PlanetZ forums - scopeusers.com

just as curious as all you other scopo-filiacs, I browsed the creamware ftp press information. It contains some high resolution pictures of the new x-ite box. However, flipping through, I noticed that the pictures are photoshopped.

Let's hope they will show something real and working at the Musik Messe!

It's real, I'm beta testing it here in Germany!
And it works great, running several synths with lots of DSP room.
cheers,
john b.

p.s. Please remember, it's not Creamware, its Sonic Core now.

Hey John,

Can you tell us if old plug-ins/modules can be directly used or is there in upgrade process to perform.
Just a question from a newbie SDK plug-in designer ?

CheerZ

Hey John Thanks 4 The Information,

I am down for the new toys no problemo.

How's about a quick taste.

When load S5 into it, and say have 8 poly. Do you notice any change in speed when you load those large quad Oscillatoe presets.

I guess I am wanting to know if Scope works at a higher speed now?


This unimportant question is just a l;ive thing, no biggie.

I am buying it regardles..


Danke.

HUROLURA wrote:Hey John,

Can you tell us if old plug-ins/modules can be directly used or is there in upgrade process to perform.
Just a question from a newbie SDK plug-in designer ?

CheerZ

Yes it seems the old plugins are working fine.
I'm sure there's some more fine-tuning they will want to do as they go through every plug-in to find out if there's any small problems, but overall, it's pretty impressive!

-jb

that sounds great! normally when new hardware is introduced its unstable at the beginning.

If this is rock solid right from start it will quickly gain a good reputation...
(which is IMO a very important thing and should not be underated.)

I cannot say how much impressed I am.



The end of the creamware story:

a happy end:
Soniccore !

Ok - as everybody else in here this new product looks very impressive! But (there's always a but), it relies on the old plugins, with which a list of bugs are known to this community. Has these bugs been solved in SFP5 and XITE-1?
My main concern are the modular and the samplers, but other plugins has bugs as well.

Cheers!
Thomas

decimator wrote:

John,

One question as you can beta test the new box: specs announce 10x more power, is it compared to a 14 DSP board? And on which operating system are you testing it? MacOS, Windows (XP|Vista), ...

Thanks, looks promising!

Fred

johnbowen wrote:

HUROLURA wrote:Hey John,

Can you tell us if old plug-ins/modules can be directly used or is there in upgrade process to perform.
Just a question from a newbie SDK plug-in designer ?

CheerZ

Yes it seems the old plugins are working fine.
I'm sure there's some more fine-tuning they will want to do as they go through every plug-in to find out if there's any small problems, but overall, it's pretty impressive!

-jb

Great news John,

Thanks for this kind answer.

Maybe you could find a few seconds to try to load my very first plug-in.
You can find it over there:
http://www.planetz.com/forums/viewtopic.php?t=23873
Not that is such a great one, but could maybe used as a quick bench to test this upward compatibility.

Many thanks anyway,

CheerZZZZZ

if overall backwardscompatibility is given I don`t see bigger problems.

Scope is a big modular, if some devices make problems, its likely that it is one or two atoms that don`t work properly. And they simply can fix this by replacing the atoms....

There may be also a potential performance gain if they optimize single atoms for the new hardware and load differents versions, depending on the old or new platform.

Now the modularity of scope seems to pay off.

Woooo,how SICK & TIRED it is,for people to engage in stress all the time,
what do you guys want ?

Flyerfred wrote:John,

One question as you can beta test the new box: specs announce 10x more power, is it compared to a 14 DSP board? And on which operating system are you testing it? MacOS, Windows (XP|Vista), ...Fred

Yes, the calculation is based on the 14 DSP Scope board, so 10 x that board.
For testing, I'm using my Sony laptop, which is just running XP.

john b.

johnbowen wrote:

Flyerfred wrote:John,

One question as you can beta test the new box: specs announce 10x more power, is it compared to a 14 DSP board? And on which operating system are you testing it? MacOS, Windows (XP|Vista), ...Fred

Yes, the calculation is based on the 14 DSP Scope board, so 10 x that board.
For testing, I'm using my Sony laptop, which is just running XP.

john b.

What would be very interesting to check is the PCI-E bandwidth with the Masterverb Test.
I know that it is more a PC test than else, but very essential.

Anyway, great job!

johnbowen wrote:It's real, I'm beta testing it here in Germany!
And it works great, running several synths with lots of DSP room.
cheers,
john b.

p.s. Please remember, it's not Creamware, its Sonic Core now.

I'm sure you know how jealous we all are that you're getting to test it out before the rest of us.

My question is: backward compatibility with old plugins only works on Windows, right? Or do the plugins automatically work in OS X, too? I can't imagine they would...but then I don't know the underlying programming of it all....

Shayne

alfonso wrote:...What would be very interesting to check is the PCI-E bandwidth with the Masterverb Test.
I know that it is more a PC test than else, but very essential....

well, guessing from the 10x power figure it's probably the ADSP-21363 at 333 MHZ they are talking about, speced at 1998 mflops versus the 198 mflops of the original 21065 on current boards...

the 21363 has 6 times the amount of onchip memory, so reverb algorithms may have less reasons to access main memory anyway
there's a calculation rule for delay amounts based on 0.5 mbit per chip for the current Sharcs, so that figure might be extendable by a factor of 6 (if I'm not on a completly wrong track...)

cheers, Tom

astroman wrote:
the 21363 has 6 times the amount of onchip memory, so reverb algorithms may have less reasons to access main memory anyway
there's a calculation rule for delay amounts based on 0.5 mbit per chip for the current Sharcs, so that figure might be extendable by a factor of 6 (if I'm not on a completly wrong track...)

cheers, Tom

Thx Tom , but a.f.a.i.k. you can have delays designed to be calculated on dsp and those designed to work with PC resources, until now the second ones are used for longer times. So I think that the Masterverb should keep working as it does and, while I think the use of more than 3 reverbs is usually a production mistake, it can be a good test for the card/pc interaction.....I imagine that not all the laptops or desktops will have the same efficiency and who knows what the videocards using PCI-e as well or sharing memory can cause on some setups.

As you know, despite the increased power and memory speed etc, no Core2Duo common PCI reaches the Masterverb Test results of a Tusl2c. In fact, the Core2Duo has problems with the STS samplers, that I can't use in my shining new PC. Fortunately the Tusl2c just sits aside, with a Scope Project and a Scope home, that I can fire up if I need the STS.

I think that the brute power is not everything, sometimes the design flaws can bring bad surprises. So is better to have good benchmarks for the PC's and laptops to use along and to know in advance what is recommended.

alfonso wrote:...a.f.a.i.k. you can have delays designed to be calculated on dsp and those designed to work with PC resources, until now the second ones are used for longer times. So I think that the Masterverb should keep working as it does ...

oops, forgot about that subtle, but important detail, hoping it would be kind of dynamic in it's resource request

...and, while I think the use of more than 3 reverbs is usually a production mistake, it can be a good test for the card/pc interaction.....

yes, I second that opinion
but there's this all so nice A100 (the real mean hog, so to say) breathing space into many things and just make them come alive...
If I remember correctly Warp69 once mentioned a few designs in his drawer capable to squeeze the bus even further. Possibly the new chips allow a re-arrangement of some mdules, but well - that's speculation of course.

it's no speculation, tho, that the TUSL still rules after all those years, I'm also happy that I still got one around

cheers, Tom

astro yep,
if there is more dsp memory on board, even if not so big, you can calculate more short delays directly on the chips (DSP delays) and for longer delays you can use the host memory(pc based delays), where the transfer bandwidth should multiplicate by two through a x1 PCIe interface.

So it is possible that we can have 4 times as much delay based stuff than on previous boards...

So I think that will be enough for most reverb scenarios ...

I mean being able to load 8-10 high quality reverbs in one project should be more than enough power for room simulations.