Menu
What's new
By:
Filters

Are we REALLY in the diminishing returns zone now?

Lee Blaske

Lee Blaske

Senior Member
I know I've said this before, but I wonder if we've really reached the point of diminishing returns when it comes to sample library development. Are there anymore major breakthrough libraries on the reasonably immediate horizon, or are we kind of in a holding zone where there will be incremental improvements/differences, and developers putting out products that in some way/shape/form imitate the products of others (e.g. low dynamic libraries, textural libraries, extended technique libraries, etc.)?

It just seems to me that a number of the products we had been awaiting from major developers that have recently been released are not the game changers some of us thought they might be.

I bought VSL Synchron Strings I during the early bird period. With all of the videos about the Synchron stage, the players, the technology involved, and the fact that VSL has been in this business for years, I thought it would be a MAJOR step forward. An earth shattering improvement over the rest of the field. Now that I've had my hands on it for awhile, I think it's a nice library, but in a lot of cases, I believe I already have a number of options that are better.

Same goes for Hollywood Choirs, which I'm on the fence about buying. It just doesn't sound that different to me than the original. I don't see it being capable of singing a jolly Christmas carol, or a serene, touching, lovely choral piece. It seems that it's still mainly for that serious choir sound in ponderous film and trailer music.

I'm looking forward to the Eric Whitacre library, but I fear that might be a disappointment, too (because I'm not sure Spitfire really wants to invest the time in making a word builder). Time will tell...

Maybe I'm missing some things, but it seems to me that we're not seeing truly amazing, jaw-dropping products anymore. For this past year, the one exception I can think of to that might be Superior Drummer 3. I do think that product really took a giant step forward. It is, however, a product that is so deep that I don't personally think I have the time to invest in it to really use it to full potential (and actually, the way I work doesn't really exploit a lot of the amazing things it can do).

There are products that come around that fill useful gaps, and are cool. I'd list Realitone's Screaming Trumpet as one of those. Not regretting that purchase.

And FWIW, some of the features that developers tout as major sell points just don't seem to be a big deal, at least to me. I don't need 23 mic positions and every surround sound format known to man. I'd rather just have a couple positions that sound REALLY GOOD. If somebody needs to put something in a a unique space, there are endless plug-ins for doing that.

What might be the next major thing to happen that will cause all of the other developers to scramble to catch up?
Click to read more...
 
Last edited:
I agree that we're seeing a lot of the same old (which isn't necessarily bad thing IMO), but there are some developers that are still very innovative and are pushing the boundaries. For me those are first and foremost Performance Samples and Musical Sampling. I think (and hope) that their focus on rather small but really playable libraries, that are fantastic for specific purposes, will be something that we'll see more and more in the future. I'm really tired of those giant libraries with a gazillion mic positions as well.
And then there is of course physical modeling, like Sample Modeling and SWAM. Advances in physical modeling might be the biggest technical breakthrough on the horizon, even though I still don't think that they will ever completely replace traditional samples.
 
lucor said:
I agree that we're seeing a lot of the same old (which isn't necessarily bad thing IMO), but there are some developers that are still very innovative and are pushing the boundaries. For me those are first and foremost Performance Samples and Musical Sampling. I think (and hope) that their focus on rather small but really playable libraries, that are fantastic for specific purposes, will be something that we'll see more and more in the future. I'm really tired of those giant libraries with a gazillion mic positions as well.
And then there is of course physical modeling, like Sample Modeling and SWAM. Advances in physical modeling might be the biggest technical breakthrough on the horizon, even though I still don't think that they will ever completely replace traditional samples.
Click to expand...
I entertained high hopes of physical modelling becoming the way forward when I bought my Yamaha Vl1 in 1996. I still use and love this instrument as much now as I did back then. Highly expressive with natural envelopes, and very musically satisfying. However somehow, listening to recent PM products, I still struggle to find any improvements in what they achieve in terms of emulations of real instruments beyond their well documented expressiveness. On the sampled front, there have been, and will continue to be vast improvements in the solo instrument arena. Huge leaps forward have been achieved by Embertones J Bell instrument, and Vir Harmonics, Bohemian Violin, for instance, rendering previous solo violin offerings all but obsolete. Listening to the achievements of the likes of Thomas Bergesen's compositions dating back from the EWQLSO (2009) virtual instruments have been capable of so much for some time already. Better use of the incredible tools we have access to will remain where personal achievement will lie. Great times to be alive!!
 
I think the only real development that can be made with samples is to have better algorithms to detect the playing and then put the correct samples together to form the desired articulation, phrasing, etc.

Other than that, I think its up to physical modelling but that probably won't be great for a very long time and won't come from companies like Spitfire.
 
I think the next leap forward will be implemented with Artificial Intelligence. We're already seeing hints of its potential in effects products such as Adaptiverb and Neutron. It shouldn't be long before it's brought to bear in virtual instruments.

Best,

Geoff
 
It's simply a question of expectations. A few years ago a lot of features were not available. Things like multiple room mics, legato, dynamic control by CC, round robin, release samples etc. Now they are. Until Wallanderinstruments and later Samplemodeling I was searching for useable pop/jazz brass for decades. Now the search is over. If I find something better or easier or just different for variation I'll get it. But I don't feel that as a 'diminishing returns zone'. I don't want to be entertained by new published jaw-dropping libraries. I want to use them. I'm really happy that a lot of innovation happened in the virtual instrument world during the last years. I'm in the comfort zone right now.
 
Last edited:
Both Sample Modeling and Swam have used a small base of samples plus extensive scripting.
Swam-S, for the solo strings, abandons samples altogether, and the results are tonally disappointing, as are the strings in Wallander (Noteperformer) and Synful.

So?
 
There are lots of rote tasks the composer has to do that can be automated and there are lots more musicality details that can be improved and can become standard. Many of these are about “interpretations” and having the variability of sounds to enable those interpretations and the automation of that Work.

Instrument attack and release variation
Phrasing
Orchestral balancing
Greater layering and cross-fading
Automating room / reverb /placement / inter-instrument resonance
Modelling of Ensembles
All aspects of orchestration, mixing and mastering

Also anything involving singing is pretty primitive at the moment.

There is still much more that can be done with current practices of sampling and scripting although some of this would require some fundamental architectural changes away from note-by-note Midi processing in DAWs and VSTs.

The new technologies of machine learning, AI and increasing sophistication of physical modelling offer huge opportunities for improvements also but it is not really clear that there are business models that would support the kind of development that those would take at the moment given the investment involved.
 
Last edited:
Ark 3 is a game changer.

Musical Sampling has done some incredible stuff too.

CSS seems to be the string library that a lot of people always wanted.

ProjectSAM has also done some really unique things.

I dont think we'll ever see a major leap forward all at once. It will be a gradual, evolving process. Thats just how technology works, and its pretty amazing how far weve come in the last 10-20 years.
 
Last edited:
A choir that can take lyrics input in plain English or Latin, and will quickly produce intelligible results, would be a massive improvement over what's out there.
 
I will agree with Mike, Ark 3 is a new approach and really added something unique this year (I do demos for OT). It is almost a must buy.
Strikeforce with it's implementation and simplicity made programming drums so intuitive it is really hard to make it sound bad (I don't do demos for them). Last years Modo Bass which seemed to come out of nowhere is a terrific physical modeling plug in. Superior Drummer 3 as mentioned above is a crazy good plug but I will likely never use most of what it can do.
 
To address the OP - quite possibly. The rate of progress has slowed, for sure.
But to flip the question around a little....how many musicians and composers can actually use their DAWs and libraries to maximum effect? I'd suggest that in many cases, it's us - the composers - who need to catch up..
 
mikefox789 said:
ProjectSAM has also done some really unique things.
Click to expand...

ProjectSAM is an interesting company. I like the way they continue to support their products and add value. I've got Symphobia 1, and I had some of their earlier brass libraries (didn't upgrade in time, though, so I can't use them now). Whenever I see they have a sale, I consider buying some of their other products. But, the thing (at least for me) that seems impressive about ProjectSAM libraries are those huge multis where you're playing the whole orchestra at once. I think that would be great for live, but it seems to me that if you're not live, you'll always get better results playing and controlling all of those individual elements separately (and I've already got all of those things covered in multiple libraries).
 
You must log in or register to reply here.
Top Bottom