Ideas for new sample libraries

Everything and anything, but nothing about the LinuxSampler project.
User avatar
dahnielson
Moderator
Posts: 632
Joined: Wed Jan 23, 2008 11:25 pm
Location: Linköping / Tranås, Sweden
Contact:

Re: Ideas for new sample libraries

Post by dahnielson » Wed Mar 26, 2008 3:21 pm

Consul wrote:And here I am, the exact opposite. I want a sampler that allows me new and different ways to tweak, mangle, and otherwise process samples to make new sounds. My big inspiration has been the videos on the Omnisphere from Spectrasonics. Other than that, though, I do want some more traditional functionality, and maybe some new ways to make realistic performances of acoustic sounds (drums are one example).

Maybe the solution to this is going to be two separate samplers.
No. You are confusing the data with the machine.
Anders Dahnielson

Ardour2, Qtractor, Linuxsampler, M-AUDIO Delta 1010, Axiom 61, Korg D12, AKAI S2000, E-MU Proteus 2k, Roland R-5, Roland HP 1300e, Zoom RFX-1000, 4GB RAM x86_64 Intel Pentium Dual 1.80GHz Gentoo Linux

User avatar
Consul
Moderator
Posts: 189
Joined: Wed Jan 23, 2008 11:19 pm
Location: Port Huron, Michigan, USA
Contact:

Re: Ideas for new sample libraries

Post by Consul » Wed Mar 26, 2008 3:52 pm

I don't think I am at all. In order to do the sample mangling I would like, certain features will have to be coded in. Those features are quite different than what's necessary for the microsample synthesis you guys are talking about. Though to be fair, microsample synthesis (look at me, I coined a buzzword!) does present its own interesting mangling and sound design opportunities.

One solution might be a modular architecture, but I somehow doubt that is going to happen. Another solution might lie within my brainstorming on making features from more atomic functions, but again, I seriously doubt that will ever happen, unless I learn to do it all myself.

And if I'm coming across as being terribly cynical, I can only say you're quite right.

I'm going to work an a mock-up of how I would like a sampler to look and behave, and if folks want, they can either chime in or tell me to get stuffed. Hell, anything would be better than the next-to-nothing I've received so far. I have a feeling, though, that I will simply continue to be talking to myself. I hope I'm wrong.

It's okay to tell me my ideas are bad ones.
Darren Landrum

User avatar
dahnielson
Moderator
Posts: 632
Joined: Wed Jan 23, 2008 11:25 pm
Location: Linköping / Tranås, Sweden
Contact:

Re: Ideas for new sample libraries

Post by dahnielson » Wed Mar 26, 2008 5:12 pm

First, it's true my interest is mainly vested into sampling of actual instruments. So I have a bias towards tools that will help me accomplish that. But those needs doesn't need to define limits of what the tools can be capable of.

Second, microsample synthesis, like in Tommasini, can be successfully implemented in Kontakt 3 or any environment with similar capabilities.

Third, you are correct that to be able to perform Omnisphere-like magic certain features need to be in place, like the ability to perform granular synthesis. But that can very well be inside the scope of a regular sample environment. E.g. if you take a look at my discussion of "next generation" design, you will find that such synthesis can be implemented as new "sample player" and "filter" blocks.

Fourth, you are absolutely correct that the most expedient way to make things happen is to do them yourself. Scratch your own itch.

8-)
Anders Dahnielson

Ardour2, Qtractor, Linuxsampler, M-AUDIO Delta 1010, Axiom 61, Korg D12, AKAI S2000, E-MU Proteus 2k, Roland R-5, Roland HP 1300e, Zoom RFX-1000, 4GB RAM x86_64 Intel Pentium Dual 1.80GHz Gentoo Linux

User avatar
dahnielson
Moderator
Posts: 632
Joined: Wed Jan 23, 2008 11:25 pm
Location: Linköping / Tranås, Sweden
Contact:

Re: Ideas for new sample libraries

Post by dahnielson » Wed Mar 26, 2008 6:04 pm

frink wrote:I just listened to the violin you refer to. This is very similar to stuff I've heard from Big Fish and London Solo Strings. I think the Stratavari tops it but very good.
Just to clarify: I took a look at Big Fish Audio "London Solo Strings" and from what I see it's just a regular library containing some articulations while missing others. Nothing wrong with a great amount of articulations and variations. The phase aligned approach of sample based modeling only take a step further than relying on articulations to solve the problem with transition.
Anders Dahnielson

Ardour2, Qtractor, Linuxsampler, M-AUDIO Delta 1010, Axiom 61, Korg D12, AKAI S2000, E-MU Proteus 2k, Roland R-5, Roland HP 1300e, Zoom RFX-1000, 4GB RAM x86_64 Intel Pentium Dual 1.80GHz Gentoo Linux

User avatar
dahnielson
Moderator
Posts: 632
Joined: Wed Jan 23, 2008 11:25 pm
Location: Linköping / Tranås, Sweden
Contact:

Re: Ideas for new sample libraries

Post by dahnielson » Fri Mar 28, 2008 1:41 pm

Just to make a point clear. I'm not trying to shot someones idea down or otherwise poo-poo-ing it. I'm all for recording a lot of samples. The more samples made available the better and even better if it's high quality samples recorded under documented conditions. Sample gathering is an enormous task always welcomed no matter what sample playback or synthesis/resynthesis will be used, because that's two separate things.
Anders Dahnielson

Ardour2, Qtractor, Linuxsampler, M-AUDIO Delta 1010, Axiom 61, Korg D12, AKAI S2000, E-MU Proteus 2k, Roland R-5, Roland HP 1300e, Zoom RFX-1000, 4GB RAM x86_64 Intel Pentium Dual 1.80GHz Gentoo Linux

User avatar
Consul
Moderator
Posts: 189
Joined: Wed Jan 23, 2008 11:19 pm
Location: Port Huron, Michigan, USA
Contact:

Re: Ideas for new sample libraries

Post by Consul » Sun Mar 30, 2008 5:26 am

http://www.sparkbangbuzz.com/flame-amp/flameamp.htm

This just opened up a whole new slew of samples I would like to record now. :D Definitely check out the video. That same mad scientist (who, by the way, is Nyle Steiner, inventor of the Steiner Electronic Wind Instrument, or EWI) has a variety of other interesting noise-making electronics and physics projects I'd like to build and sample, like this one:

http://home.earthlink.net/~lenyr/alsounds.htm

Listen to the sound sample. :)
Darren Landrum

User avatar
dahnielson
Moderator
Posts: 632
Joined: Wed Jan 23, 2008 11:25 pm
Location: Linköping / Tranås, Sweden
Contact:

Re: Ideas for new sample libraries

Post by dahnielson » Sun Mar 30, 2008 12:47 pm

Consul wrote:http://www.sparkbangbuzz.com/flame-amp/flameamp.htm

This just opened up a whole new slew of samples I would like to record now. :D Definitely check out the video. That same mad scientist (who, by the way, is Nyle Steiner, inventor of the Steiner Electronic Wind Instrument, or EWI) has a variety of other interesting noise-making electronics and physics projects I'd like to build and sample, like this one:

http://home.earthlink.net/~lenyr/alsounds.htm

Listen to the sound sample. :)
Cool.

Yeah. And if you plan to be up against Omnisphere then you have at least 50-60 GB of samples or so to record and edit, man. :mrgreen:
Anders Dahnielson

Ardour2, Qtractor, Linuxsampler, M-AUDIO Delta 1010, Axiom 61, Korg D12, AKAI S2000, E-MU Proteus 2k, Roland R-5, Roland HP 1300e, Zoom RFX-1000, 4GB RAM x86_64 Intel Pentium Dual 1.80GHz Gentoo Linux

User avatar
Consul
Moderator
Posts: 189
Joined: Wed Jan 23, 2008 11:19 pm
Location: Port Huron, Michigan, USA
Contact:

Re: Ideas for new sample libraries

Post by Consul » Sun Mar 30, 2008 3:17 pm

Yeah, I'd better get a move on, I suppose. :D
Darren Landrum

User avatar
dahnielson
Moderator
Posts: 632
Joined: Wed Jan 23, 2008 11:25 pm
Location: Linköping / Tranås, Sweden
Contact:

Re: Ideas for new sample libraries

Post by dahnielson » Sun Mar 30, 2008 3:53 pm

Anders Dahnielson

Ardour2, Qtractor, Linuxsampler, M-AUDIO Delta 1010, Axiom 61, Korg D12, AKAI S2000, E-MU Proteus 2k, Roland R-5, Roland HP 1300e, Zoom RFX-1000, 4GB RAM x86_64 Intel Pentium Dual 1.80GHz Gentoo Linux

User avatar
frink
Newbie
Posts: 8
Joined: Tue Mar 04, 2008 9:52 pm
Location: Borneo

Re: Ideas for new sample libraries

Post by frink » Sat Aug 30, 2008 12:03 pm

Okay sorry for my long hiatus. A move, family, friends and new studio construction come first. I'm glad that I generated some interest with my previous longish post. Now I will attempt to address a few of the replies. Basically my whole point is that we need more sample than what is provided.

Consul: You exactly understand what I'm after: better sampling. Record more stuff and you get a more accurate sound... I'm not against the features that you're talking about either either. I think it would be wonderful to have more synthesis capabilities in a sampler so that sampling and synthesis can be mix. In fact, historically that has produced the best sounds all around. So I'll definitely vote for everything you suggest in that arena. While we are requesting feature that may be considered the opposite sides of the spectrum. Ultimately we want the same thing: better sampling with more options.

dahneilson: You don't get me at all. Most of this reply will be directed at your comments...

"Capturing the performance" - I don't mean simply grabbing the sound that comes out of the instrument or the sound of the microphones, hall etc. I'm talking about the articulation that is unique to a performer and her instrument at the precise moment of recording. When I talk of capturing the performance I'm talking about capturing as much as possible about the way the sound is achieve so as to create articulation rich performances in the future. Violin portamento is a good case in point. Currently I don't know of a sample library that allows you to choose samples to slide between notes. I'm stuck with synthesis which often mangles the sound. Many instruments change the sound when they bend a note. We either have to model this through synthesis and/or provide original samples. I'm talking aobut providing as much in the way of original samples.

"Composers don't know how to play" - For example, most composers cannot play the oboe. If they can it is likely that their ability is less than that of a virtuoso oboe player. Even if their skill were perfect it would be limited by their own physics. Male wind players typically sound different from female wind players. What I propose is a way to pull enough articulations out of the instrument in a sampling session as to recreate their style. Certainly it takes someone much more skilled than a monkey to produce quality performances from these sample. (perhaps an elephant could do it it the elephant can paint - LOL) Playing sampled instruments is an art to itself. But as much as possible we want to make it easy to produce these sounds so that even if a monkey handed composer plunks out a melody it will provide a realistic performance.

My original recommendations stand: millions rather than thousands of samples allowed for a single library file. (granted the chance that a sound architect would put millions in their file is rather low.) Complex scripting that is contextually aware of what has been played before. (I should be able to script when a flute player is loosing his wind to produce the proper sounds.) Certainly there is a lot of art that goes into sampling. It takes someone with a trained ear and a lot of technical understanding to translate the studio performance of one interface (the original instrument) to a new interface (the keyboard).

On breath control I was showing the need for complex filtering. If you play flute (the breathiest instrument) even to a small degree you will understand that it is very difficult to get good wind articulations such that a keyboardist could play a flute patch and get anywhere close to creating a believable solo performance. It's okay in the mix but in actuality it's pretty dinky compared to a goo musician. The amazing part is that flute is one of the easiest instruments to model through synthesis. But capturing the performance nuances of articulation are near impossible without hours of work to produce a single track. In the end it's still easier to learn to play a flute!

I speak not from the perspective of the programmer that I am but from the perspective of composers who use these samples in lieu of original musicians for any one of a number of reasons. These composers need a much larger sample base to produce their compositions. Many of these dream of the day that they can export a file from Finale and have it rendered to a near perfect symphonic performance. There are two things stopping this dream from becoming reality. One is the limitations of sampling and the other is the lack of articulation vectors provided by MIDI. (The limitations of MIDI are well known to anyone who deals with acoustic instruments regularly...)

Hopefully this explains again and underlines the necessity of changes to the sample library format.

-FRINK

Post Reply