Discussion in 'SAMPLE Talk' started by Mystic, Jan 11, 2018.
Those are already a dime a dozen usually for hip hop and trap.
Less gigabytes and less resources.
Well, if that doesn't sound exactly like my post in the "diminishing returns" thread! Lol
We have some good sounding tools by now. I think it is quite impressive what we can already do with sample libraries.
In my opinion it is the usability, a streamlined workflow using those tools that needs to see a quantum leap forward. AU3, expression maps in every major DAW, really stable and self explanatory ways to set up templates. All those acrobatics around Midi channels, banks, sampler plugins, external plugin hosting solutions, articulations switchers, and don't forget not to load a certain plugin before another one... now it is 2018 and should we really have to spend this much time and effort with these creativity killers, just to get up and running (and plenty of times in between)...? That's what I wish to change before I need a lot more new sounds...
Ok, this may not be the next big fad, and it is not a sample library, but...
A fast (using GPGPU) wavetracer (accurate physical modeling similar to 3D raytracer).
Now, I know there's not always much love for physical modeling (although it matches with the hybrid fad), but lets think about it:
- A wavetracer does not 'fake' the sound with fancy equations. The air displaces like in the real world and the surfaces react and vibrate like in the real world (hence the need to use GPGPU to calculate an immensely huge amount of trigonometric data).
- One could use it in conjunction with a CT scan (or similar) to map a real Stradivarius and get its 'real' sound.
- The output could be 100% dry, but one could also scan the room (same laws, same engine) and get the 'real' live sound without reverb plugins.
- If needed, the output could be 'rendered' to allow 'low-end' machines to run it as sound samples, while allowing the big guys to play with a full wavetraced orchestra in real time (maybe using some Tesla P100).
- A performance AI assistant could improve realism and playability, maybe using machine learning algorithms to 'learn' from real players how an instrument is played (this can also be applied to current sample libraries).
All this can be done today (the tech already exist), so... Why not?
Depends on what Hans does in his next score.
Misquoted for truth.
No libraries above 200 USD..
One can always hope...
Played by Dolly Parton.
‘Quiet Desperation’ libraries came to mind.
I'd like to throw "wrapper plugins" in with that, complete with "mixer within a plugin contained within a wrapper plugin."
So you have exhausted the RBB library!
WaveSkimmer - libraries are the NEXT BIG THING!
You heard it hear FIRST!
I would think most if not all of this could be done now at least at the level of research projects. So your string library can be "played" by emulated performers in a user defined manner"[/QUOTE]
Sort of where Apple are going with their AI drummer and Percussionists… Obviously a Bass player is imminent given the iOS app does it already…
A great "Steel Guitar" put out in Kontakt by kalago.
Sort of where Apple are going with their AI drummer and Percussionists… Obviously a Bass player is imminent given the iOS app does it already…[/QUOTE]
String articulations are hard to understand, even by humans. The "user defined" aspect would have to be very advanced. It's always the "dead giveaway".
Have you ever noticed the sound of so many libraries sound like sample libraries?
There's a certain tone to many sample libraries, but they tend to sound like leftovers from somebody's project.
Or they used the same mic in the same room in the same position through the one preamp for everything.
Many sample libraries are the aural equivalent of buying tyres with 90% tread.
I don't care what they release, as long as it doesn't sound like a sample library.
Decent accordion libraries are rare nowadays. Even the existing ones are short in possibilities, they all have some lacks and shortcomings. Modern well trained accordionists play the modern bayan now, which stands so far from the popular folk accordion (in technique and sound!). They all have a double discant (full scaled keyboards for both left and right hand) and most of them are 'convertible': they can switch from (left hand) bass/chords to full keyboard like the right hand, which makes 4 voiced counterpoint possible.
So far I haven't seen any library which can do that. I sometimes write for accordion/bayan and I have to use two right hands to simulate the modern accordion. Hardly imaginable in a time where libraries offer such quality!
Well, that all depends on your perspective. In your case, yes, I can definitely understand where you're coming from. I'm surprised a modern bayan with full range free bass hasn't been sampled yet, given the popularity of the instrument.
Now that you mention it, Dominus Choir does sound a lot like Shreddage 2.
But seriously, one possible explanation is that is many sample libraries are recorded raw so that you, the producer, can process and mix them as required by your project. That, and your skills as a MIDI composer, is where the magic happens. If you drop libraries into a mix without creatively seating them, everything will sound lackluster. You still must treat a sample library as if it were a recorded instrument.
This is so true, putting all the bells and whistles on a library instrument will probably leave little room for diversity.
Of course recorded raw samples, as Polkasound says, will require more FX, and controllers, which makes things more complex and can require considerable knowledge and skill to produce midi. Not to mention the patience to examine the manuals thoroughly which can take a lot of time.
I think developers think about this a lot, trying to decide how far to go and how complex their library(s) should be. They have to depend on their users to have the ability, patience, and desire to explore and learn what makes the library click.
There is (was) one exception: The Hohner Gola accordion (Best Service), which was like a convertible (free bass), but with few registers and some scripting problems (basic volume resets after every shut off and at every restart it is zero).The automation can't change that, the helpdesk stated that it simply the way it is... So hardly usable.
Separate names with a comma.