KVR Audio

kuniklo · Post by **kuniklo** » Fri Sep 26, 2008 4:35 pm

It seems that software instruments are powerful enough now that you can achieve just about any sound you can imagine in very high quality with them. However, the interfaces still mostly mimic traditional synths and samplers and are fairly slow and tedious means of exploring the sound universes of these tools. Isn't the next step completely rethinking the UI of these tools? Why not try something radical and put all that CPU under our hoods to better creative use? The evolutionary patch approach in the Nord G2 is a step in the right direction but there's so much more that could be done.

Benedict · Post by **Benedict** » Fri Sep 26, 2008 4:59 pm

I think that the breakthrougs we need are in

1. Processing Power - so we can be less worried about running out of notes etc.

2. Integration - VST is great but the way it integrates into Hosts is not really well sorted yet. This may be an issue of User Interface more than software protocols.

3. Interface - the hardware paradigm tends to stick because it appeals to us. What I think we need is coming from the ditection of things like Novation's Automap but that (for good reason) is only part of the way. I think if we had a kind of morphing hardware interface that made genuine representation of the VST in hardware then it would be the perfect interfaceof hard and soft. To use existing paradigms imagine having a project with a VST Jupiter 8 and VST Moog and as each instrument takes focus the controller instantly looks just like it with all the knobs positioned correctly. A fair stretch I know but that wold be a great step forward.

Sorting 1 & 2 will take us a long way alone. 3 wold be magic (literally I think)

kuniklo · Post by **kuniklo** » Fri Sep 26, 2008 5:05 pm

I think all of these things would be good in the short term but I'd like to see some more radical approaches.

If you think of a synth's sound world as a multidimensional space, with each parameter as an axis, and points in that space representing sweet spots, or interesting sounds, there are a lot of ways you can imagine exploring that space that don't involve manually manipulating each parameter one at a time. Current randomizers are a very crude way of jumping to random points in that space but you can imagine lots of other ways to move around in there.

glokraw · Post by **glokraw** » Fri Sep 26, 2008 5:52 pm

kuniklo wrote:It seems that software instruments are powerful enough now that you can achieve just about any sound you can imagine in very high quality with them. However, the interfaces still mostly mimic traditional synths and samplers and are fairly slow and tedious means of exploring the sound universes of these tools. Isn't the next step completely rethinking the UI of these tools? Why not try something radical and put all that CPU under our hoods to better creative use? The evolutionary patch approach in the Nord G2 is a step in the right direction but there's so much more that could be done.

The WusicVM to select any of your synth collections sundry sounds from one browser is
great. The simple multiFX/multi-timbral setup of IK Sonic Synth 2 etc is wonderful. The 16 part multi-timbral setup of Zynaddsubfx is also childsplay. Synth1 color selection for text and backround is nice. ProteusVX has a 'type' button, that splits the 1024 sounds into subgroups the preset arrrows access. The numeric values displayed in Zebra make precision easy when you want or need it. Along with skins, such things make it easier and more enjoyable to use the instruments. Getting all those in one great sounding synth would sure be sweet. Off topic, but some hosts would benefit from obvious 'Load Vsti' and 'Load Effects' buttons, instead of disguising the most important things in some twisted sicko gui

koalaboy · Post by **koalaboy** » Sat Sep 27, 2008 1:13 am

Benedict wrote:I think that the breakthrougs we need are in

1. Processing Power - so we can be less worried about running out of notes etc.

2. Integration - VST is great but the way it integrates into Hosts is not really well sorted yet. This may be an issue of User Interface more than software protocols.

3. Interface - the hardware paradigm tends to stick because it appeals to us. What I think we need is coming from the ditection of things like Novation's Automap but that (for good reason) is only part of the way. I think if we had a kind of morphing hardware interface that made genuine representation of the VST in hardware then it would be the perfect interfaceof hard and soft. To use existing paradigms imagine having a project with a VST Jupiter 8 and VST Moog and as each instrument takes focus the controller instantly looks just like it with all the knobs positioned correctly. A fair stretch I know but that wold be a great step forward.

Sorting 1 & 2 will take us a long way alone. 3 wold be magic (literally I think)

1) Track freeze is a technology that fixes this.

2) Get rid of Steinberg & Midi.

3) This does need a lot of work, but people are obsessed with pretty over usability, generally.

Novata · Post by **Novata** » Sat Sep 27, 2008 1:26 am

koalaboy wrote:
Benedict wrote:I think that the breakthrougs we need are in

1. Processing Power - so we can be less worried about running out of notes etc.

2. Integration - VST is great but the way it integrates into Hosts is not really well sorted yet. This may be an issue of User Interface more than software protocols.

3. Interface - the hardware paradigm tends to stick because it appeals to us. What I think we need is coming from the ditection of things like Novation's Automap but that (for good reason) is only part of the way. I think if we had a kind of morphing hardware interface that made genuine representation of the VST in hardware then it would be the perfect interfaceof hard and soft. To use existing paradigms imagine having a project with a VST Jupiter 8 and VST Moog and as each instrument takes focus the controller instantly looks just like it with all the knobs positioned correctly. A fair stretch I know but that wold be a great step forward.

Sorting 1 & 2 will take us a long way alone. 3 wold be magic (literally I think)

1) Track freeze is a technology that fixes this.

2) Get rid of Steinberg & Midi.

3) This does need a lot of work, but people are obsessed with pretty over usability, generally.

Ok, I don't quite understand number 3, but for number 1, 'track freeze' is a workaround not a fix or a solution and getting rid of the initial innovators of vst is no good as this seems to be the current standard of things. The future options to be initiated will have to take these older technologies (vst) into consideration for there to be a steady progress into newer technologies.
I'm certain that Midi will be replaced one day as it does have it's limitations. It has had one hell of a good life span though eh?

aciddose · Post by **aciddose** » Sat Sep 27, 2008 1:40 am

midi itself doesn't even exist really in vst. what has had a long life is the idea that we need fundamentally a set of commands to control an instrument: note on, note off, controllers both whole instrument (modulation, after touch) and per note (velocity, poly after touch) and in addition a few system functions like panic (allnotesoff/allsoundoff) adjustment of polyphony (not very good in midi but it's there somewhat) tuning and other global settings. we also need an expandable communications channel where messages specific to a particular instrument can be sent and received.

now we've actually realized the shortcomings of midi: unlike vst it doesn't have "getparameter" "setparameter" type functions since midi was designed with a single controller -> instrument mentality while vst was designed with midi + bidirectional communications and dynamic controllers in mind.

we do need to keep the existing systems in mind when designing new ones, but only so we can see what was left out or what was done wrong. not as an evolutionary step.

even the "note on", "note off" idea is fundamentally wrong. a lot of instruments are never "on" or "off" but always at some graduation between those two points and constantly changing. those sorts of instruments can't be very accurately controlled using a boolean system. to control the bowing of a synth violin for example requires several controllers to be used in addition to frequency, velocity and on/off state.

if anybody can come up with an efficient way to make the whole idea of a "note" more fluid with the variety of instruments out there from boolean to anything else feel free, you might end up the author of a new standard.

Casar1973 · Post by **Casar1973** » Sat Sep 27, 2008 1:42 am

From the point of view of what's going on in the universities I would say this is something that will be a major theme in general music technology:

*replicating human performances and human dynamics in how the human interacts with instruments and music. There are two universities I read of with dedicated research in how to decode the various aspects of humans interacting with instruments.

So if they succeed at any rate, we'll see virtual drummers, violinists, pianists etc in a new way in the future. Possibly new songwriting tools as well to assist at the level of songwriting to take the tune into various optional routes.

mellotronaut · Post by **mellotronaut** » Sat Sep 27, 2008 1:59 am

back to cpu-efficiency

i'm bouncing and freezing all the time again like in the old vst-days with a pc, i bought new not even three years ago ... .

loose snare · Post by **loose snare** » Sat Sep 27, 2008 2:00 am

Collapse of modern civilization ........ back to acoustic instruments only, yay!

mellotronaut · Post by **mellotronaut** » Sat Sep 27, 2008 2:06 am

loose snare wrote:Collapse of modern civilization ........ back to acoustic instruments only, yay!

not a bad idea ... better for our social life

koalaboy · Post by **koalaboy** » Sat Sep 27, 2008 4:26 am

Novata wrote: Ok, I don't quite understand number 3, but for number 1, 'track freeze' is a workaround not a fix or a solution and getting rid of the initial innovators of vst is no good as this seems to be the current standard of things. The future options to be initiated will have to take these older technologies (vst) into consideration for there to be a steady progress into newer technologies.
I'm certain that Midi will be replaced one day as it does have it's limitations. It has had one hell of a good life span though eh?

But Track freeze isn't a workaround. It's a solution to let you run really powerful software on a machine that can't cope with doing twice as much work, and then keep the result.

If you honestly think software can just keep getting more efficient on the same hardware, beyond a certain point, you're somewhat mistaken.

Personally, given the choice between a really powerful VST that could only ever use by itself, but I could freeze the result, or a not-so-powerful VST that I could run more instances of, I take the first (if I had to choose). Of course, the sensible option would be to use each where it was best suited.

MIDI is still great for many things, but VSTs should have moved beyond being tied to it, a long time ago. Replace the note-on/note-off messages with sound objects, that have a start and stop, but can then vary their pitch (and any other parameter) at any point - this could still translate to/from MIDI for existing devices, but with limitations on what legacy sound generators could handle. With software instuments, object-based messages are long overdue.

whyterabbyt · Post by **whyterabbyt** » Sat Sep 27, 2008 4:34 am

Novata wrote: getting rid of the initial innovators of vst is no good as this seems to be the current standard of things. The future options to be initiated will have to take these older technologies (vst) into consideration for there to be a steady progress into newer technologies.

I stenuously disagree. A rethink of a technology used by an industry on the basis of contributions of all the members of that industry makes more sense than encouraging one company to dictate that technology solely by their own needs.

Technical 'innovations' in VST pluginsd (mostly by subversion, such as MIDI-processing VST plugins, multiple inputs/outputs and sidechaining) have come from the small developers, while Steinberg seems more concerned with reinventing a wheel that these developers already thought up.

Novata · Post by **Novata** » Sat Sep 27, 2008 4:55 am

whyterabbyt wrote:
Novata wrote: getting rid of the initial innovators of vst is no good as this seems to be the current standard of things. The future options to be initiated will have to take these older technologies (vst) into consideration for there to be a steady progress into newer technologies.
I stenuously disagree. A rethink of a technology used by an industry on the basis of contributions of all the members of that industry makes more sense than encouraging one company to dictate that technology solely by their own needs.

Technical 'innovations' in VST pluginsd (mostly by subversion, such as MIDI-processing VST plugins, multiple inputs/outputs and sidechaining) have come from the small developers, while Steinberg seems more concerned with reinventing a wheel that these developers already thought up.

And what about people who have spent a lot of money on these vst products?
Hosts that run this new technology of plugins will have to be capable of running older plugins.
Coming up with new plugin technology is certainly the way forward. I must stress that I do not only include vst in this argument as rtas and other formats of plugins will need to be taken into consideration. I agree that not one company should dictate a standard but unfortunately Steinbergs vst format does seem to be the current leader in plugin technology with regards to the amount of vst plugins there are available. You can't ignore it, that's why it's incorporated in many of todays hosts.
One standard of plugins technology democratically decided by all members in the industry. I've been around a bit and that is almost laughable.
Yes it makes sense but since when did competing businesses in the same industry agree on a 'one standard' view?

Novata · Post by **Novata** » Sat Sep 27, 2008 5:03 am

koalaboy wrote:
Novata wrote: Ok, I don't quite understand number 3, but for number 1, 'track freeze' is a workaround not a fix or a solution and getting rid of the initial innovators of vst is no good as this seems to be the current standard of things. The future options to be initiated will have to take these older technologies (vst) into consideration for there to be a steady progress into newer technologies.
I'm certain that Midi will be replaced one day as it does have it's limitations. It has had one hell of a good life span though eh?
But Track freeze isn't a workaround. It's a solution to let you run really powerful software on a machine that can't cope with doing twice as much work, and then keep the result.

If you honestly think software can just keep getting more efficient on the same hardware, beyond a certain point, you're somewhat mistaken.

Err, yes it is a workaround.
Problem = CPU load is not enough for the software running on a machine.
Solution = Increase CPU speed or use multiple CPU's (Dual, Quad or however many) = fix

Workaround = Bounce audio down to hdd (freeze).

I never mentioned that you should run the same hardware.

where next with software instruments?