future of scope?

lagoausente · Post by **lagoausente** » Wed May 24, 2006 10:49 am

The only question I have in my mind is why Creamware have been on finacial problems. Really, buying a Scope Home is not much more expensive that buying any other pro soundcard, with the advandages of DSPs. I really don´t understand many people here in Spain selling and selling their Pulsar. Probably one reason is that is PCI, and many people go for a laptop, and buying a Magma seems really expensive.
If Creamware company finally don´t go up on their financial position, and the company stops, it will be a very absurd lost. Really is one thing that works, while OS based sofware is giving problems, cracks, pops, just see any Sequencer forum.
Since I have changed my RME for the Scope home I have now, I have been playing and recording much more time than before. Just sending audio from devices to sequencer, all is in time, without any midi time problem, just like a tape recorder, what i think it must be, if we are really musicians, or trying to be.

[ This Message was edited by: lagoausente on 2006-05-24 11:49 ]

astroman · Post by **astroman** » Wed May 24, 2006 12:37 pm

On 2006-05-24 11:49, lagoausente wrote:
...I really don´t understand many people here in Spain selling and selling their Pulsar. Probably one reason is that is PCI, and many people go for a laptop, and buying a Magma seems really expensive. ...

maybe they cannot hear (the difference) ?
one has to buy the software - no w*rez
they always follow the 'majority'

cheers, Tom

[ This Message was edited by: astroman on 2006-05-24 13:38 ]

darkrezin · Post by **darkrezin** » Wed May 24, 2006 3:48 pm

Lago - if you had MIDI timing and crackling problems with Scope, then it's a fault of your system... perhaps Scope hardware is more sensitive to average/poor quality systems than RME, but it does way more complex stuff.. you cannot compare them. The RME specializes in a *miniscule* fraction of what Scope is capable. It's no surprise that it may do that part better.

If you don't need Scope you don't need it.. IMHO with the lack of real-world studio knowledge nowadays, a real-time sample-by-sample system like Scope is less appealing for the consumer hordes. However, they are the ones losing out, suffering flat and lifeless mixes in their floating-point native apps. I don't care for that dimension-less sound.

astroman · Post by **astroman** » Wed May 24, 2006 4:32 pm

just to be fair - optimastered on-air stuff is the plague...

okantah · Post by **okantah** » Thu May 25, 2006 3:39 am

Certainly customers expectation to use SCOPE dush out,when they realised that it take's same time to lean to use it.
Cheers

craighuddy · Post by **craighuddy** » Thu May 25, 2006 4:47 am

Well, for me , Scope does what it is supposed to do. I think sometimes we put the PC ( or Mac) in front of the platform. A Platform as serious as Scope deserves to be on it's own machine, meaning, Scope first, OS second. If you want to use scope use a system that is compatible. I guess that depends on how serious a user is though. I would not buy a standalone Roland hard disk recorder and expect to write documents on it or play games on it. Likewise, I bought and setup up my DAW for one purpose, produce music. If Vista comes out and it is not compatible with scope, I will stay on XP.The Daw and OS is there to support my Sequencer and Scope, not the other way around.

That said, I would not discourage in any way the development of 64bit and OSX drivers as there are several native plug ins that are exclusive to that platform that would benefit from faster CPUs etc. The premise that microcode extensions are DSP is a little stretch, but floating point is king in Native land. If a new CPU can help out , I Kind of hope scope keeps up to the point where they are not obsolete so I can update as needed. They basically are on the mac now, and I fear will be on the PC soon enough. That said, I see my self using Scope on XP for at least the next couple of years. It really is a brilliant platform.

bronYaur · Post by **bronYaur** » Thu May 25, 2006 10:13 am

For the pci Express the creamware card seem OUT bat just get 5 volt directly to power supply like Z link power supply conctor ,pciE work only at 3 volt ,scope card work also at 5 volt.
I'm happy CWaudio user with three scope cards,I hope creamware don't stop scope platform in next years and find solution to be alive scope system

[ This Message was edited by: bronYaur on 2006-05-25 11:15 ]

symbiote · Post by **symbiote** » Thu May 25, 2006 1:05 pm

On 2006-05-25 05:47, craighuddy wrote:

The premise that microcode extensions are DSP is a little stretch, but floating point is king in Native land.

Mind explaining the difference/distinction then?

organix · Post by **organix** » Thu May 25, 2006 3:10 pm

Ok People, some really nice answers here.
To a few things I would like to take position.

You must imagine Scope more like an hardware instead software. If U had buy i.e. a Korg Wavestation EX in 1990, do U now think that is obsolete??? And a hardware LA2A compressor? I think not so...

If Scope would be only Hardware, I never bought it.
The DSP Cards are hardware, yes, but the essential, the "Scope Fusion Platform" is software with all their advantages and disadvantages.
One of the advantages is, that it's possible to extend and improve it without buying a complete new gear.
Computer Software lives in an environment, that makes it necessary to maintain it.
As a developer of software it's necessary to keep your programs up to date to the current available systems.
If a software is not adapted, it will be not more used on short or long. The most of us, maybe all, don't use Windows-3.11 Applications anymore.
For today it's necessary that a computer software runs on Windows-2000/XP. In a few years it will be necessary, that software runs on Windows-Vista or any future operating systems.
When CW Scope will not run on future computer systems, it'll die. I'm sure about this.
Maybe you can keep your current system alive for a few years, but the time will come and you'll get no spare parts for your old system and you'll be forced to use a new computer system.

Well, you are new here to the forum, and your questions have been on everybodies mind for quite a few years.

I don't know if you know that Creamware are still recovering from big financial problems

I'm on this forum for a few years, but more as a reader and I know that Creamware was broke.

Let me explain a few thing, so no false impression on me develops.
First, I'm a german guy and speaking englisch isn't something I do everyday. I hope you all understand what I'm writing here.

I'm a scope user for about 5 years and I love to use it. It's not necessary to explain me the advantages of Scope.
I know the big advantages of scope against some other systems, that is why I bought my CW cards in the past.
But, I want to use it in the future too.

Creamware seems able to develop time and cost consuming hardware gears (ASB). It seems they concentrate their work,on emulating legendary synthesizers and neglect their hard- and software masterpiece, Scope.
Maybe I'm wrong and they still do developing on scope, but how long do they still work on OSX compatibility? Maybe, too long. For a Mac enthusiast buying a Windows-PC to run Scope isn't a really good option.
I'm lucky, because I'm not a Mac user.

CW don't give me the feel, that they constantly keep Scope up to date.
Maybe some of you don't need any updates, because they are lucky with their current systems. "Never change a running system", is absolutely ok.
But, the time will come and you have to change your systems. Computer hardware will fail and getting any spare parts for older system will become harder and maybe impossible.
Check the market for AT-Components, EDO-Rams, Pentium-1 CPUs, EISA-Cards etc. Parts that there up-to-date not 10 years ago.
You'll run into problems to get some older parts if you would need them now.
Maybe in 5-10 years you'll have to change your current computer to a newer hard- and software platform and if Scope don't run on them, you can trash your CW equipment.
So I think, you can't think about CW Scope like a hardware gear. Constantly development is necessary to keep Scope usable.

If Vista comes out and it is not compatible with scope, I will stay on XP

And what do you think how long you can stay on XP. Be honest, do you still would like to use Windows-3.11?
Making Scope with updates compatible to Vista is a must or Scope will die.
I'm really sure, or hopeful, that CW will make scope compatible to Vista.

Creamware officials every now and then indulge in hieroglyphic or cryptic innuendos about a 'next generation' or a '5.0' or a 'OSX driver'.

This sounds good to me. Even if there is only rumors, I'm happy about such statements. This gives me a small light of a future with Scope.

VST doesn't have any infrastructure to support oversampling, so you end up with plugins that eat up a whole 2GHz CPU for 8 voices, which can be accomplished on a 200MHz CPU. need more power?

We all don't know what VST will be capable in future.

Think about how fast the developing of computer systems occur in the last 10 years. Computer prozessors are speed up with factor more than 10, the available RAM extended of factor 1 million and more.
Maybe in 5-10 years, will have 20 or 200 GHZ CPU with multiple cores for parallel processing. Perhaps, you'll be available to load hundreds of native hardcore synthesizer plugins and latency of 1 ms.
You think, this is only a dream. For about 20 years people thought, that we will never need more than 128 kilobyte RAM. Now, we have Gigabytes of RAM. Calculate yourself how much more this is in 20 years and think further.
Maybe, in 10 years we'll laugh about the computers of today. Big companies (like AMD for example) today have concrete planes for the next 2 years of multiple core (>4) cpu or 128bit architecture.
The next server operating system of Microsoft will have first support of 128 bit computer architecture.
Maybe, if you're interested, check the revolutionary Cell Core CPU that will build for Sony Playstation 3. This piece will have something like 8 cores and able to carry 10 instructions the same time with a processing speed of over 4Ghz.
The current Intel dual core cpu, are only a development cut of a further planed multicore cpu with 4 or more cores. But, because the overall development needs more time, they released the current state to public.
People want to see progress in development on products, not on reports.
You believe if you see it? Look to the past and maybe you can imagine the future.

But, I'll not say, that DSP based systems like CW-Scope, UAD, Powercore etc. will die. If they're consequently developed they will also get better and computer capacities are not the only reason of using audio DSP cards.
Most DSP Plugins are accuradely developed with really good algorithms for a really good result in sound. This is also the reason, why still many hardware gears are often better.
But, native Plugins were really improved in the last few years and become more meaning in our studio environment. They were improved, because they're consistently developed.
Our CW-Scope will be even improved, if it's consistently developed. Without such development, many scope plugins and features will be overhelmed by native software.
For example, I've used the Masterverb-Pro a long time for audition and the final mix. But since I have a good convolution reverb, I'll use the Masterverb for audition only and never for the final mix.
Searching here on PlanetZ regarding a convolution reverb for scope, results in something like an impossible!
"Impossible" isn't a good view into future.

I hope, that CW will be encouraged and able to bring scope over to the next generations of computers. I would love it.

Thanks.

regards.
Markus

_________________
http://www.neotrax.de
German Soundtrack and Film-Score Artist Community.

[ This Message was edited by: organix on 2006-05-25 16:11 ]

craighuddy · Post by **craighuddy** » Thu May 25, 2006 3:34 pm

On 2006-05-25 14:05, symbiote wrote:
On 2006-05-25 05:47, craighuddy wrote:

The premise that microcode extensions are DSP is a little stretch, but floating point is king in Native land.
Mind explaining the difference/distinction then?

as explained to me by both hardware and software engineers, the extentions are exactly that, added micro code to the x86 micro code syntax. They simplify the execution of code by sending one command for what it prior took in some cases , over 30. Basically they wrote a few more words into the language.

DSP, as the "P" notes, is a seperate processor specifically for manipulating a digital signal.By definition a DSP is a seperate piece of hardware, not an extention of a programming language running on a general purpose processor, but a seperate processor designated to doing a specific task.

astroman · Post by **astroman** » Thu May 25, 2006 6:48 pm

On 2006-05-25 16:10, organix wrote:
...Think about how fast the developing of computer systems occur in the last 10 years. Computer prozessors are speed up with factor more than 10, the available RAM extended of factor 1 million and more.
Maybe in 5-10 years, will have 20 or 200 GHZ CPU with multiple cores for parallel processing. Perhaps, you'll be available to load hundreds of native hardcore synthesizer plugins and latency of 1 ms.
You think, this is only a dream. For about 20 years people thought, that we will never need more than 128 kilobyte RAM. Now, we have Gigabytes of RAM. Calculate yourself how much more this is in 20 years and think further.
Maybe, in 10 years we'll laugh about the computers of today. ...

computers aren't faster because they are higher clocked - the 'performance advantage' is mainly used to cut down the most expensive resource - human wages.
Jobs (in developement) are automated, schedules are tightened and code quality doesn't really matter (regarding performance).

I admit that I overquote it a bit, but since you mention a time span of 20 years:
you won't find anything reaching even 10% of the original Mac's 128 KByte Rom code quality in any M$ product.
As such I'd rather laugh about todays machines

time is money and the IT business has turned from an industry to a self-fullfilling prophecy.
Apple scores again:
1984 they used excerpts from shell scripts as an antipole to position the Mac OS as a user friendly alternative - 23 years and a 2-figure billion $$$ profit later the Terminal is integral part of the OS

As frequently mentioned - the point is not to change this (as its impossible), but to be aware of how this industry is run.

In that context your concerns are simply too dramatic, imho.
The 'full' SFP including SDK is a self contained environment for audio processing, that's almost independant from the host machine's OS.
Like it or leave it, but the reason it doesn't run on OSX or Linux yet is the protection of intellectual property.

The ASBs are simply the reverse of what TC Works does with their hardware units - they transfer the code to ProTools, CWA transfers existing DSP code to new hardware boxes.

Looks like there's some market success - more will follow and allow the company some progression again.
Which will most likely be new products, based(!) on SFP experience - but almost certainly not an attempt to queeze current cards in whatever may come along.
This doesn't make sense economically and technically.

For the lifespan of the card you can buy tons of hardware for barely nothing, so there isn't even the need for CWA to catch up with certain nonsense.
If they weren't already broke on NOAH (?) they'd certainly been from an OSX port - considering the sh*t Apple released in early versions.

Imho everything is fine, I completely agree with Craig about usability, in particular since I gave the Pulsars it's own P3 - so much more convenient

As one can tell from recent releases, it obviously took 3rd party developers some time to realize the (true) potential of SFP - and that it doesn't need custom machine code to make a good sounding product.

I still don't see which product could replace SFP - and since I know (at least a little) about software developement, I'm also convinced that noone else will show up on the market with something similiar

cheers, Tom

[ This Message was edited by: astroman on 2006-05-25 19:54 ]

symbiote · Post by **symbiote** » Thu May 25, 2006 9:20 pm

Maybe in 5-10 years, will have 20 or 200 GHZ CPU with multiple cores for parallel processing. Perhaps, you'll be available to load hundreds of native hardcore synthesizer plugins and latency of 1 ms.

Nope! Well, perhaps, maybe, hypothetically, yes, but very unlikely. Read this if you want to know why:

http://www.gotw.ca/publications/concurrency-ddj.htm

Forget about processor's clock getting faster anytime soon. This is why EVERYONE switched to dual/quad core architecture: because they can't push the clocks higher, at least not without a drastic cut in density to reduce heat (for example, some electronic circuits interfacing with optical signal run much faster, but they are far less complex), which will damage your performances a whole lot more than the gain you will have with a faster clock.

This doesn't mean clocks won't get faster, but it DOES mean that they will go up much slower.

Thankfully, audio processing is easily parallelisable (?), but even then you will run into the throughput problems I described.

Other points you seem to forget:

- DSPs and most other types of circuits whose performance is based on transitor density DIRECTLY BENEFIT from any advance made in CPU architectures. As the CPU clocks/performance get faster, so does the DSP clocks/performance.

- Seeing how the biggest performance consumers in the consumer market (i.e. other than military, engineering, scientific, financal stuff) are gamers, there is as much, if not more performance-pushing for the DSPs that power most/all the modern graphics card.

Try and look up some benchmark on how fast (in raw floating points ops) todays GPUs are compared to CPUs, and you will see that the GPUs are much faster (but more specialized.)

symbiote · Post by **symbiote** » Thu May 25, 2006 10:00 pm

On 2006-05-25 16:34, craighuddy wrote:

as explained to me by both hardware and software engineers, the extentions are exactly that, added micro code to the x86 micro code syntax. They simplify the execution of code by sending one command for what it prior took in some cases , over 30. Basically they wrote a few more words into the language.

What you describe is just "micro-coding". While it is true that there is micro-coding involved with SSE, they are much more than just a set of commands.

The SSE extensions adds two things: registers for its calculations (in addition to the processor's own standard set of registers,) and SIMD (SSE stands for Streaming SIMD Extensions) floating point and/or integer ALUs. These processing units are much faster than the standard ALU of the general purpose processor because of their SIMD architecture, i.e. they can do anywhere from 1 to 16 different calculations in the same clock tick.

I won't really waste time trying to convince you though, the AltiVec and SSE/SSE2/SSE3 specifications and implementation details are fairly easy to get. I would strongly advise you to read up on the subject if you are to argue out about it further (comp.eng graduate here) =P

Here have fun! They'll usually link to all the meaningful places at the bottom of the articles if you want to check the gory implementation detail:

http://en.wikipedia.org/wiki/MMX

http://en.wikipedia.org/wiki/Streaming_SIMD_Extensions

http://en.wikipedia.org/wiki/AltiVec

http://en.wikipedia.org/wiki/Vector_processing

http://en.wikipedia.org/wiki/SIMD

Choice quote:

"In the past there were a number of dedicated processors for this sort of task, commonly referred to as Digital Signal Processors, or DSPs. The main difference between SIMD and DSP is that DSPs were complete processors with their own (often difficult to use) instruction set, whereas SIMD designs rely on the general-purpose portions of the CPU to handle the program details, and the SIMD instructions handle the data manipulation only."

Look, more!

http://www.embedded.com/showArticle.jht ... =175800222

" Recently processor designers have been reducing power consumption and part cost in a new way - by combining RISC and DSP features into a single core, known as a ‘convergent’ processor core.

Two examples of architectures that were designed from the beginning to be convergent are the Analog Devices Blackfin processor and StarCore processors (SC1000, SC2000, and SC v5). Some other examples of convergent architectures are based on well established RISC architectures and have been modified to efficiently fulfill DSP functions. These include the MIPS 24KE, Renesas SH3-DSP, PowerPC with AltiVec, and ARM966E-S."

DSP, as the "P" notes, is a seperate processor specifically for manipulating a digital signal.By definition a DSP is a seperate piece of hardware, not an extention of a programming language running on a general purpose processor, but a seperate processor designated to doing a specific task.

First off, the "P" can stand for either "processing", describing the field of digital signal processing, or "processor" like you say.

In this case, I was indeed refering to the processor part of things, but notice how I used the term "DSP core", for good reason. I was refering not to the whole separate hardware chip, but only to the circuit inside (the major chunk of it that does the actual calculations, ALU + registers if you know what those here.)

Both general purpose DSPs (like ADC SHARC, TI, Motorola 56k etc) and CPU sporting SSE/AltiVec instructions use SIMD (single instruction, multiple data) circuits to accomplish their speed performance, and if you do a bit of research on the subject, you will find that the bulk of the FLOPS performance of general purpose processor come from the SSE/AltiVec extensions (hence my comment on the subject.)

This being said, you are a bit off on the instruction set too. Most DSP architectures, in the separate-hardware-processor sense of it, ALSO describe a complete instruction set, that is usually implemented across several different models and generations of processor architecture by the different vendors.

So yeah, nah, sorry, calling the SSE/etc instructions "DSP cores" really wasn't anywhere near a stretch, they're pretty much the exact same kind of technology and circuit architectures.

organix · Post by **organix** » Fri May 26, 2006 1:40 am

On 2006-05-25 19:48, astroman wrote:
I still don't see which product could replace SFP - and since I know (at least a little) about software developement, I'm also convinced that noone else will show up on the market with something similiar

In my opinion, ProTools for example is a product that is comparable and maybe better, but for a higher price.
But yes, I agree. From a overall point of view, SFP is a unique product with a really good price/perfomance ratio.

Some parts of SFP could be replaced by other, better products, but not the whole system.

As good SFP is, but there is still room for improvements.

The ASBs are simply the reverse of what TC Works does with their hardware units - they transfer the code to ProTools, CWA transfers existing DSP code to new hardware boxes.

Looks like there's some market success - more will follow and allow the company some progression again.

I wish them the best success and much money, so they can pimp my Scope to the next level.

-
Markus

Shroomz~> · Post by **Shroomz~>** » Fri May 26, 2006 1:55 am

SFP itself is sweet as it is imo.

As for pimping it, well that has been & continues to be done by anyone with the inclination & the sdk. The platform has dozens & dozens & dozens of different & amazing 3rd party devices with dozens & dozens more on the way for sure. So I really don't understand all the fussing & ranting.

Scope rules.

astroman · Post by **astroman** » Fri May 26, 2006 7:28 am

I don't read any rant - the questions at the beginning of the thread are reasonable and probably a lot of folks ask themselves in a similiar way.
It doesn't hurt to learn a bit about the background of DSP coding and DSP hardware either, along the way.

regarding (possible) Scope customers noone will spent one or more k Euros on an obscure description that's contrary to all industry and marketing preaches and even some of the 'leading' musicians' statements

of course, after the Scope experience one knows better...

the sales of Scope itself are certainly less than one would appreciate, but on the other hand this may lead 'competitors' to (possibly) looking down on the company with regret

cheers, Tom

astroman · Post by **astroman** » Fri May 26, 2006 7:46 am

On 2006-05-26 02:40, organix wrote:
...In my opinion, ProTools for example is a product that is comparable and maybe better, but for a higher price...

I see your point, you probably have 'productivity' in mind.
Scope is in fact a bit 'spread all over the place' in this context.
I don't have ProTools myself - but I wouldn't hesitate to buy an old (Mac based) one if, yes if there were silent SCSI disks...

But afaik PT doesn't have Scope's free routing capabilities, while Scope lacks the integration of the very good, yet isolated VDAT recorder.

On the other hand this would never be a ready-to-mass-market solution anyway, so possibly a better automation and other improvements in this context might be a niche for ambitious 3rd parties (OEM)

cheers, Tom

lagoausente · Post by **lagoausente** » Fri May 26, 2006 12:30 pm

Lago - if you had MIDI timing and crackling problems with Scope, then it's a fault of your system

I have just told reverse. I told I´m not having any problem with Scope, and was refering to sofware VSti, and samplers in general. I´m telling Scope is better for me than sofware. Really hate when people reply before reading.

darkrezin · Post by **darkrezin** » Fri May 26, 2006 12:32 pm

Chill out dude, I just misunderstood the language in your post. Apologies.

craighuddy · Post by **craighuddy** » Sat May 27, 2006 12:25 pm

So yeah, nah, sorry, calling the SSE/etc instructions "DSP cores" really wasn't anywhere near a stretch, they're pretty much the exact same kind of technology and circuit architectures.

I still maintain that DSP by definition is a seperate piece of hardware with seperate discrete inststuction sets. SSE etc qualifies for the second requirement but not the first. Your first "choice" quote as much as says such. Having it do "DSP like" funtions does not make it a dedicated DSP.
From your own links:

"The main difference between SIMD and DSP is that DSPs were complete processors with their own (often difficult to use) instruction set, whereas SIMD designs rely on the general-purpose portions of the CPU to handle the program details, and the SIMD instructions handle the data manipulation only. DSP's also tend to include instructions to handle specific types of data, sound or video for instance, whereas SIMD systems are considerably more general purpose."

I obviosly am not going to change your mind about this any more than you are going to change mine. That said, lets move on to the topic at hand and not hijack this thread any more than we have

[ This Message was edited by: craighuddy on 2006-05-27 13:28 ]