Trendsetting Sampler Functions - let´s look into the Future

[GRUMP]

Hi All,

I have used lots of trendsetting Plugins in the last Months and checked out "unusual" ways for Sound Design. I don´t believe in the Vintage Hype and all these "modeled after" Tools.

I feel that the Future belongs to the Sampler. Tools like Iris2, MCharacter, MTransform - Spectral Editing/Filtering, Multi Band Splitters and Stuff like that - already Point into a really exciting Direction and these new Methods will make Samplers really mighty and superior Sound Design Tools in the next Years. AI (IZotope...) should also be mentioned in this context.

My Questions:
- Which Functions would you like to see in the next years and why?
- Which VSTs or HW does already offer trendsetting Possibilities you would highlight and recommend?
- Have you already seen a Tool that offers Sampler Faunctionalities and the Possibility to Split the Signal up into multiple Bands (that can ideally be processed individually)?

Greetings!

Grump(y)

[DSmolken]

As far as samplers, a lot of people seem unimpressed by the Kontakt 6 update, and Sforzando (the main platform I use) hasn't been actively developed much for the past few years, so there is a chance for someone. UVI Falcon seemed promising and powerful when it was released, but I haven't really seen anything that used it for heavily manipulated sounds. HISE seems to be picking up a bit of steam, though. Spitfire have their own engine, Play got updated, and OT are also working on their own engine, so there's also that side of it, where more developers of sample-based instruments might write their own samplers, too. But for the question you asked, HISE seems the most promising right now.

I really do think the future isn't so much in new samplers, as it is in better understanding stuff we're trying to model. Sample Modeling Strings and Neocymatics Violin Ensemble are the most interesting-looking things I've come across lately. Both use Kontakt, but that almost doesn't matter. Both milk a lot out of not-so-huge sample pools of only a few GB each (tiny for a modern string library) with extensive modeling of not just instrument behavior (like applying different body IRs to the same set of string samples) but more interestingly, modeling the behavior of humans making up the ensemble.

When (not if) somebody does that with a choir, and models, for example, how singers start a note slightly off then drift towards a common pitch, that will be groundbreaking. Identifying and modeling human behaviors. That I think will be the biggest growth area IMO.

[GRUMP]

I really do think the future isn't so much in new samplers, as it is in better understanding stuff we're trying to model.
When (not if) somebody does that with a choir, and models, for example, how singers start a note slightly off then drift towards a common pitch, that will be groundbreaking. Identifying and modeling human behaviors. That I think will be the biggest growth area IMO.

That is an interesting Point of View, for sure, but my Subject is not and never has been the Recreation of Real Instruments. I can´t say much about this and I have never tried Kontakt Librarys. Modeling Real Instruments is also a Topic for the Core-Funcionalities of a Sampler, the sampled Materials and its Combination. The Result is more or less something existing.

What I think about are Functions like Groove Agent´s "Decompose". The Sample is split into Body and Noise. Or Iris2 Spectral Filtering. You can deconstruct the Material and combine it with other Elements. Melda´s MB-Technology which opens a "Modulation Heaven" for formerly static Sample Content (very interesting for me, my Topic is Synth VOX and Pads).

I must admit - I assume that Parts of the traditional Sound Industry won´t like this Development because it gives an endless creative Potential to the User. That´s surely why no Sampler on the Market is capable of MB-Processing until now. But ... the first Steps are made already - Decompose is pioneering - and I think we should not underestimate the Power a "Granular Deconstruction". We have time, pitch and Formants today. If e. g. Bands and Spectrums follow the Synthesis Capabilities of the Sampler are bigger than they have ever been with what we call a Synthesizer. The Melda Plugins already pint into this Direction ...

[Tj Shredder]

One technology which did not move into commercial plugins is concatenative synthesis. I heard of one non-prominent exception, but that is for just orchestral sounds, forgot where I heard about it. Its developed at Ircam. I suspect some Max patches being way ahead of commercial sampling...

[whyterabbyt]

One technology which did not move into commercial plugins is concatenative synthesis.

well, except for the various vocal synths that use it. seems to be the main method in that niche.

[Tj Shredder]

Yeah, and Atlas comes to mind, not quite concatenative synthesis, but automatic analysis and AI aids will be more and more incorporated into sampling techniques...

[Obsolete236871]

For creative sampling / mangling of audio material I really like the direction that Serato Sample and it's standalone equivalent Serato Studio go. No cumbersome mapping and preperation of samples before you can start. Just load the sample and immediately start having fun with it. It has automatic detection of the key that the samples are playing in and you can shift it easily to any desired key. That makes it easy to tune the sampled material to the key of your song or to sample material from other sources. I had the trial version of Serato Studio and it was the most accessible, immediately usable and intuitive sample-centered "beatmaking" application I have ever used. Unfortunately Serato decided to release it as subscription-ware, which I don't support. But the plugin version is available as regularly sold product and currently on sale at Plugin Boutique.

[GRUMP]

For creative sampling / mangling of audio material I really like the direction that Serato Sample and it's standalone equivalent Serato Studio go. No cumbersome mapping and preperation of samples before you can start. Just load the sample and immediately start having fun with it. It has automatic detection of the key that the samples are playing in and you can shift it easily to any desired key. That makes it easy to tune the sampled material to the key of your song or to sample material from other sources. I had the trial version of Serato Studio and it was the most accessible, immediately usable and intuitive sample-centered "beatmaking" application I have ever used. Unfortunately Serato decided to release it as subscription-ware, which I don't support. But the plugin version is available as regularly sold product and currently on sale at Plugin Boutique.

A nice tool for OneShot-Sample-Mangling, OK. But that is noting new ... and represents Limits and not Progress i. m. O.

But as I just said in the Melda Thread: there are various Use Cases for Sampling - and a good Sampler should balance well between them..

[GRUMP]

As I just remark what I do every day is quiet special indeed. Imagine - I just have my MPC auto sampling the 2nd Layer of an additive (and already layered) VOX-Pad. What I want is the Band Split - second Layer is 2001 - 8000 khz, 1st is below and above. I´ll load it into Halion now, loop the first Layer and make a Grain-Zone from the 2nd. Afterwards I´ll make a Loop-Envelope for Modulating the Volume of the second Layer. That´s why I asked for a Sampler with Multi Band Splitter Capabilities ...

EDIT: Quick and dirty Demo ... This Pad was a Zebra one Day. Three Zebras to be true...

VIPstack2BandsQDDemo.mp3


But anyway ... watch the Market. We´ll experience severe Changes in the next years that will let us forget Waveforms and Wavetables quickly. Forget Vintage. The Target is cloning complex Spectrums and I don´t think that it will take long until we get there.

[GRUMP]

For creative sampling / mangling of audio material I really like the direction that Serato Sample and it's standalone equivalent Serato Studio go. No cumbersome mapping and preperation of samples before you can start. Just load the sample and immediately start having fun with it. It has automatic detection of the key that the samples are playing in and you can shift it easily to any desired key. That makes it easy to tune the sampled material to the key of your song or to sample material from other sources. I had the trial version of Serato Studio and it was the most accessible, immediately usable and intuitive sample-centered "beatmaking" application I have ever used. Unfortunately Serato decided to release it as subscription-ware, which I don't support. But the plugin version is available as regularly sold product and currently on sale at Plugin Boutique.

Question: don´t the most DAW-Sampler-Tracks offer these Possibilities anyway?

[Obsolete236871]

A nice tool for OneShot-Sample-Mangling, OK. But that is noting new ... and represents Limits and not Progress i. m. O.

Represents Limits? I don't know what you are trying to say here. Serato Sample is not really about creating one-shots, but about enabling uncomplicated and creative work with the audio material, regardless of the play-mode. You can play samples multitimbral, as well as loop samples. So, your "OneShot" remark really misses the point.

Question: don´t the most DAW-Sampler-Tracks offer these Possibilities anyway?

DAW sampler tracks? Could you elaborate on what you mean? Cubase has a function called "Sampler Track", but that's rather recent (I think since version 9) and is not a universal feature in other DAWs, but Cubase specific.

Unfortunately the feature set I've mentioned is not standard in sampler plugins. Most sampling plugins are not intended for the creative and intuitive work with audio material. Plugins like Kontakt and Sforzando have been mentioned, which I use as sample library playback engines and which are great for library creators, but are not really useful for creative and intuitive work with sampled material. An instrument like Iris is great for sample based sound design.

But when I talk about creative work with sample material, I mean quickly slicing up an audio sequence into smaller parts, which I then can edit, timestretch, tune to the rest of my song and re-sequence into something completely new etcetera. And all that without too much hassle and without clicking myself through dozens of menus.

That's exactly what I like about Serato Sample (by the way: I'm neither employed by them, nor do I receive any benefit from endorsing their product) - they clearly understand the workflow and know what sample based producers / crate digging kind of guys want in a sampling plugin. The combination of features that I found useful in Serato, I haven't found in other sampling plugins:

- useful automatic slicing / chopping of audio sequences with proper transient detection (obviously you can also manually slice a sample into smaller pieces)
- categorizes and color codes the sliced pieces according to their musical content (eg kick, snare, instrument bass sound)
- automatic key detection and global key setting (great, if you want to tune your sampled material to other elements of your composition, which is usually a quite fiddly, trial and error affair with typical DAW sample functions / plugins)
- high quality realtime independent stretching and slicing of individual sample slices

If you ever did some intensive creative sample work, you will understand how incredible useful and rare those almost trivial sounding functions are in sampler plugins. But you might have a completely different understanding and entirely different expectations, which is okay.

[GRUMP]

Represents Limits? I don't know what you are trying to say here. Serato Sample is not really about creating one-shots, but about enabling uncomplicated and creative work with the audio material, regardless of the play-mode. You can play samples multitimbral, as well as loop samples. So, your "OneShot" remark really misses the point.

Din´t you say that it can not use Multisamples? So what else should it be then but a "One Shot" Sample Mangeling Machine? I mean ... as I already said before: it all depends on the "Use Case". If that is what you want to do it may be a good Tool. Butr me - personally - I wouldn´t even consoder it. My Sounds are usually based on Multisamples and several Layers in the End.

Question: don´t the most DAW-Sampler-Tracks offer these Possibilities anyway?

DAW sampler tracks? Could you elaborate on what you mean? Cubase has a function called "Sampler Track", but that's rather recent (I think since version 9) and is not a universal feature in other DAWs, but Cubase specific.

As I just said: I wouldn´t even consider these Tools and waht I know about them from the Marketing is not much. But as much as I know no serious DAW comes without such a Tool anymore.

Unfortunately the feature set I've mentioned is not standard in sampler plugins. Most sampling plugins are not intended for the creative and intuitive work with audio material. Plugins like Kontakt and Sforzando have been mentioned, which I use as sample library playback engines and which are great for library creators, but are not really useful for creative and intuitive work with sampled material. An instrument like Iris is great for sample based sound design.

Ok +1 for Kontakt. NI wants to sell and nothing else. Their Strategys seem to model the User as a dumb Consumer. Or maybe a "Musician".

But when I talk about creative work with sample material, I mean quickly slicing up an audio sequence into smaller parts, which I then can edit, timestretch, tune to the rest of my song and re-sequence into something completely new etcetera. And all that without too much hassle and without clicking myself through dozens of menus.

Sorry, but that is simply not true. I already did that intensively with my ASR10. What about the MPC? You´ll say Halion is too complicated, but I can tell you: it´s really easy with Halion. Believe me. I´m quiet sure that there are +20 Tools on the Market that offer such Opportunities, but maybe focus more on Possibilities than on toime Consumption and Easyness?

Furthermore: we´re not talking about the same Thing. I´d like to discuss Boders and Opportunities. Tools that enable you to make ceratin things quick and easy may be nice (for EDM ...), but they have severse Restrictions. Otherwise it wouldn´t be easy.

That's exactly what I like about Serato Sample (by the way: I'm neither employed by them, nor do I receive any benefit from endorsing their product) - they clearly understand the workflow and know what sample based producers / crate digging kind of guys want in a sampling plugin. The combination of features that I found useful in Serato, I haven't found in other sampling plugins:

If you ever did some intensive creative sample work, you will understand how incredible useful and rare those almost trivial sounding functions are in sampler plugins. But you might have a completely different understanding and entirely different expectations, which is okay.

As I already said - "a long, long time ago" ... and I´m used to do everything myself (Slicing). It´s easy and often very inspiring. And ... oK - you like Serato - but the mentioned Functions are nothing but trivial. Simply basic Functionalities - in a nice Package (Colors and so on). People do such Stuff for more than 30 yrs. now and nothing will make me believe that Serato and Co will tell us much about the Future of Sound Design (what has a lot to do with Harmonics for me .

[GRUMP]

And an other Example for the creative Power of Band Splitting ...https://bluelab-plugs.com/?product=air

[funky lime]

That is an interesting Point of View, for sure, but my Subject is not and never has been the Recreation of Real Instruments. I can´t say much about this and I have never tried Kontakt Librarys. Modeling Real Instruments is also a Topic for the Core-Funcionalities of a Sampler, the sampled Materials and its Combination. The Result is more or less something existing.

I think maybe you missed his point a little bit. It's not just about recreating real instruments, it's about recreating actual human performances of those instruments through use of clever scripting and circumnavigating the limitations of a traditional sampler. The idea is that, ultimately, you will be able to "perform" the sampler with the level of control that an instrumentalist would have over the actual instrument modeled, which is way different than 99% of orchestral sample libraries where the performance/expression is more "baked into" the samples and the composer only really has to worry about dynamic crossfades, vibrato, and volume/expression.

The newer tech, on the other hand, lets you get into the finer details, like bow position/pressure/direction, rosin/bow noise, resonance of the instrument's material, which string to play on, etc, so that you can actually model/program a "performance" instead of just sequencing/playing a traditional violin sample library where what they recorded is basically what you get.

It may not be trendsetting for the kind of music you make, but I think it's pretty new and exciting for purveyors of virtual orchestral music. It's about clever people pushing the limits of sampling as we have known them, which does seem to align with what you're talking about.

But then, maybe that's more rompler territory, now that I think about it.

[DSmolken]

Yup, stuff like controlling bow noise, or deconvolving violin samples to remove the "color" of a particular instrument and then reconvolving them with different violins' IRs - that's stuff in the Sample Modeling strings which is being used for realism but could also be used for weird out-there sounds. A thing that I haven't seen used yet but could be - what I mentioned about singers in a choir drifting towards a common pitch. Use non-music flocking algorithms to control detune, and beyond choir singers drifting towards a common pitch, you suddenly can start making unison a lot more interesting than it's ever been before.

A lot of our favorite sounds started as ways of making earlier sounds which didn't quite work out. Roland tried to make the supersaw a better synthesized string ensemble, and unintentionally made a bright, powerful lead that's stuck around for decades.