For what it's worth, I have spent so many years trying to get VSL Solo Strings to sound as good as a real player that it has become obvious to me it's an impossible thing to create.  Unless the part is so basic and so minimal, and then also only spands a few measures I can pull it off.  But if I expect to pull off more than that, I have revert to hiring a real solo player.  I tried for 3 months to get a simple performance to sound strikingly realistic but in one day, I accomplished this with a real player and a small amount of cash.

I do find the solo strings great for fast, shorter note stuff though such as classical Bach and other musical literature that doesn't require extremely expressive long notes, or dynamic moments.  They are amazing to layer with as well.  They have a beautiful raw sound to them.

I know that great things can be done with the solo strings but even being the best ones you can buy on the planet, they have limitations and as Daryl said, I can't even imagine what it would take to recreate that type of performance.  Perhaps a Sample Modeling type string instrument but the problem is, it's far more difficult to create a sample modeled string instrument than say a brass or wind styled instrument.  It would take much skill to operate a sampled instrument of a Violin and create a perfect performance that contains every aspect of a top notch world class player (which is what our expectations are/would be).  Also, how much time do you have to edit and edit and tweak?  If you have unlimited time, and devote yourself to perfecting it, even more possibility opens up to you but it's often hard to remember the vast amount of possibilities that are already available, let alone add a gazillion more options that I will forget are there.

I am happy with them for what they are, but they simply cannot replace a real player (and honestly, I am glad for this to save all of our musician friends).

Maestro2be

I don't think VSL is worried about anybody else "catching up".  I don't think they're going to do anything at all to keep competitors at bay  If they were worried about this, they would need to put out an expansion any time any competitor makes any library that's even closely related to an instrument VSL has recorded, since every new library offers something different.  We would have a new VSL expansion every three minutes and it would just be ridiculous.  Let's look at the facts though.  VSL's instruments are what, ten years old?  More?  And they're still heavy competition for the very newest libraries.  In that ten year span, VSL's instruments have been honed to perfection and their sampling methodology has been so fine-tuned that new instruments and software come out virtually bug-free.  Also VSL is far ahead of the competition just by releasing something like Dimension Strings.  I'm not sure if a competitor will catch up to that any time this decade.  Honestly I'd be running over to Orchestral Tools' website posting comments like "VSL is way ahead of Orchestral Tools--we need expansion packs!" because of DS.  But OT doesn't care what VSL is doing and vice versa because the libraries are so different and it would be the death of any developer to watch what everyone else is doing and try to match it.

As a side note, the list of expansions and upcoming products that OT has posted is in no way detailed.  It's a stretch to say that they're going to wipe out VSL with all these upcoming articulations when we don't really know what those articulations will be.  For example, Expansion B of Berlin Strings provides sul pont, sul tasto, harmonics, and col legno.  VSL already offers all of these in some way or another.  To be fair I'd love if VSL would provide these articulations with legato, but I can't remember the last time I've missed it.

@DG said:

OR's Agile legato sound like a cheap synth,

I think same after first listen to demos. And after about half year from release this library is unusable due to bad programming.

With VSL, the expansion pack is when you purchase the Extended, also labeled L2, edition.

@Rolf_Music said:

The problem to alot of the issues regarding "dynamics" and "conflicts with crossfading" would be solved if we had the option to "automate the in built eq on a matrix level" and if we had a "high pass filter". Including the option to change eq and filters on a patch level. I also think it would be great to introduce a melodyne similar feature which enhances vibrato without changing the patches speed. As a result we would have a VI Pro which is alot more powerful than it is now.

I actually have a few suggestions with screenshots which I will post later in a different topic.

IE: for a solo instrument I do not velocity xfade, I think there is zero point to it. For a section, reinforcements and cancellations of phase are what happens. So where it occurs in velo xfade it's somewhat related to reality.

I don't think VSL can't be surpassed but I don't think OT is where this is going to happen. In terms of winds, Sample Modeling is the next evolutionary step. I'm not a great strings maven, but I think modeling of instrument behaviors applies, but the sound of strings is a different problem than brass or silver.

And that you're reading the marketing drivel by 8Dio quite uncritically; you see them go on about how they've solved the crossfading issue (no), and then this 'vibrato' purely through these devices of pitch/LFO as in synthesizers is the way to go. It's not exactly revolutionary.

If I can't be persuaded by demos...

I know nothing from Orchestral Tools strings via experience. However I have Berlin Woodwinds. They sound good I think. But it is basically articulations loaded individually in .nkis and there are multis to give you keyswitching. And there is layering of that. Big fuggen deal.

And it's buggy, it won't even load here. I got Orange Tree Samples Mind Control to give me an instance with more than one articulation. I'm pretty skeptical of VSL needing to stay competitive with this vendor. The "Articulations Performer" multi is following VSL from a decade ago with Giga Performance Tool in a sort of half-assed implementation. And they have abandoned it in the strings (buggy?), so we have the argument now that loading single .nkis is superior because you want to do a lot of work with crossfading anyway, individually. Sure, let's go back to that and pretend VI Pro didn't happen. Get real.

Hello everybody, 

Of course we appreciate your personal opinion in this forum, but let´s try to keep insulting phrases, both on a personal level and about software products, out of this discussion. 

Everybody in this business is doing their best to follow what they feel is right, and the diversity makes the scene as interesting as it is.

Thanks, 

Paul

VSL is the best without doubt. After purchasing some of their collections I started asking myself: "Why didnt I purchase from them before? Lost alot of money through Kontakt libraries".

Anyways, just thought I would share what I was thinking :D

I still think that it would be kind of nice to see specialized packages for each articulation, specifically for strings as they are the most complex :D

I just wanted to add something here about the relationship between the complex issues of realism and sustained notes. It is on sustained notes that one is more likely to become aware of the vibrato in the sample repeating - which sounds unrealistic. Awhile ago I questioned whether it would be possible for VSL to create a vibrato equivalent of the global humanize function in VPro. The thread is here:

http://community.vsl.co.at/forums/p/36559/224706.aspx#224706

There wasn't much response to this and no word from anyone at VSL as to what they thought of it.

To recap, my suggestion was to start with a non-vibrato sample. The software could be triggered by any note long enough for the vibrato to be audible (starting at a couple of seconds?). The software would apply random types of vibrato to the single sample. The humanize software already includes five types of vibrato. This would mean avoiding the need to do more recording and create more samples, and the software would be added to VPro.

This isn't meant to cure all the issues to do with achieving realism but I think it would make a big difference - especially on a solo violin line.

I agree that the main problem are sustained notes on solo instruments with vibrato, especially strings. I am trying and experimenting around in order to get different results but I think Im getting very very close. Just need to work more on the realism.

Here one test where you see 2 versions of a Violin and Cello. One is the Raw Patch and the other always the modified version. Deliberately raised the bow noise but check how the vibrato is changing the entire time. I did alot of changes in the tone. Used patch combos etc.

http://vocaroo.com/i/s0fOZrkdSjvh

@Oceanview said:

The software would apply random types of vibrato to the single sample. The humanize software already includes five types of vibrato. This would mean avoiding the need to do more recording and create more samples, and the software would be added to VPro.

 

The problem I've always had with any kind of randomization functionality built into software, whether it's legato/vibrato varieties or quantizing notes (like in the old days), is that the software always chooses to use the most awkward and/or inappropriate note, start/stop time, velocity, etc. then I have to go back and re-edit everything over again anyway.  The reason for these inconsistencies is really simple: Software will never have any common sense. 

@ Rolf Music,

I'm not trying to pick a fight or anything as you sound like a sincere person who is just trying to write better music, like the rest of us, but it could be that you are writing music that is outside the scope of any sample library.  The costs associated with producing these "expansion packs" could be substantial to VSL to say the least and personally I don't think I would be interested in them because they would just over complicate my workflow.  Speaking from a marketing point of view I'm not sure there is enough of a demand for this to justify the expense.     

I agree on both things. Randomizers are bad and you are right absolutely right about the expansion packs idea. I noticed that is true, especially from a marketing point of view. I have to admit, you are absolutely right. However,  I somehow noticed that you can create all articulations you want. Im not just refering to the slot x fade patch combinations but also a rather surgical approach with automation using an equalizer by targeting all relevant areas.

In the first test which I shared here I was experimenting with bow noise and very long sustains. Also the second cello version was rather harsh sounding because I wanted to emulate a very strong bowing. Strong bowing techniques usually expose alot of overtones. The realism in the ears of professionals however can be questionable but by analyzing the frequency of a reference piece (or a few notes) you can actually reproduce a nearly authentic "extreme bowing sound" by properly exposing the corresponding  overtones.

I also have another experiment where I tried to figure out ways to create a strong and fat vibrato. I used portamento for this but every patch gives different results. I could create romantic vibratos, vibrato to gradually less vibratos and viceversa... the options are endless. Again all surgical procedures. Not much reverb.

Here the second test:

http://vocaroo.com/i/s1Oz4ehuMYb3

Again this can greatly be improved and im experimenting around and around. This experiment does not use velocity for the volume / dynamic change, neither any filters but just by automating a few EQ gain parameters yet keeping all frequencies 99.9% in phase.