SFP "not suitable for advanced concepts"

wavelength · Post by **wavelength** » Mon Sep 01, 2003 8:29 pm

it seems to me that "coders" don't always make the right musical decisions and musicians basically are lost with a page of code in front of them. SCOPE DP puts well-written and stable code in the hands of musical artists... in a way that makes more sense creatively, to an artist, than a screen full of numbers and letters. that's why the devices sound so good. could Einstein have invented funk?

i am not saying that "DSP is better than native"... blah blah blah... just that SCOPE DP should not be arrogantly passed-off as less serious or less capable than the more conventional method of coding devices. it is *different* and this difference seems to be yielding some impressive results. this platform is not "musical lego".

[ This Message was edited by: wavelength on 2003-09-01 21:50 ]

Michu · Post by **Michu** » Tue Sep 02, 2003 12:01 am

Stephen,

oscillators and filters on SFP sure sound great, but do you know any method of modulating loop points in sample?
other things like fft or disk streaming were mentioned already...

I seem to remember, that major Angus' grief was inability to use custom GUI code, and he could not implement functionality he wanted and he had in his native plugins.

btw, there are native filters and effects that are at least on par with Creamware's.
Let's not get elitist

kensuguro · Post by **kensuguro** » Tue Sep 02, 2003 12:30 am

i am not saying that "DSP is better than native"... blah blah blah... just that SCOPE DP should not be arrogantly passed-off as less serious or less capable than the more conventional method of coding devices. it is *different* and this difference seems to be yielding some impressive results. this platform is not "musical lego".

Yeah, I think this seems to be the final line. The impression I get is that the whole SFP and its development platform is geared towards complete realtime signal processing. If I were to compare it with Max/MSP, it has the characteristics of the MSP part, which does the signal processing. While certain things in Max is possibly using the MSP portion, MSP can't do things off line, like writing text files or temporary tables to ram. But this is a valid choice, because SFP is built as a realtime synth/effects processor so basically, you don't need to do things off line.

Strange thing is tho, some of the CW devices seem to have off-line analysis. Not sure how they do it. Notably, STS5000 with its off-line stft analysis. They've got some serious juice going on with that device I think. And also Optimaster, with its automation option. That definitely stores a certain amount of info in a buffer so that it can be analyzed. I'd suppose the platform itself is capable of doing stuff like that. I mean, take 3 analysis results of STS5000 and if they were fft tables, you could some funky interpolation between the 3 and that'll be some great spectral synthesis. I can see that some of the important parts are there.

[ This Message was edited by: kensuguro on 2003-09-02 01:32 ]

marcuspocus · Post by **marcuspocus** » Tue Sep 02, 2003 12:49 am

Sorry, i can' t really add to all this. I just wanted to say that i'm really impressed by the level and quality of this discussion.

It give me the impression that all of the sfp user base is really skilled and mature in all this.

The only thing i might add is that either native/sharc code, a combination of both is possible, and totaly desirable imho.

The STS5000 is a good example, but also VDAT, and the old tripleDAT, even Optimater as said before.

I can easily imagine devices that can use both resources to do even more spectacular things than native alone or dsp alone.

I think developers should at least have a look in that direction. This could totaly blow away both native/dsp devices by combining both of their strong points.

I always wondered why it wasn't more used, i guess this is due to difficulties in programming such stuff, but hey, nothing is easy no more these days!

Again, i'm really impressed by the maturity and level of understanding of participant of this discussion.

I love the Z / SFP / third party developpers

wavelength · Post by **wavelength** » Tue Sep 02, 2003 12:58 am

something is a bit odd to me, here... Steve Duda is the "mastermind" behind the BFD recordings? I remember him having a nasty pissing-contest with CreamWare back in the early days of Pulsar (some of you might remember), when someone "dared" compare Pulsar to ProTools. It all did not end very well, just as I am assuming, FXpansion's exit from the platform did not happen on good terms. there seems to be some personal agendas against CreamWare here + SFP is technically competition for FXpansion now...

why is it that i am being accused of being elitist here when I am simply defending my position against comments like:

"the development environment has major shortcomings - it simply isn't suitable for developing advanced concepts like DR-008, MindFX or BFD"

and

"It's musical LEGO just like Reaktor, SynthEdit et al but with some fundamental technical flaws that even SynthEdit doesn't suffer from. It just doesn't have *anything like* the flexibility needed to build a BFD, a MindFX or a DR-008"

this is elitist, is it not? just because devices like these haven't been the focus, as of late, in the CreamWare "world" does not mean that they are impossible, or that the platform is suddenly incapable of "advanced concepts". sampling is simply something that does not vastly benefit from what DSP can provide... + there is so much sample-based stuff out there already for native, why not focus on something different?

Michu · Post by **Michu** » Tue Sep 02, 2003 1:38 am

I see your point
but
do you see the way of building something like DR008 in DP? with all the functionality?
IMO all samplers for SFP (ie STS line...) blow. All the way.
i don't mention MindFX and BFD cos they are vapour for now...
I am just used to that Angus usually weights words, so if he found it limiting, then it probably was. And he already mentioned that investment in developing suff for Pulsar simply didn't pay back.

wavelength · Post by **wavelength** » Tue Sep 02, 2003 3:41 am

On 2003-09-02 02:38, Michu wrote:
I see your point
but
do you see the way of building something like DR008 in DP? with all the functionality?
IMO all samplers for SFP (ie STS line...) blow. All the way.
i don't mention MindFX and BFD cos they are vapour for now...
I am just used to that Angus usually weights words, so if he found it limiting, then it probably was. And he already mentioned that investment in developing suff for Pulsar simply didn't pay back.

well, i personally have no interest in designing such a thing, as this is not my forte, but i don't see why someone else couldn't design one... it is true that the sample modules are somewhat limited in our version of DP, at the moment, but certainly much has been done in this area that isn't available to us all yet (3rd parties), as the STS-series represents. if there was an idea that was truly great and no existing DP atoms to get the job done, CreamWare would certainly partner in the project to help get things done... just look at the Bowen/ CreamWare collaboration which resulted in the excellent SCI products.

i, personally, just don't see what is so cutting-edge about sample manipulation that my k2000 has done for over ten years. granted there are some things that recent developers are adding to this concept, but things like Granular clouds don't get timbrally interesting until they have 1000s of grains (otherwise you just get stutters and hiccups which become unappealling to me very quickly), which would require a sick amount of processing power. i think CreamWare is trying to move things in a slightly different direction and that is their prerogative. there is, of course, nothing wrong with disagreeing with this direction and finding a different platform for your ideas... but this doesn't negate what CreamWare is doing or make it less "advanced".

as per the non-profit aspects of this platform... well, I'm still here!

(stirring up trouble, as usual)

[ This Message was edited by: wavelength on 2003-09-02 04:57 ]

scary808 · Post by **scary808** » Tue Sep 02, 2003 5:43 pm

http://www.mynetcologne.de/~nc-dietzvo/#SDS

I'm not too sure of the specs of the DR008 but here is a great sample drum player. It's an old device but I still use it. It's simple yet flexible. Volker ,if your out there, a new version of this handy device would be spectacular.

[ This Message was edited by: scary808 on 2003-09-02 18:44 ]

darkrezin · Post by **darkrezin** » Tue Sep 02, 2003 6:48 pm

While the SDS is a great, useful device, I think it's really unreasonable to compare it to DR008. It really is like comparing a Ferrari to a Nissan Micra... they both essentially do a similar thing, but one is clearly light years ahead of the other.

DR008 is not just a Drum Sampler - it is a semi-modular beast, with extensive sample-mangling capabilities. It also does synthesized drums (with an open architecture for 3rd-party modules), and allow sequencers to be assigned to pads, so you can trigger sequences realtime with MIDI keys. It is a very deep machine: I reccommend it to anyone whole-heartedly. I myself use it a lot, alongside EXS-24 and Kontakt... it really does offer certain things the others don't. In the Native world, Linplug RM4 and Battery do not even come close to what it can do. And much as I love the Creamware platform, there is just no way I would expect to see such a device for SFP.

I think, in general, sampling is Creamware's weakpoint.. speed of use is IMHO crucial when working with samplers. STS is extremely cumbersome. The fiddly method of working with samplers is one of the first things people remark upon when I show them what SFP can do.

peace

Michu · Post by **Michu** » Tue Sep 02, 2003 11:34 pm

<snip> i think CreamWare is trying to move things in a slightly different direction and that is their prerogative. there is, of course, nothing wrong with disagreeing with this direction and finding a different platform for your ideas... <snip>

Don't get me wrong, i find Pulsar fine platform for me ideas, just very weak on sampling side (see thread in wish forum, where people talk about multisample support in modular, well, stereo sample support would be a starter).
i enjoy very much flexibility of connecting and ease of integration with native side, also Modular powered up with Flexor, practically don't use anything else recently, cannot wait until others will be able enjoy it as well.

Nikko · Post by **Nikko** » Sat Sep 18, 2004 5:37 am

To create new atoms you need to have the tools to do so.

When I created the DSPLIB they game me only 1000 crypts, that means that I was able to compile only 1000 time and that it!!!

Imagine that you have Visual C++ and you are doing a tool and suddently, Microsoft call you and tell you stop!

That's exactly waht did happened.

Also the DSPLIB atoms are not all finished, the one I use in my device are stable.

Later, Creamware proposed me to send them the ASM source code so they will copile them for me.

Imagine that you are developping a Software with Visual C++, and suddently Microsoft say, you can't no more use Visual C++, but now, you need to send us you source code and we will compile it for you.

Of course to get your source compiled it takes at least on week!!!

That's exactly what did happened with Creamware.

SO I think that we can create good atoms, but Creamware does everything possible to stop us, and keep the technology for their exclusive use.

wavelength · Post by **wavelength** » Sat Sep 18, 2004 3:52 pm

On 2004-09-18 06:37, Nikko wrote:
To create new atoms you need to have the tools to do so.

When I created the DSPLIB they game me only 1000 crypts, that means that I was able to compile only 1000 time and that it!!!

Imagine that you have Visual C++ and you are doing a tool and suddently, Microsoft call you and tell you stop!

That's exactly waht did happened.

Also the DSPLIB atoms are not all finished, the one I use in my device are stable.

Later, Creamware proposed me to send them the ASM source code so they will copile them for me.

Imagine that you are developping a Software with Visual C++, and suddently Microsoft say, you can't no more use Visual C++, but now, you need to send us you source code and we will compile it for you.

Of course to get your source compiled it takes at least on week!!!

That's exactly what did happened with Creamware.

SO I think that we can create good atoms, but Creamware does everything possible to stop us, and keep the technology for their exclusive use.

can't argue with that...

astroman · Post by **astroman** » Sat Sep 18, 2004 4:18 pm

On 2004-09-18 06:37, Nikko wrote:
... Later, Creamware proposed me to send them the ASM source code so they will copile them for me. ...

what a generous offer - how could you refuse ?
...to provide that little educational service

ASM source for free - LMAO

cheers, Tom

[ This Message was edited by: astroman on 2004-09-18 18:51 ]

dehuszar · Post by **dehuszar** » Sat Sep 18, 2004 4:59 pm

Keep in mind, this thread was in reference to the old system. Let's wait until more details surface about the full, commercial SDK before getting all salty.

Sam

astroman · Post by **astroman** » Sat Sep 18, 2004 6:01 pm

ok Sam, I don't remember the circumstances - if Nikko had trouble with CWA (or CWA had trouble with Nikko) - so it could well have been that they made him an offer he only could refuse.
At least I found it more amusing in it's naiveness - or stupidity, if one takes the compile offer 'by word'.
For those who wonder: for a programmer that's almost the same as if parents would sell their children.
But lets just keep the past where it is...

cheers, Tom

dehuszar · Post by **dehuszar** » Sun Sep 19, 2004 8:57 am

My memory is that they both had trouble with each other. CW's old SDK and licensing policies were fairly draconian. But they are in the process of revamping those policies. While I can't imagine what frustrations Nikko had to endure, I think it's a little pre- and immature to go blasting the "free for all to use" SDK as being theft. The commercial SDK is not written yet, and if memory serves, the majority of the cost of the DP package and a good number of resrictions had to do with Analog Devices technology licensing to Creamware. So who knows what's been agreed upon at this point, but I'll be very surprised if anyone's work gets "stolen".

Anyone who does remixes will know that you have to REALLY make a name for yourself and they pay for all the parts you're using before you own any distribution rights to your own creation, because it uses the copyrighted work of another. It may not seem logical, or fair, but it is the way many things are done in the industry in which we operate.

Also, --and I may be wrong on this one-- but my understanding is that Steinberg's Development kit is not a "Photoshop for audio-tool creation" style SDK, but a couple of isolated tools that operate on a more raw-coding level. Anyone who could confirm or disconfirm, please pipe in.

This is not to diss Nikko or any troubles that he has had, just that the whole story has not been told, and we should wait and see what the whole picture looks like when the commercial SDK has been released and then re-open the conversation with CWFrank and the CWA crew.

Sam

Post by **garyb** » Sun Sep 19, 2004 3:15 pm

well said.