The Future Studio

[Koisan]

Let's have a little fun...given all the moaning about not having enough power - some here remind me of Scotty on the Enterprise... :lol:

I predict that we are slowly going backwards..my future studio (2008)

a) 4 - 6 PC's interlinked via an Ethernet variation of the present TCP/IP

b) Control surface + 2 expansions using same link

c) DSP cards per pc - 1 or 2 {hence the growing number of pc's}

d) Expansion DSP racks from TC, UA, Focusrite, Yamaha, Waves - unfortuanely not using the obvious Ethernet link
{too simple for marketing teams to agree - felt their company would lose some advantage}
{ requiring additional interface cards and bussing}

e) 6 LCD screens to keep up with the lack of tactile control - 2 screens - touch sensitive.

- In nutshell my dream of a simple powerful box to to the mix math - destroyed....

 

[audio-buddy]

Let's have a little fun...given all the moaning about not having enough power - some here remind me of Scotty on the Enterprise... :lol:

I predict that we are slowly going backwards..my future studio (2008)

a) 4 - 6 PC's interlinked via an Ethernet variation of the present TCP/IP

b) Control surface + 2 expansions using same link

c) DSP cards per pc - 1 or 2 {hence the growing number of pc's}

d) Expansion DSP racks from TC, UA, Focusrite, Yamaha, Waves - unfortuanely not using the obvious Ethernet link
{too simple for marketing teams to agree - felt their company would lose some advantage}
{ requiring additional interface cards and bussing}

e) 6 LCD screens to keep up with the lack of tactile control - 2 screens - touch sensitive.

- In nutshell my dream of a simple powerful box to to the mix math - destroyed....

my setup:

a) UAD-X DSP host card with plenty of power (8 super-duper DSPs)
b) Lexicon DSP host card with 960 engine (...oubs...) 8)
c) ONE PC - one box, not a dozen external HW units linked with 100000 different drivers.

ok, that might be the setup of 2010 ... and I now have to start saving money for the Lexicon card... :?

 

[Staccato]

Mine will be a little different. A single PC, Quad core processors with dynamically allocated resource management, MOBO with multiple 16x PCI-e slots with full user control over bus/device/lane, bandwidth/resource allocation, dime a dozen PCI-e expansion chassis with multiple slots, configureable for any combination of PCI-e devices from x1 to x8.

But what's to come after PCI-e?

 

[Eric Dahlberg]

I think Liquid Mix & Duende are going to make big, long lasting impacts & that people will move in quite the opposite direction from what's been described here because of them. Once proper dynamic impulses become commonly available, how much outboard are we really going to need anymore aside from preamps, DI's, converters, & a Nicerizer 16 (stretching on that last one, of course )? How much is the power of the CPU going to matter when such great sounding & powerful solutions will be so easily & cheaply obtainable? Native plug-ins will surely improve along the way, too, but patent laws ensure Liquid Mix its place at the top, not to mention that it's a tactile solution, as well.

 

[RobPain]

I agree that some things are going backwards whilst the underlying technology is still moving forwards.

There is no question that CPUs will increase in processing power. However, while loading everything in one box simplifies some things, you are severely limited by the standard input devices of a mouse and keyboard. It is clear that being able to work quickly and intuitively is a huge bonus, and a high priority for professionals in the area.

My prediction is therefore simply a continuation of the current trends - software will continue to increase in complexity and capability (e.g. at some stage, it will be common for hobbyists to be mixing at 32-bit, 192kHz), and hardware control will continue to remain at the forefront of many designers' minds (just look at Euphonix System 5-MC, or NI KORE for examples).

 

[Koisan]

Commenting backwards...

a) I agree a mouse and keyboard just won't cut it - but how about those touch screens? Being able to physically work a desk that does not exist - and yet it does!!! :wink:

b) Convolution - proprietary/patent?
Remember the Sony sampling(convolving) reverb? remember the price? not? ...my point exactly - long live Voxengo :lol:

\" Sintefex?Why does no-one mention Waves Q-Clone? They have been doing EQ convolution for some time now\"

c) Mobo's with more PCI-e slots? that implies less MOBO sales... guess not then :|

d) 32 192 - probably - however our imperfect hearing would still result in endless debates of real merits of this technology even in 2010..viva forums! :lol:

 

[Eric Dahlberg]

b) Convolution - proprietary/patent?
Remember the Sony sampling(convolving) reverb? remember the price? not? ...my point exactly - long live Voxengo :lol:

" Sintefex?Why does no-one mention Waves Q-Clone? They have been doing EQ convolution for some time now"

The Sony, Voxengo, & Waves convolution devices do not use dynamic convolution like the Sintefex/Focusrite do.

 

[Koisan]

b) Convolution - proprietary/patent?
Remember the Sony sampling(convolving) reverb? remember the price? not? ...my point exactly - long live Voxengo :lol:

" Sintefex?Why does no-one mention Waves Q-Clone? They have been doing EQ convolution for some time now"

The Sony, Voxengo, & Waves convolution devices do not use dynamic convolution like the Sintefex/Focusrite do.

..but what does that really mean other than a marketing angle.. :wink:

 

[cAPSLOCK]

Dynamic convolution has a huge advantage. It can be argued that it is a valid way to get the non-linear elements of something like a compressor or EQ without having to model at the component level. Those who like the focusrite and Sintefx hardware will argue that it works.

As far as reverbs go, there is no real advantage to dynamic ocnvolution. But as far as EQs or Compressors go it is really impossible without it.

It is not a marketing thing... And the patent is legit.

As to whether it really works as well as component modelling... the jury is still out.

cAPS

 

[haze]

my children's children's studio won't have any wires in it ... :roll:

 

[Eric Dahlberg]

..but what does that really mean other than a marketing angle.. :wink:

That's a sad comment to make. Why is your first reaction to assume it is only a marketing ploy? If everyone did this, there would be no point for companies or individuals to do anything innovating.

If I may comment on cAPS' point, I'm looking forward to the day we start seeing dynamic convolution used for tape saturation & amp modeling. I'm not sure how necessary dynamic convolution is for EQ's, though. Waves & Tritone Digital seem to make people happy enough with their static convolution EQ's.

 

[Koisan]

..but what does that really mean other than a marketing angle.. :wink:

That's a sad comment to make. Why is your first reaction to assume it is only a marketing ploy? If everyone did this, there would be no point for companies or individuals to do anything innovating.

If I may comment on cAPS' point, I'm looking forward to the day we start seeing dynamic convolution used for tape saturation & amp modeling. I'm not sure how necessary dynamic convolution is for EQ's, though. Waves & Tritone Digital seem to make people happy enough with their static convolution EQ's.

Sad - not really - been round long to enough to become a little cynical.

Why do you use the description "dynamic convolution" like it is something radical? Look at the UAD guru's description
http://www.uaudio.com/webzine/2004/july/index2.html

Basically this is the method most dsp developers employ for developing a software emulation of a piece of electronic hardware - Yamaha started way back, Roland did it, SansAmp, Line 6 etc

Now, personally I think UAD have taken a better approach by attemping to capture the design, component by actual component (note 'actual')

 

[Eric Dahlberg]

Basically this is the method most dsp developers employ for developing a software emulation of a piece of electronic hardware - Yamaha started way back, Roland did it, SansAmp, Line 6 etc

Can you cite some specific examples? I've never heard this about these companies before.

 

[Koisan]

..but what does that really mean other than a marketing angle.. :wink:

...I'm looking forward to the day we start seeing dynamic convolution used for tape saturation & amp modeling. I'm not sure how necessary dynamic convolution is for EQ's, though. Waves & Tritone Digital seem to make people happy enough with their static convolution EQ's.

..I see your sarcasm... and raise you - maybe you were hoping for power amp modelling - Hiwatt - FET model maybe? :wink:

 

[svs95]

Dynamic convolution has a huge advantage. <snip> as far as EQs or Compressors go it is really impossible without it. <snip> As to whether it really works as well as component modelling... the jury is still out.

Not with some of us, it's not. As Dave Berners said in the article Koison linked to, "Dynamic convolution can be used successfully for a very specialized set of nonlinear systems, but in general is not appropriate for systems with hidden internal states, such as dynamics processors."

The number of impulse responses required to accurately represent the nonlinearities of an analog compressor is too unweildy! This is not what convolution (dynamic or otherwise) is best at. EQ is a different matter, being essentially linear.

Regarding dynamic convolution, Berners again: " for nonlinear systems, it is possible that an infinite set of measurements is needed to do the characterization. Moreover, in most cases it is impossible to determine when enough measurements have been made. For linear systems, a device can be characterized fully with just a single measurement, but only for the particular knob settings tested. This means that for devices with many controls, it may be cumbersome to characterize a device using signal modeling. Without sophisticated analysis, data needs to be collected for every conceivable set of knob positions, and some method must be introduced to interpolate between measurements"

UA looked long and hard at doing this, and in fact they do use some (static) convolution in their EMT Plate 140. But as for its electronic circuitry, and all their other emulations, they stick to circuit modeling for all the above reasons. Talk about hidden system states, UA has even modeled the quantum states of the optical photocell in the LA2A.

I have seen/heard the difference it makes with compressors. The Fairchild, 1176 and LA2A on the LiguidMix sound very little like the UA versions, and don't produce the signature program-dependent phase-dynamic response envelopes.

The thing is, these focusrite boxes make nice sounds, and people will buy them. It just isn't accurate to call them "emulations." They have some of the tonal qualities and general dynamics, but not the signature responses of the real deals.

BTW, Duende does not use not dynamic convolution, AFAIK. It is a port of the algorithms from the digital SSL mixer processors.

 

[svs95]

I also like the idea of a hardware box which can move between workstations, controlling VST plug-ins inside the workstation, and I especially like the idea of one which emulates the manual interface of the emulated hardware (or simulated, in the case of original designs). A touchscreen is one nice way to do that. That would be a cool future product.

As for potential PCI-e power, have a look here:
http://www.cyclone.com/products/pci_express_expansion_system.php

 

[Eric Dahlberg]

..I see your sarcasm... and raise you - maybe you were hoping for power amp modelling - Hiwatt - FET model maybe? :wink:

Actually, it wasn't sarcasm. I'm American, we're not good at that stuff over here. :wink:

 

[Eric Dahlberg]

It's been said on this forum that the UAD plugs do a better job of emulating the dynamic qualities of certain processors while the Sintefex does a better job of emulating their tonal qualities. I'm looking forward to getting to use the devices together to get the best of both worlds.

Portability is definitely a huge consideration. I'd rather carry everything in a single PCI expansion rack than as separate units, though, especially if firewire proves to be as much as complication for SSL & Focusrite as it is for TC Electronic.

 

[Koisan]

Just to move on a bit...


I would like to go on record (probably not the 1st unfortunately) by stating the next phase in modelling vintage and modern circuits will be the function for the user to choose certain \"components\" .

Imagine that instead of the factory standard TLA input transformer we get to select one with a different turns ratio! or ferrite core!

How about a choice of vacuum tubes?

Customising your plugin!! what could be cooler :!:

 

[Eric Dahlberg]

Creamware gives you that level of control when you sign up for the free SDK. They're better for synth stuff than studio stuff, though, & I don't think they've modeled any transformers yet.

edit - I may be wrong about the transformers, Creamware's VinCo compressor sounds a lot like the DBX 160SL & that has Jensens in it.