PDA

View Full Version : hmm...


xsitar
02-05-2007, 05:52 PM
so i had some thoughts recently. not trying to offend any companies here, but i'd like to hear some of your opinions about this.

looking at blender it seems so quickly to adapt all the highend features from apps that cost hundreds or thousands of dollars. i don't say it's necessarily all exactly as powerful. but for a free program it's damn fine and/ or getting there.

blender has sculpting tools, lscm unwrapping that even xsi doesn't have. it has an integrated compositor that seems almost better than xsi's and the animation tools shouldn't be too far away too.

and i thought a open-source program like blender should motivate those other companies who have much higher budgets to come out faster with upgrades and more innovative features. but unfortunately it seems not so. maya & xsi have kind of been slow in that regard while smaller companies with much less money do the cool stuff (nevercenter, luxology, skymatter)... and even a open-source program like blender seems to come closer.

so what do you think?

3dtutorial
02-05-2007, 06:05 PM
so i had some thoughts recently. not trying to offend any companies here, but i'd like to hear some of your opinions about this.

looking at blender it seems so quickly to adapt all the highend features from apps that cost hundreds or thousands of dollars. i don't say it's necessarily all exactly as powerful. but for a free program it's damn fine and/ or getting there.

blender has sculpting tools, lscm unwrapping that even xsi doesn't have. it has an integrated compositor that seems almost better than xsi's and the animation tools shouldn't be too far away too.

and i thought a open-source program like blender should motivate those other companies who have much higher budgets to come out faster with upgrades and more innovative features. but unfortunately it seems not so. maya & xsi have kind of been slow in that regard while smaller companies with much less money do the cool stuff (nevercenter, luxology, skymatter)... and even a open-source program like blender seems to come closer.

so what do you think?


.....ummmm nevermind, I withdraw my comments on account of the fact I don't want to take part in a btch fest.

Sorry.

Sil3
02-05-2007, 07:17 PM
and i thought a open-source program like blender should motivate those other companies who have much higher budgets to come out faster with upgrades and more innovative features. but unfortunately it seems not so. maya & xsi have kind of been slow in that regard while smaller companies with much less money do the cool stuff (nevercenter, luxology, skymatter)... and even a open-source program like blender seems to come closer.

so what do you think?

What do i think? I think that you are seing things in a wrong way...

First of all: INNOVATION


Tell me a feature that Blender has that is NEW to the 3d World, i dont recall none from my mind, but in all this years i "know" Blender they were always playing "catch up".

- Hey Package X just got Feature Y, lets see if we can code a similar feature one too.

This has been the "spirit" of Blender so far.

Innovation cost's money, after something is done it's easier to copy/mimic it, the really dificult part is inventing it in the first place.

I can be completly wrong, but Blender Devs seems like a bunch of Students playing software engineers because they like it, no one is forcing them thats for sure, they are doing it for the thrills, dont know if any of them does this full time, but withou money i kinda doubt it.

On the other hand, the software devs from the commercial packages are people like you and me that need to get payed or else they wont work. Companies need to have products to sell to have income to pay their workers, it's impossible for a company to not release a new version of their software to make cash flow during certain timeframes, and majority of the times, like it happens in what we do, sometimes it's impossible to do better/implement all what they liked it in that specific version.

We all would love all the cool features and more in all new releases, but it will never happen, some features have priorities over others not to mention that a day only has 24h and the human factor at those companies is a finite thing, they dont have acess to all the people in the world to help them out and even if there were people glad to work for free for those same companies, it's an impossible scenario too... companies protect their code and none would give it away for the sake of free workers...

About Maya and XSI no being inovative enough in the last years... are you using the same XSI i am? Do you have a clue how much work must Alias programmers had to make Python even more acessible to all of Maya features like MEL was? Might not look like something big after all, but in the long run it will be, of course you might say that you didn't need Python in Maya, well no one is forcing you to buy the latest version after all.

At the end i think we all should be glad some cool blokes decided to use their time to give away free tools (Wings 3d, Blender, Gimp, Star Office, etc etc) that very very closely match the big bucks ones, but then using this argument to say that the payed ones are being lazy or not innovative is over the mark IMO.

toonafish
02-05-2007, 08:17 PM
I've tried it but never seriously used Blender so I can't judge for myself how good it is.

But I suppose it's a good stepping stone for starting programmers and students that will have to make a buck or two at some time in their life to support their family and have a beer with some friends.

So that's when they start working for companies like Soft, and suckers like us pay big bucks for stuff we could also get for free ;-)

xsitar
02-05-2007, 09:14 PM
sil3 i agree with the most of what you've said.

but it's usually the case that one company innovates on one part and the others follow with that.

the first time i've seen the lscm unwrapping was in modo (i'm not saying they invented it). but that was like a year ago when they introduced 201. - how come blender has it in there before xsi? and since that year xsi could have been upgraded with such a feature...

again the sculpting tools that we know from mudbox. it's going to be soon in silo2, modo as well (of course not meant for high-detail as mudbox. but close or at least medium-res). it's already in blender and i've tried it, it's pretty good actually (again of course not as high-detail oriented as mudbox or zbrush). where in xsi? nowhere.

i don't even expect softimage to be super-innovative always. but i was just wondering how come am open-source app implements such features sooner than a much bigger app/ company like xsi/ softimage.

to come back to the lscm example again.

the pelt-mapping idea for xsi has been around since 1 or more years (with helge's addon). why didn't softimage implement the lscm feature in the te-editor since then, in a possible x.5 upgrade? now it seems like we have to wait from siggraph to siggraph. - not that i'm disappointed with the v6 features, but a lot of "features" are just minor upgrades (look at the shrinkwrapping, material viewer, etc.) to existing stuff.

SheepFactory
02-05-2007, 09:49 PM
i don't even expect softimage to be super-innovative always. but i was just wondering how come am open-source app implements such features sooner than a much bigger app/ company like xsi/ softimage.



It's because a open source app does not have to support thousands of customers and productions and make sure the features they add dont break the workings of the app. Softimage , Autodesk etc have to worry about these things unlike blender and wings 3d. Sooner or later it will get those features and when it does I am sure it will be great. Meanwhile what is stopping you from using blender or whatever to do your uv's in and export to xsi. I dont see what the problem is especially considering blender\wings 3d is free.

attanze
02-05-2007, 10:06 PM
... Tell me a feature that Blender has that is NEW to the 3d World, i dont recall none from my mind, but in all this years i "know" Blender they were always playing "catch up". ...


Well, actually Blender have at least one feature no other 3D software package have: A Video Editor!

Im mainly a Blender user, but, beacouse Blender isnt perfect (at this point of time) Im evaluating XSI to integrate into my workflow.

Per-Anders
02-05-2007, 10:10 PM
I would suggest you try using the application before evangelising it.

Feature checklists are just that, they don't tell you how well something was implemented, nor it's workflow or design.

Opensource can introduce features apparently quickly because all their users are alpha/beta testers, there's no roll out or dealing with marketing or having to get manuals written and printed or any real logistics of anykind, all the stuff that slows down a more traditional softwares release if the company hopes to ship a product and have support that will be of an acceptable standard.

Additionally while the previous post about the merits of innovation was rather impassioned it's true to say that once the mold has been set it's easy enough to make a copy. It takes a great deal of effort, research and development to come up with a truly good new concept or idea, but relatively little effort to copy that idea. You can copy what's already been done very quickly because the model is already there, all the problems faced, the design issues, the workflow problems, they've been resolved for you (though somehow that still doesn't stop people getting it totally wrong...).

Per-Anders
02-05-2007, 10:20 PM
Well, actually Blender have at least one feature no other 3D software package have: A Video Editor!

Well you'd be conveniently forgetting the dawn of consumer 3D then with the VideoToaster package, originally Lightwave was just the transition rendering component of this.

It's not as if XSI doesn't have a compositor which allows basic editing in there either, or as if editing hasn't been around for years either or used in 3D, so not what I'd call an innovation or "new" and certainly not bringing anything new to the world of 3D.

xsitar
02-05-2007, 10:27 PM
agreed per-anders.

but i wouldn't say that a feature like lscm-unwrapping would upset a entire pipeline... (i would even go so far to say that sculpting tools wouldn't too. it can all co-exist)

so if "motor" & "delta referencing" is being introduced between siggraph '05 and siggraph '06, why can't this much smaller feature like lscm not be done too?

xsitar
02-05-2007, 10:35 PM
by the way. just to make sure. i didn't create this thread with the intention of flame-wars like who did this or this first. because that doesn't really matter. i was looking for some insight/ opinions. thanks for keeping it clean here.

xsitar
02-05-2007, 10:46 PM
@sheepfactory

maybe because i'm tired of exporting/ importing my .obj's...

in most cases poly's here and there get messed up or are missing. and it's just annoying to fix it.

MikeMD
02-05-2007, 10:54 PM
It's easy to add features. It's hard to make them work well.

Just because blender has feature x or y or z doesn't mean it's any good.
Example: xsi, maya, truespace, lightwave, all have bones, some of these you can actually work with, some not at all ( guess which is which ). Same goes for blender. Features mean nothing until they work the right way.

If blender were as good as xsi or maya, nobody would buy xsi or maya, so..............

xsitar
02-05-2007, 10:58 PM
mike. again if you read most of this discussion, i wasn't saying that all the features are yet as good as the highend-apps. but in a lot of cases it's pretty close or it has even more features. i'm not only surprised that blender has sculpting tools but also that it has a fluid-simulator that is much better compared to xsi's blobby stuff.

and specificly talking about the lscm feature (which i was pointing at since the beginning) it does actually work as you'd expect it (like modo's). such a feature is not the biggest thing in a 3d app but a small one that can ease up a lot of tedious work.

withanar
02-06-2007, 12:08 AM
If there are features in app A or B that aren't in XSI or some other app of choice, it's because they didn't have enough time or resources or desire to get around to it. Nothing more.

There are only so many developers at Soft, with only so many areas of expertise, and only so much money to allocate, and the "to do" list in this industry is growing faster than the amount of revenue supplied to drive it.

Blender has a potential dev pool of... anyone who wants to contribute. This is great when somebody wants to test a new algorithm for unwrapping UV maps or something of that ilk. Not so great when design consistency is of concern, or when you're in a production crunch and need debugging help.

Mic_Ma
02-06-2007, 01:00 AM
I haven't used Blender, but I have used other 3d apps. "Features" can be livesavers but "implementation" is at least equally important. A new way of unwrapping something is hardly a feature that will save much time in big projects. The bottlenecks are in other places.

I think there is a big difference between how hobbyists, freelancers, small and big studios use software and attack projects. As the tasks grow and multiply, managing your dataflow becomes more and more important. In the end features are just a little part of a whole puzzle. While a new something-shader might justifiably be crucial for the small developer, it wouldn't raise an eyebrow in a large team. It's all about scale I suppose.

AngelDream
02-06-2007, 02:30 AM
I've used Blender for a while, but I can't say it's suitable for a pipeline. Sure, there's the OpenMovie Elephant's Dream and the upcoming argentinean short Plumíferos (they're using stuff like Cinelerra, Gimp and so).
I think the main problem of OpenSource is the human interface, the usability, how user-friendly the software is, and of course, it's integration on a pipeline. If we talk about features, Blender has more features than LW, including fluids, something that LW maybe never will. But it's not user friendly, neither it let you develop a nice workflow.
The same goes for GIMP, Jahshaka, Audacity and others.
However, Blender is growing into a more evolved app, we'll just have to wait.

xsitar
02-06-2007, 02:36 AM
but come on. a lscm feature is so small compared to other tools in 3d. within the past 2 years this could've been easily implemented. it's not like it would tear apart the entire package just because you can select edges to unwrap the uv's.

i just maybe don't really understand why this couldnt have been possible since there's such a big price difference between xsi and blender that doesn't cost at all.

and it's not just about resources & time. i'm talking also about quality. one might expect to get better quality for money. i've tried the lscm unwrapping in modo. and to my surprise blender's was instantly giving me perfect results (not kidding). while in modo it required me to drag my mouse around and not even after many approaches it got as clean as in blender. and modo is around 600-700usd.

i guess i'd just like to see a little more pressure on the commercial-packages-companies.

Per-Anders
02-06-2007, 02:52 AM
You could always try and create the feature yourself. After all the tools are there to do so and it is as you say a small thing.

Sil3
02-06-2007, 08:27 AM
but come on. a lscm feature is so small compared to other tools in 3d. within the past 2 years this could've been easily implemented. it's not like it would tear apart the entire package just because you can select edges to unwrap the uv's.



Again it's an issue of priorities, i would love to have a LSCM UV Unwrap in XSI like Blender, same way i would love to have a real symmetry modeling option ala Modo and LW.

Even this things are for sure "in the list" of features at Softimage Dev desks, they must have other "more important" things to implement/improve the workflow.

Do yo think that a Gator, Motor, Ultimapper, Animation Layers are such trivial things to implement? This things have taken many many months of code writing, testing and debuging until they were production ready, not to mention fixing Bugs and still improving older features.

Yeah sometimes i find myself thinking...man... how can this not be implemented already? Everybody has it by now... LSCM is one of those thing, but it's not there, so i'll use Blender that has an amazing LSCM feature. Also, i would not trade XSI for Blender :)

Mic_Ma
02-06-2007, 08:38 AM
It reminds me, I did a bit of game testing a while back and we had over a thousand bugs, trying to keep track of them, rating them by severity, etc. Not to mention that fixing a bug might introduce another one.

I guess it is the same for any software dev work. And they have to update the documentation, translate it into various languages, etc, etc. I am sure it is easy enough to hack a feature together, but a big headache to implement it properly, especially if it is something users can access in another app.

But I agree. I'd like to have those nice uv-thingies too.

:o|

toonafish
02-06-2007, 09:42 AM
Do yo think that a Gator, Motor, Ultimapper, Animation Layers are such trivial things to implement? This things have taken many many months of code writing, testing and debuging until they were production ready, not to mention fixing Bugs and still improving older features.


Don't forget Face Robot..


runs for cover... :scream:

attanze
02-06-2007, 10:29 AM
I've used Blender for a while, but I can't say it's suitable for a pipeline. Sure, there's the OpenMovie Elephant's Dream and the upcoming argentinean short Plumíferos (they're using stuff like Cinelerra, Gimp and so).
I think the main problem of OpenSource is the human interface, the usability, how user-friendly the software is, and of course, it's integration on a pipeline. If we talk about features, Blender has more features than LW, including fluids, something that LW maybe never will. But it's not user friendly, neither it let you develop a nice workflow.
The same goes for GIMP, Jahshaka, Audacity and others.
However, Blender is growing into a more evolved app, we'll just have to wait.

The next version of Blender (2.5 and not 2.43 who will be released this days) will have a refactored UI, with customizable shortcuts and more customizable UI. I want to sow the same degree of UI customization into XSI.

Kabab
02-06-2007, 11:55 AM
Think of it this way..

Your the development manager at Soft and you have 2 choices to make..

1. Spend time developing a tool which is in a free competitive application

or

2. Spend the same time developing a tool a studio has requested and is willing to pay money for or buy more seats.

One option is going to generate revenue and one won't, after all these are publicly listed companies their #1 goal is to make money!

I think some of the reasons open source won't take of is, commercial software is actually fairly cheap, well supported and has a large skilled userbase.

If you go open source your are basically going to need more programmers etc on hand to fix any issues with app as you go a long and if something is drastically wrong who do you call? Commercial app come with support which covers this and its very unlikely that Soft/Autodesk isn't going to exist in 10 years time can the same thing be said about any open source package?

1 good programmer = $100k per year
50 seats of xsi/maya = $100k

So if you can save on just one programmer buy using commercial app's you have paid off your investment.

Software costs are usually far cheaper then people time its all about saving the man hours.

Unkelben
02-06-2007, 06:42 PM
I didn't know about the Blender sculpting and painting tools... I tried it at once and altough the texture painting looks pretty stable, the sculpting crashed 3 times in less than an hour. Of course, this is still a beta release.

I agree that 3d painting and sculpting would be very nice in XSI (along with pelt mapping) as you can't do without these tools anymore and it's really a hassle to do all the importing and exporting.

luceric
02-10-2007, 02:27 AM
Ho I get what this thread is about.

It's about XSI not having lscm unwrapping and complaints about that hidden behind some thread that purports to be some discussion about open source vs commercial.

XSI doesn't have lscm unwrapping, deal with it.

Instead in XSI 6.0 the team has worked to enhance the existing UV manipulation tools, in collaboration with a couple of large game customers who had really big issues with the texture editor. They also had huge issues with material managements in XSI, and the team also worked on that.

The softimage team works IN THE REAL WORLD where XSI is used in a pipeline. The XSI clients the team works with do not want a new type of unwrapping *before* the texture editor's basic function are fixed, or if the material management sucks. Or if they can't transfer scenes between Max and XSI, etc. If they need a feature of modo immediately, they buy a copy of that.

XSI is not about buzzwords, adding some bullet point to a product just because some other products has it. XSI 6.0 fixes things that BLOCKED clients completely from using the product.

It's easy to add bullet-point features to impress people, but very few people actually use the product and check for themselve if the stuff works as they think it does.

From what I've seen of the compositor in Blender, I'm surprised anyone thinks the compositor inside it is 'almost better than the one in XSI'. It might look better on paper, but it's still just starting. It's VERY easy to start something, to make a basic compositor that does 'over' and color curves and writes out the result. It's after that that everything becomes more complicated and expensive to develop. At the siggraph booth, it wasn't yet working even though it was advertised, and people were already saying it would kill shake based on the screenshot alone. It's all about getting a good screenshot these days.

The XSI team also worked for months on very boring and un-sexy things in the Animation Editor and Dopesheet that nonetheless BLOCKED many users from using or being productive with XSI.

It's all in the detail, you can't tell that from a web page feature list.
Unwrapping is something many people want, but it wasn't the top issue people have with XSI 5.0.

NRG-Alpha
02-10-2007, 04:56 AM
Wow. Threads like this seem to be just begging for high tension and flame wars. I'm actually impressed that it has not depreciated into a complete lock down.

I think it definately boils down to basically commerce economics vs freedom of developing whatever you want. Anyone in their right mind can understand that commercial public companies have obligations to their legal userbase as well as its shareholders. Freeware doesn't have the restrictions of time, bugets, schedules, etc.. As a result, apps like Blender and Silo can freely develop any number of feature sets, as these programmers are not bound like their commercial counter parts. Therefore, having said that, I am a true believer that commercial apps are at a disadvantage to some degree, simply due to the restrictions mentioned above.

Having said all of this, I think that it is ultimately (in all likelyhood) inevitable that sooner or later, XSI will have no choice but to include, say pelt mapping for example (it seems only natural for a 'character package'). The demand will arrive sooner or later in my opion. In the meantime, I would advise resorting to external packages that are extremely strong and specialised for such features (Headus' UVLayout comes to mind). Yes, it is a pain in the ass having to start including external specialised apps to perform some of the functionalities that you want but XSI does not yet have. Until XSI gets such features, and you still want and use XSI, there doesn't seem to be really any other alternatives.

Rest assured, it's obvious that Softimage is aware of what their customers are asking for (large and small). I suppose it is only a matter of time. Again, with what resources are available within Softimage, we have to understand that there is a delicate balance between important flaw fixes, wish list items and the like. Be patient. I hardly doubt that a company like Softimage is deliberatly ignoring wish list items to piss people off. In the meantime, there are other alternatives.

-NRG

luceric
02-10-2007, 05:03 AM
NRG: Silo is not freeware, it's a commercial app. it's an application that only needs to do one thing correctly: modeling.

Also, if any programmer decides to develop something for free for an open source app, they could also consider developing the same thing for free with the SDK for a commercial app. Commercial applications are not limited to what the internal programmers can do.

NRG-Alpha
02-10-2007, 05:53 AM
NRG: Silo is not freeware, it's a commercial app. it's an application that only needs to do one thing correctly: modeling.

Also, if any programmer decides to develop something for free for an open source app, they could also consider developing the same thing for free with the SDK for a commercial app. Commercial applications are not limited to what the internal programmers can do.

Oops. I stand corrected. You are right. For a moment there, it completely slipped my mind about Silo. It is definately a commercial app for sure. I suppose the advantage about Nevercenter is that not a public company. So once again, the restrictions that bind companies like Softimage and Autodesk may not apply if this is indeed the case (especially if Silo developers work on Silo on the side and do not solely rely on sales to support themselves, and that they themselves are the only shareholders (not sure about their circumstances though)).

Yes, correct about programmers being able to develop those features for SDKs. I suppose the question is whether they want to or not.

'Commercial applications are not limited to what the internal programmers can do'.. I suppose you are still refering to 'external freeware' programmers and freely developing for commercial SDKs. In this case, yes. I agree again. But again, it all boils down to whether someone out there will do it or not. Personally, I'm not going to hold my breath on it. Besides, I would rather see it done at Softimage at some point in time. Stated simply, to me this means that it will ship out of the box as opposed to users having to hunt down and install such features. I personally frown upon companies relying on freeware programmers to add value to their apps (renderers aside.. I am speaking of feature sets linked to modeling, unwrapping, etc.). This is just my opinion though.

-NRG

el_diablo
02-10-2007, 08:03 AM
I've been playing with blender alot recently.

Blender does have sculpting in the latest beta...but what can you do when you finish sculpting your model? In blender only render it out as a still. Blender doesnt have means to do a displacement map from lores model, neither a adaptive subdivision displacement rendering. Blender has fluids...but it does'nt have the comparable featureset of Realflow or Glu3d or whatever. So you are down to basics.

Some of its features are just small programming projects added to blender. Lot of loose ends. Some things work some don't. Amateurs arent bothered about that, they have a nice 3d program for free.

As soon as you try to use it professionaly...you see only problems and unfinished implementations.

Still, its a nice addition to a pipeline. For one thing i seem to use its modeling functions more and more...

janimatic
02-10-2007, 11:06 AM
missing feature like LSCM would be annoying if that was breaking my workflow but thanks to xsi|o, it's not the case... I think solid bridges of all kinds are a very good answer to that problem, because you take the best tool for each task.
The advantage of opensource is the amout of free developpers but the obvious advantage of nevercenter is that silo is focused on 3 things only, so they had a chance to collaborate with the right person for this specific LSCM problem.

CGTalk Moderation
02-10-2007, 11:06 AM
This thread has been automatically closed as it remained inactive for 12 months. If you wish to continue the discussion, please create a new thread in the appropriate forum.