Softimage2014?

[luceric]

IMHO you've always been able to shoot down any release of any software, people just forget.

early versions of XSI? just caching up to what was in Softimage 3d, just now with gray beveled buttons, more bugs and slower.

Animation mixer? stolen from Hash Animation Master and Mirai. subd? stolen from Mirai and lightwave. new RBD? hair? cloth? just bundles of third party libraries.

etc..
these criticisms always work. always. it kind of depends how strong the fanboy factor is to fight them back. I think right now no one cares to go to bat to defend their software anymore, everything is mainstream and therefore you're not part of a special clique.

[gaboraa]

Animation mixer? stolen from Hash Animation Master and Mirai. subd? stolen from Mirai and lightwave. new RBD? hair? cloth? just bundles of third party libraries.

Stolen or not it does not concern us, I consider if we have them or not. Long ago there was a rumor that Maxon had used the same code used in Shave an a haircut but whatever the truth is users got a decent Hair module. Also in the latest release of Cinema 4D they implemented a feature called sticky keys and yes it is the same feature of XSI so who wins? Of course the users. Implementing archetypal features into your own software should not be always considered as theft but it is also a caring.

[Bullit]

Animation mixer? stolen from Hash Animation Master and Mirai. subd? stolen from Mirai and lightwave. new RBD? hair? cloth? just bundles of third party libraries.

That fights between fanboys which almost disappeared, not what happens now. Actually it is more the contrary: the grass is greener in the other side.

[luceric]

I'm just saying the greater cg community could have made these argument against earlier version of XSI if the product had been developed in the world of today. the Softimage community doesn't really bitch that way about its own product.

Softimage did get some of that snark, especially in the early years when we had to bundle Softimage 3d in the box to make up for the shortcomings of XSI. the early demos (ex Jaqua) were not even created in XSI, because there was no modeling or uv editor in it. there was a lot of smoke and mirror in those days.

[Letterbox]

There's an interesting quote from a CTO of one of the big VFX studios, and he was asked what's the next big product/feature. His response was that it's not a product, but a pipeline.

Well then you can tell that CTO from us - he is wrong. Or rather he's right for today, but wrong for tomorrow.
In fact his own words show while he understand the dollars he doesn't understand the sense.

The overall efficiency of ANY 3-d package, is to put it midly, exceedingly inefficent on multiple levels. One only has to look at other software to see that well designed software that fully utilizes the hardware, can and does make people more effective and more productive, and that always effects the bottom line.

What he'd doesn't understand is the software and hardware variables. Given that every one here has upgraded to a faster machine at some point, only to be impressed, ask youself how long does it take you to adapt, ie how soon the feeling of speed wears off, and the productivity gains with it?

The way software today is writtten and the effectiveness of it, given modern multiple cpu and gpu hardware, is to put it mildly in it's infancy. And those hardware costs are falling all the time. Yet conversely development time and productivity gains are taking inversely longer to attain.

For example read http://informatics.indiana.edu/rocha/complex/csm.html

I think that's what Ang Lee means as only after he'd realised what R&D at R&H had to do, and in that realization he saw just how complex the r&d was, and the time it took to perform it took, for what seems a simple task. And that clearly shows how far software hasn't come.

Or simply put, how complex and long it can be to set up a RDB simultion with something like momentum (which is a good piece of software) for what seems an incredible simple mechanical problem. Or read some of the questions here, where it's apparent to the user that the answer is unintuitive. That needs to change.

And in that way he (the cto) is wrong, as efficiency gains in the area of staff performance will always far outweigh the cost of any equipment, to solve such simple, yet invaribly complex issues in 3d apps. That always effects the bottom line.

In those area's software and design choices and so productivity gains, are and have been for a very long time, lagging dramatically behind what the hardware can do.

People are realizing this both directly and indirectly, as can be seen with lower and lower expectations / excitement in new versions of the software, never mind the amount of bugs that are found in them and the time it takes to address them.

[Bullit]

I'm just saying the greater cg community could have made these argument against earlier version of XSI

Understood.

Or simply put, how complex and long it can be to set up a RDB simultion with something like momentum (which is a good piece of software) for what seems an incredible simple mechanical problem.

I would call it a good piece of software for the current bullet technology. Go try to make a wall break visually ok and without the 0 weight trick, you can't.
DMM is probably the next big step. That would be a considerable advancement if the promises are real.

[Letterbox]

DMM is probably the next big step. That would be a considerable advancement if the promises are real.

Not really Bullit, its just another step in the same pseudo-direction.

Take for example the reality given by certain renderers, where you actually use the real-world physical data, for glass, metals etc. Now since your doing that why not add the reality of actual physical properties to that data such examples as...

http://www.keytometals.com/page.aspx?ID=CheckArticle&site=kts&NM=174
http://www.keytometals.com/page.aspx?ID=CheckArticle&site=kts&NM=120

Google "dentritic growth metal fractures Crystalline growth".

Similarly back Ang Lee's (I feel) misunderstood comments - we have material here that have a wide array of animals that show every bone, every muscle, weights etc. Yet all of that material has to be painstakingly (and i mean pain) re-created as none of the physics / dynamics / movement / density / weight work out of the box.

More cpu/core's are here and there's more coming - the hardware can cope - but where is the software to match it?

Without it productivity falls, and everything is getting visually more complex and more realistic. So when people say software is at it's peak and we don't need new features, i can understand that from today's simplistic 'fake it' perspective. But certainly not for tomorrows, not unless we take years have massive staff counts and massive amounts of money to achieve those goals.

[angus_davidson]

I have already gone over my trepidation about the fact that we are nowhere close to where we I feel we should be.

Part of this I feel is because of the following.

1) To many bloody file standards. All the effort that is continually combined to reinvent the Wheel all the time. As an industry we have to say (and just using an example of the top of my head)

Texture Format & Render format (EXR) Make sure that newer things like ptex work 100% with it

The idea being that no matter which software you use there is no need for an import function. Simply File Open.

2) Given that massive amounts of money that is spent on development , why is it that we are still predominately stuck with operating systems that are mostly designed to do everything well except what we actually want it to do.

Windows is designed to work at the lowest common denominator level, Linux is like android fragmented far too much to be of much use to anyone except people with very good TD's. Even my beloved Mac OSX for all its elegance and ease of doing most things, has inferior graphics support, both from a hardware and software level.

We are effectively hamstringing ourselves before we even begin. To take possibly the easiest case. Why isnt there a Linux kernel that is designed specifically to make the maximum usage of the power of machine available to do what it is we want it to do. One that is officially supported by the big Developers, (just certifying is seriously not even close to what is needed). One that we know that if we put in that quadro card its a) actually going to perform at its peak, b) not take 3 days to get it working to begin with. The simple fact that people are getting better performance from Games cards says volumes. Like the file formats we need to have a consistent graphics standard. OpenGL was the original idea. It to became too fragmented and wasnt implemented properly on all systems (looking at you Apple). Imagine how advanced our viewports could be if could have developed them consistently for 3 decades.

In a way similar to how the #vfxprotests are bringing people together to sort out how we interact with the other industry players we really need to rethink why we are using such a poor base to develop the tools that allow us to put food on our tables. Its like turning up at a F1 race in an old datsun and expecting to compete.

rant over for the night. Guild Wars time with the Wifey ;)

[gustavoeb]

To take possibly the easiest case. Why isnt there a Linux kernel that is designed specifically to make the maximum usage of the power of machine available to do what it is we want it to do. One that is officially supported by the big Developers, (just certifying is seriously not even close to what is needed). One that we know that if we put in that quadro card its a) actually going to perform at its peak, b) not take 3 days to get it working to begin with.

Isnt that what old 3d software running on Silicon Graphics and current Flame/Smoke Linux incarnations were all about? While that is great, over time the use of more popular OSs seemed to have one big advantage. Price

[angus_davidson]

Isnt that what old 3d software running on Silicon Graphics and current Flame/Smoke Linux incarnations were all about? While that is great, over time the use of more popular OSs seemed to have one big advantage. Price

There isnt really a price difference, you get windows/mac/linux all together in the bundle and I am assuming there isnt much of a difference in Maya Windows vs Maya Mac if you bought them separately. You are right about the old SGI rigs though. Why they lasted so long when faced by overwhelming Speed from Intels was because they were designed to work well at what they did. What SGI did that was incredibly stupid was they didnt compete pricewise, and when the writing was on the wall speedwise they should have shifted over to Linux running on the Faster Intels. If they had done that their would be an Octane mark 9 on each of our desks now.

We cant change the mistakes of the past, but we can sure as hell learn from them, and do it properly going forward.

[Hirazi Blue]

For those of you anxiously awaiting the next "big" Softimage Technology Preview, here it is:

softimagetechpreview2.jpg

(BTW: that was a joke!!!)

While I agree the yearly "feature hunt" and its faithful companion "feature disappointment" can be quite predictable and annoyingly repetitive at times, I personally do think the blind faith we as customers have to place in Autodesk whenever we renew our subscriptions - as the "classic" upgrade path has been made increasingly less attractive - should permit us to "bitch and moan" come "new release time", even if we are often only repeating ourselves. A "common foe" strengthens a community and as long as the community still keeps on buying the software of this "common foe" no real harm is done IMHO.

[bottleofram]

Drink whenever someone makes this hand gesture:



Happy friday everyone.

Cheers

[Draise]

lol

[Matic]

FYI, information on 2014 launch events, in case it hasn't been posted yet.

http://www.digitalartsonline.co.uk/news ... -march-26/

http://area.autodesk.com/2014unfold

[mirkoj]

jsut fresh from the list.. probably not so new but cmon

http://www.youtube.com/watch?v=L5fOwSmSaW8&feature=youtu.be

and in Softimage we never got even shadows to show correctly.. I mean SI viewport IS ancient looking at current technology. Anyone can even whisper if anything at all is done on that side?
And I will say again.. being silent with new features in software is really retarded.... What like another company will say oh they have this in Maya, Si or something, let's make something like this but better before they release their version!

....

[Nizar]

jsut fresh from the list.. probably not so new but cmon

http://www.youtube.com/watch?v=L5fOwSmSaW8&feature=youtu.be

and in Softimage we never got even shadows to show correctly.. I mean SI viewport IS ancient looking at current technology. Anyone can even whisper if anything at all is done on that side?
And I will say again.. being silent with new features in software is really retarded.... What like another company will say oh they have this in Maya, Si or something, let's make something like this but better before they release their version!

....

not comment about technical side...
But the model, the texture and the overall style showed by this guy is great!

For the rest... why maya look to me like a in incredible mess? guess I'm too used to softimage.