there is a fair amount of sound design going on behind the scenes I'm sure)

i thnk what I'm trying to say is I can understand their current mind set, with the time that it takes to develop some patches in giga, and the limitations currently available. Things may change with future sampler formats and options....not to mention you might see other options that they've been planning show up that will kill your need for "generic" patches.

I made patches that combine VSL with other libraries that are "more playable" to MY idea of playable, they definitely arent the best at realistically creating an orchestra, they are just "quick" patches to match with other quick patches I've made for other libs. I'm not saying that it wouldn't be cool to have "quick" patches, but man....these dont give me the best results. The best results are from switching articulations manually. I've also found that these "simple" patches (in all libs) tend to make me NOT use the rest of the library and make me "write for the patch". Still I use them, and move on because I "settle" on it,...mostly because I'm lazy, sometimes because of time (but again there's more things that will help with time than just "generic" patches), and also cuz I like that "fake" sound sometimes...with all the weird quirky stuff I've been doing for gigs.

now back to the point of the thread and connecting it with this post, more velocities, that doesn't automatically equal "more playable"

I think practice using VSL might be a better thing, than relying on "quick" patches. Practice leads to faster working.

About future stuff from other companies. Gary is pretty creative and has alot of talented people working for him on designs. He's also a big advocate of "keeping things simple, but giving the most options" now, (as am I in most libraries...just not VSL because its a broader concept). I'd expect the new stuff to be realyl cool. If its some of the older development concepts that I was involved in, then it will be very cool and unique stuff. If its newer stuff, then who knows, itmight be even more revolutionary.

I think a ombination of sampling, morphing and physical modeling would make for an amazing "virtual instrument"

however I also think that an off beat approach using a variation of granular synthesis/time manipulation could make for some really amazing results and still offer real-time playback.

Point is, theres so much out there in other peoples heads that you cant know it all, and exactly why things are done a certain way.

If it all starts to suck at some point, I'll make some patches, and you all can kiss my ring and buy me beer.

These are very interesting posts from King and Charl. Isn't one question to be asked about all of this: what is your goal?

Is it to have a playable musical instrument, like a piano, but with orchestral samples?

Or is it to realistically program on a computer a virtual orchestra?

In other words, these are two separate worlds: one is the creation of a "Super-Instrument" that is a real-time performance-oriented musical instrument but allows infinite orchestral timbers. The other is a composing-oriented, non real-time process like notating a score but is done by building up sounds instead of notes. Both of these enter into the process to some degree but you have to decide which you want to emphasize depending on your mindset and your talents.

I notice something similar to what King said in that I get lazy and use load-and-play patches because they allow one to concentrate more on the music. I also have a real inertia and sometimes aversion to finding the exact crescendo that will perfectly express part of one line in one instrument in one section of one cue - when there are a hundred others waiting to be done. This principle of laziness has a lot more influence than I think people are willing to admit. They like to talk about awesome systems and everything done perfectly with infinite precision, but don't actually use those in practice. It is also important to distinguish between the ultimate in technical possibilities, and what can actually be used in a practical way: the difference can be huge. EVen though I accept the idea of "programming" music, I keep on trying to create a more directly playable, instantly loadable Super-Instrument.

@Another User said:

Is it to have a playable musical instrument, like a piano, but with orchestral samples?

Or is it to realistically program on a computer a virtual orchestra?

I'd like both, please! And I believe it'll happen.

@William said:

I notice something similar to what King said in that I get lazy and use load-and-play patches because they allow one to concentrate more on the music. I also have a real inertia and sometimes aversion to finding the exact crescendo that will perfectly express part of one line in one instrument in one section of one cue - when there are a hundred others waiting to be done. This principle of laziness has a lot more influence than I think people are willing to admit. They like to talk about awesome systems and everything done perfectly with infinite precision, but don't actually use those in practice.

I think that´s a very wise observation. I experienced the same personally. Still I won´t never admit that! [[;)]]

I use the "_all" patches alot... I hope they will be expanded upon and made better once GS3 allows for it

Sometimes I'm working on some cues or tracks and I like to have everything at my fingertips.. I want to be totally focused on the music... Looking for articulations doesn't help me in those circumstances.

Now we all know that in reality writing music is always also reinventing the procedure of writing as well, but to cut things short, I could generalise my way of using the library like this:
When the main stuff is done, I will gradually dig deeper through the 'production' & will usually prioritize on parts that require the most 'fixing'.

Fixing can be indefinate but there's a point where you simply have to stop and deliver the music to who ever has to pick up on the job or just switch to another job or cue..

Personnally my experience has made me realise that the better my 'all' patches are, the less fixing I have to do afterwards... Again I'm being schematic here, but just to illustrate my point, here's an excellent real life example that happened this afternoon.
Earlier today I loaded a "basic all" solo viola (or violin or cello) I can't remember exactly...I'll check when I get back to work tomorrow. I was composing a quick snippet for a portion of a commercial.
Well, the patch had 2 velocity layers per articulation.. I played for 20 secs, clearly understood that 2 layers wasn't enough for me at that time. And I'm afraid I then turned to another library.... And found an 'all-round' patch that performed really well. The line was done in 3 min. & I didn't have to load any other patch or fix anything at all.

Frankly I would've preferred samples recorded on the silent stage but there was no way I wasn't gonna go the quickest way on this job..

I agree there may indeed be sometimes when the the fact you're using an 'all' patch is going to influence your musical idea... And it's a good idea to keep an eye on this phenomena.. I know I try to..
There's a kind of natural balance to find..

A good counter example is that I will sometimes load a very specific patch with a precise instrumental technique or something of the like even though I don't know exactly what I'm gonna play on it.
It's just my instinct that's bringing me to it ... It pops to my mind that I should use it at that point of my work. It'll sometimes work very well, sometimes not that well at all. [:D]

Also, King's comment on 'practicing' is a good one. I think with VSL the more you use it the faster you work. (Using it is also practicing). Maybe there WAS a VSL patch I missed this afternoon. I just went where it seemed the fastest easiest way .... 1 patch for writing + 3 patches for 'fixing' was not an option..


King Idiot :

@Another User said:

not to mention you might see other options that they've been planning show up that will kill your need for "generic" patches.

As I was saying, that's exactly what I hope (& trust) they'll do.. So why are we actually waisting our time on the subject of velocity layers...[[;)]]

Untill then, even though I agree more velocities doesn't 'necessarily' equate to more playability (michi's point on sequencing is a good one), in many cases I feel it actually really does provide more playability (By this I mean more depth in the reactivity of the samples thus more expression for an equal effort in the execution)...

Obviously I don't care much about the quality I'm only looking for a great 'return' and a minimum of mouse clicks.... ...... [6]

where jsut at a point where technology isnt there yet. I mean all the control in the world and options from jsut one patch jsut means you have to have 8 controllers to do it all in realtime, or have a MIDI instrument that pretty much makes it so that you might has well learn the real thing and record it.

I'm with you on gettin high return with less mouse clicks, believe me. With all the editing I do....BE freaking LEAVE me! Its worse with the actual software makers [;)]

What d'you mean you use software? who needs software...

You know how I get the best results with VSL don't you?...
I like to do it the real way :T.A.S.P. ( TRUE ANALOG SAMPLE PLAYBACK)

When I need some staccato tuba, for instance, I spin the corresponding DVD on my left hand little finger. I then read the tuba samples file with my right hand thumb nail (grown to this effect). - Aiming 'does' takes a bit of practice + a good knowledge of the DVD content.

What about 'playback'?

Good question.

Insert all your left-over fingers in mouth, and bite vigorously to insure proper proliferation of the wave form, from the finger bones to the teeth, and then from there to the jaw bone, ensuring maximum resonance of the samples in the whole of the cephalic cavity (from chin to forehead right through to the cervical vertebrae)..

Remember what John Lee Wonder said about the collection :

"In the right hands, those samples will make wonders"

Now is this groudbreaking or what? [8-)]


PS. Did I forget to mention how enjoyable and highly addictive it is to perform the 'Peter Framton like (talk-box)" effect on your favourite VSL timpani crescendo rolls?

My long-standing struggle with "basic" instruments is the way they're thinned out. They cut out entire velocities. We must go from yellow to red. We lose the orange. So when I'm writing in a hurry, paradoxically, I can't use basic instruments because the dynamics are so "either-or," and I haven't got time to tweak.

This is obviously more prep work to do, but if samples of all velocites were cherry-picked and then stretched just one key down, the dynamic spectrum would remain. It's still a compromised sound, of course, and I'd be reluctant to use it in a final track. But I'll suggest that a C3 root key sample stretched down to B2 is less of an audible difference than a leaping, snake-bite forte jumping out because the note velocity is one integer too high. This way, our basic instruments would be more playable and the RAM savings and palette availability would remain.

A classic example of this is what Dave Govett did in the reduced RAM versions of the GOS. Those are more than "basic" patches. They aspire to deliver the whole dynamic spectrum and save resources at the same time.

Pretty neat...