Has Blender really made zero progress to attract professionals since 2.4x?

General Forums Blender and CG Discussions
21

How is this anything other than a false choice? The things that will make Blender more attractive to “professionals”, however you want to define that, will also make it better for everyone else.

The value of the open movie projects isn’t that they promote Blender. Their value is that they put artists, TDs, and developers in the same room and force them to solve real production problems. I don’t think it is a coincidence that some of the best software in the industry started off as in-house tools at big studios. The best tools are always born in production.

I don’t know I see some very legitimate well articulated criticsims in that thread. I’m probably biased because they happen to align with a few of my reasons for using Blender less than I would like.

1. Poor view port performance.
Assets aren’t getting any smaller and watching Blender struggle with a couple million polys while the devel list explodes at the mere suggestion that support for a 10 year old operating system be dropped is a little discouraging. Meanwhile modern apps like Clarisse are pushing billions of polys.

2. Doesn’t play well with others.
The somewhat unconventional interface and related training costs certainly play in to this but there are also technical difficulties incorporating it in to a pipeline. FBX compatibility has been less than ideal for quite awhile and it is great to see it getting some love. I know there are challenges supporting proprietary formats like that but what about Alembic, Partio, OpenVDB, and OpenSubdiv/Ptex? The difficulty of integrating Blender in to existing pipelines is something that gets brought up fairly regularly on the development list.

3. Fear of removing a feature, ever.
Seriously, have you ever watched a new user try to setup texture painting? Sometimes less really is more.

4.Jack of all trades master of none.
Although I see where the critics are coming from, I disagree about this one a little bit. Although it does contribute to it’s complexity, I think the all-in-one nature of Blender is one of its greatest strengths. On the other hand, there are a lot of areas where Blender almost works great but fails in some subtle but important way. This wouldn’t be as much of an issue if 2 were solved.

Fortunately, all of those issues are fixable and I’m hopeful that the BF will put some more emphasis on them in the near future. Blender is far from perfect, but it is a great example of how good desktop open-source software can be.

22

The yearly subscriptions cost 1/3 the software price… and let me give you an example on what the professional support gave us…

In the studio we had our own Motion capture setup… the mocap device has its own software which sends the captured motion data to Motionbuilder. The motion builder uses biped system for mocaps (you can also use a standard rig but won’t be much help outside motion builder). The reason we use biped is it has built in tools for mocap and retargeting.Also it can copy poses for bones, which you can’t do with standard bones. However our animator cleans up and edits these motion capture data and sometimes adds some custom movement like arm and shoulder repositioning.

In one of these situations, we couldn’t get the biped skeleton clavicle bone (shoulder) to rotate on x axis… the original motion had to be edited to get the correct upper swing of the arm but couldn’t be done… After our chat with AD support we were told that biped skeleton had a build in constraint on that bone (clavicle) only for that axis… “Uhm yeah you pay us top dollar so that we can put hidden constraints on bones”

That support was useless, it was the only time we used it and with no benefit…

We had to make a custom skeleton which the animator would use to edit animations; we baked the animation from biped skeleton to custom skeleton so the animator could edit the animation and export from there…

By the way Blender has a system for supporting pipelines , you find a technical artist or a programmer from Blender Network and hire them for pipeline integration…

23

I am not sure if I should participate in a flame war, and I could not disagree more with @yii7 about Andrew Price. He has helped as many people learn blender as anything on the Internet, and his “hobbyist” attitude embodies what blender stands for. Open to all, and free to learn via other artists good graces. He consistently punches out as excellent of quality artwork as blender is capable of.

That being said, the reason why Blender is not seen as a professional grade application is very easy to understand.

1. Lets Sculpt

Blender has sculpting tools and you can sculpt just about anything in it. But I use Zbrush 4r6 and I can tell you that the ease of use, the natural artistic feel and various features of Zbrush blow Blender out of the water. Even still, I ENVY Mudbox users because Mudbox probably has the most awesome and intuitive texture painting system out there. Much much better then Zbrush spotlight.

Both apps are affordable for any artist achieving even modest success and if you want cheaper there is 3D coat. 3D coat has the most amazing Re-topology tools. IMO it is much better then either Zbrush or mudbox in this regard.

2. Let’s Animate

3DS max, I believe most people agree, does not deserve it’s reputation as “the” great application of choice. It’s a modeler and all modelers are about the same because modeling has the same basic tool sets. So Blender is as good of a modeler, but 3d graphic arts do not stop at modeling.

Maya on the other hand really is an amazing application for producing animations. And rendering those applications in Mental ray, Vray or Arnold on a server farm is going to do much better than trying to do the same thing in Cycles. I have seen people on this very forum rig and texture models in Blender only to render it in 3ds max Vray for this very reason.

I don’t have allot to say about Cinema 4d because I don’t know much about it, but have you heard of Modo? It has an excellent built in render engine and an extremely intuitive set of modeling tools. It’s like Blender only smarter. That does not mean that you can make something in Modo that you cannot make in Blender, it just means that the work flow can be smoother.

3. Let’s Render

Even now Blender to Vray is not done, as I understand it it is still very buggy. Octane for a long time needed an external exporter script, and even though recently they have added an internal Octane render for blender it ultimatly required Otoy to create their own version of Blender itself to make it work. It is not a “plug in” but a fully modified version that you have to download because blender trunk could not support it. As far as I know maxwell is the same way and it still has to run on a script, and thus maxwell fire will not work inside of Blender.

Even if you love Cycles, it is missing some very important render passes. Compositing is what studios do and not having them is easily a deal breaker, regardless of the quality of the actual engine.

4. Lets Make A Game

Above all else, this makes the least sense. Forget about Unity the Cryengine is free, and why on earth would you want to run a game in blender when you can run it in Cryengine? As I understand it you only have to pay money if you sell your game and it starts to make money, you will then only own them a reasonable % of your earnings. To my understanding if you don’t sell your game or distribute it as a “demo” then you own them nothing. If anyone knows better feel free to correct me if I am wrong :smiley:

Thus…

Blender being free has nothing to do with big studios not using it. They have sooooooo much money that purchasing multiple Maya licenses is the normal. Professional artists can earn enough in a few gigs to purchase their version of 3ds max and Vray so that is not as big of an issue as people make it out to be either.

Not be insulting, but it is the non professional hobbyist that does not generate significant revenue with their artwork that wonders “why doesn’t everyone use this because it’s free?”. If you have the money then it does not matter if the software is free or not.

Give blender texture and painting like mudbox.
Give it a dedicated and intuitive re-topology section like 3dcoat or topogun
Take a look at Maya’s animation features… we want them.
Make the blender trunk capable of more fully interrogating with the most widely used render engines.
Hello…Alembic?

…and then, Blender would be want everyone wants it to be. I thing that you could not easily look past or disregard. It would be a thing that everyone would want on their desktop regardless of what their primary software was. It be like “I animate in Maya and render in Vray but I texture paint in Blender” or “I sculpt in Zbrush but retopo in blender because it is so much easier and intuitive”. And so on and so forth.

24
  1. Blender can sculpt 68 million vertices real time… if you have a capable computer zbrush can go 25 million with uber lag…(not HD sculpting thats a different issue).Outside sculpting the Opengl improvements aside I think they are going to optimize it.

I remember back on 2010 you could barely open a 700 k decimated sculpt object in max. About Clarisse i dont think it can push billions of polys real time.If they are instanced maybe, but can you show an example ? It is definitely a cool software though…

  1. OpenVDB in development, Opensubdiv next on the list… Alembic,partio and Ptex well don’t know about those…

  2. Really ? Try Max Viewport Canvas…

  3. Disagree completely…

@Jon Smith

I am not sure if I should participate in a flame war, and I could not disagree more with yii7 about Andrew Price. He has helped as many people learn blender as anything on the Internet, and his “hobbyist” attitude embodies what blender stands for.

Well I already stated it is my opinion, you can disagree however you like. I never claimed his tutorials helped nobody or his hobbyist attitude was a wrong thing… However I presumed he was a professional since he constantly writes articles about professionalism or issues of of a profession (as a 3d artist). He created a paid tutorial service called “Nature Academy” where he tutors other people.

So his over exaggerated announcements , podcasts or videos like “Why is this broken ?” , “Why doesn’t professionals use blender?” seem very slapdash to me. Frankly I have to say there are more subtle ways to talk about this stuff without a making a scene and hurting the image of Blender.Again I would like to point out this is my personal opinion of his actions nothing more. But you can see what I mean by his Blender Conference speech title “Why People Laugh at Blender” . I wonder who is laughing at who ?

25

key for attracting professionals alike new people is and always will be first and most needed an outstanding incredibly easy to understand good documentation and presenting good stuff. <- in that order

video is good but as an example I get more than frustrated to be forced to hear nonsensical/private stuff of the presenter for the first four minutes, to see one minute where the information sits in.
meaning? in some cases text and image outperforms video.

text and images should get a priority if it comes to showing node systems.
video can be good for showing work flows but it needs to be crisp and tedious tasks to be time lapsed and SPEAK clearly don´t smoke or drink during the process and explain what you do,
good instructor for that would be Kent Tramell.

bad instructor would be Wayne Robson.
he is really skilled but his smoking and drinking during his tutorials is making me either get frustrated about him for doing this or simply ignoring his stuff which is actually a loss.

cycles is on a good way of getting used, many people are presenting good stuff but it could really benefit of a well documented node manual for “real life” cases
about skin shaders, fabrics, wood, metal, carpaint, hairshaders and so on. a “real” manual if you would like to say…
such a manual must be free without charge because every more serious rendering solution has such a document or even goes so far (not needed in my opinion) to give you a huge library of materials, to quickly get to speed.

at the moment I am looking at cycles and it baffles me on how to quickly get to where I need to be.
want an example? the sss skin shader.
it gets mentioned with a pretty picture in the to be release of 2.69 but there is no example of a node system, what so ever to give a user a hint on how to get to such a result.

until that changes even cycles which seems to be on par or even superior to some otehr gPU/CPU rendering solutions out there,
(hey cpu and gpu rendering and no node locked renderer? a hair particle rendering solution and sss how AWSOME!? but what a bummer ! :confused: where is the manual for the materials?)
people wouldn´t use Blender and cycles simply because they don´t know where to look for helpful solutions and even this forum which can give fast answers, could be to slow in some cases.

good examples needed on how it could be made?

http://www.topogun.com/Docs/index.htm

Octane for Blender
http://render.otoy.com/manuals/Blender/
rudimentary manual for nodes so far but it follows a promising path of teaching you, on how to archive stuff and I guess the node part of that manual will be massively documented.

Zclassroom


even if it is all in video format it shows a viable direction- the best so far for videos.

giving information quickly to get to speed with the program in all necessary topics from beginner to intermediate or even pro is crucial for the success of a program.

btw Greg Zaal
has an awesome website for Cycles nodes which is what I would like to be taken much much further becuase that was and is a real helper.
http://adaptivesamples.com/

so in short;
to really attract people; give first good free documentation for all areas to an intermediate level even professional level out.
People need to be able to quickly dive in without the need to pay for external tutorials or even worse to search for tutorials
and then show them cool pictures,movies to proof on how cool the program is.

26

I wanted to post a long point by point explanation about how I started a BIG commercial project in Blender involving fluid sims, heavy geometry and HD render, and needed to switch half-way to Softimage to be able to finish it. But let me just link one of the posts I made after it was done.

http://www.blenderartists.org/forum/showthread.php?300645-Finishing-a-big(er)-project-in-Blender-here-are-my-findings-bug-reports

I must note thats not all, there been many other showstoppers along the way I haven’t written down.

As far as am I a professional (for the relevance to the original question), I have about 200 TV commercials behind me since 2001, yes they were local and not at the same level of production as those that come out of Glassworks or TheMill, but they still used same post-production/animation principles.

27

You can get just as much in Zbrush if you know what you are doing, the approach is different though. That said though, a higher poly count doesnt necessarily mean better details or a better sculpt. This is why the polycount issue is meaningless.

Take an alpha and plug it into both zbrush and blender, adjust the settings so they are as identical as you can get and then see the difference in quality. Zbrush wins hands down, even if the mesh in blender were twice the poly count it still comes out more accurate and with finer detail in zbrush.

The point is that things like polycount dont make a difference if the end result isnt what they need it to be.

This is pretty much one of the main points in Andrew Price’s videos…and to be quite frank it didnt start with him but he is doing a good job getting idea out there.

That idea being… its about the goal. If I need to sculpt a detailed mesh easily for use in baking normal maps, the only thing im going to care about is getting the best result as quick as possible with less interface tweaking. Thats the issue we need to focus on. The goal. If blender isnt achieving that goal compared to other apps or its not making it as intuitive and easy as it should be, then it wont be used. Its as simple as that.

Another good example is the UV layout. Blender has a great UV toolset, but many, even blender users, will use alternative programs just to get the best results faster and with less work. I would rather use Roadkill UV than Blenders UV editor…and those are based on the same code…yet the interface and controls are different. One is more end user friendly, it focuses on the goal.

So his over exaggerated announcements , podcasts or videos like “Why is this broken ?” , “Why doesn’t professionals use blender?” seem very slapdash to me. Frankly I have to say there are more subtle ways to talk about this stuff without a making a scene and hurting the image of Blender.Again I would like to point out this is my personal opinion of his actions nothing more. But you can see what I mean by his Blender Conference speech title “Why People Laugh at Blender” . I wonder who is laughing at who ?

You may say that, but I know some fairly big vfx supervisors, those who have worked on major hollywood titles that know of blender and even of Andrew Price. The few who do now of him say good things about his tutorials and opinions… so its kind of funny when these people see actual blender users attack him for stating the obvious. Many in the industry are watching Blender, seeing how it developers, whether it goes in the right direction or not, whether it helps them achieve their goals better or not, is whats important… The ones with the most experience generally carry more weight than some teenage hobbyist in a third world country. I think we need to keep that objectively in mind. Blender uses money, given in donations and via products/dvd sales to fund itself, it uses that money as well as the community to further its development. Those who can offer the most in development and money, because it makes the tool more important for their livelihood, are critical to the growing success of Blender.

Never forget that one of the best things about Open Source is that its not beholden to any to the point where it can never change. One of the strengths of open source projects is that it can change and more drastically than some commercial entity with those paying for as little interface change as possible.

28

The killer is the continuously changing plugin API which prevents the formation of the third party ecosystems which are the real power of 3DStudio or Maya or SketchUp. Pros rely upon these plugins automating their tasks and they want to keep having being able to use them in time (no plugin rot).

Come on, Blender 2.69 has no Greeble plugin (2.49 had it)… do you expect a 3D artist to build greebles by hand in late 2013?

29

Actually, yes. You make your own greeble packs and import them into your work as you need - detail depends on the size of the object. The Discombobulator was okay, but it was visually obvious when used in blender projects, mine included. Watching JDaniels’ videos and checking out SyFy Meshes forum, I saw a lot of prep work done on greeble packs that were used to kit bash - and like I said, detail and style depends on the project at hand.

30

Even if the API was adequate to support industry-standard plugins (which it isn’t), the GPL license still effectively prohibits most commercial plugins from working in a way you’d expect.

Can’t we just accept that Blender likely is never going to take over “the industry”? Who cares what people in LA think? They’re a tiny industry that struggles as-is. If you’re looking for money, better look elsewhere. Look for people that are easier to please and larger in numbers, such as… hobbyists in emerging, overpopulated third-world-countries.

31

The real vertex count is what I was talking about and it is per subtool.Not per scene.I have also stated that HD sculpting is something different and is out of scope for comparison.

Ofcourse a higher poly count doesnt make a better sculpt but it does mean finer details, but not only that. Depending on your sculpt and your end product/usage it might make a huge difference. If your target is to bake maps then the resolution of your textures will define the final subdivision level.Because if you bake a 4 million sculpt to a 1024 texture you will lose 3 million vertex of information.Because a 1024 texture can hold only 1.2 milion of data without loss.

Also if are making game assets, there are conditions where you have to have high detail but also need to have a single subtool.For example a bear or a wolf. It happened during production so let me tell you; either divide the mesh up to multiple subtools or use photoshop or something similar to bake a map… So please dont expect me to take you seriously when you say something like polycount issue is meaningless. If it was really meaningless and if Pixologic thought the current system was fine they wouldn’t add a “Dynamic” option to “Solo” view mode. You can also find the reason in their update log for Zb4r6…

Take an alpha and plug it into both zbrush and blender, adjust the settings so they are as identical as you can get and then see the difference in quality. Zbrush wins hands down, even if the mesh in blender were twice the poly count it still comes out more accurate and with finer detail in zbrush.

The point is that things like polycount dont make a difference if the end result isnt what they need it to be.

First of all each sculpt tool behavior changes depending on the curve type it uses. Also it changes which kind of projection plane you’re using for the brush and the Brush mapping for the texture mask area plane or view plane…I suggest you experiment…I think you don’t bother with experimenting, or give up to easy…

Another good example is the UV layout. Blender has a great UV toolset, but many, even blender users, will use alternative programs just to get the best results faster and with less work. I would rather use Roadkill UV than Blenders UV editor…and those are based on the same code…yet the interface and controls are different. One is more end user friendly, it focuses on the goal.

I really wish you didn’t say something like this, Roadkill are you serious… I would understand if you would use roadkill instead of Maya UV unwrap but choosing it over Blenders UV tools is just absurd…You really quit to quickly…

You may say that, but I know some fairly big vfx supervisors, those who have worked on major hollywood titles that know of blender and even of Andrew Price. The few who do now of him say good things about his tutorials and opinions… so its kind of funny when these people see actual blender users attack him for stating the obvious.

Well it is ok if you or Andrew or your Big vfx supervisors have opinions but when someone else says their opinion its an attack.Well let me explain yet again (since i already explained it on this thread before). I object andrew’s way of saying these things, not because he is saying it. You can make a nice valid argument and explain it in a non belittling way. So with many people watching blender, old blender user which the community respects so much, shouldn’t create such a negative image of blender all over the internet.Its not like he’s not heard…

The ones with the most experience generally carry more weight than some teenage hobbyist in a third world country.

I am sorry is this directed at me ?

32

Seriusly to all persons that are saying that Blender viewport performance is not that bad. Are you f kidding me? I am modeler for games. and i make HP models. My gun have 3 milions poly. + one subdiv for final stuff. when i turn all layers on. Without that last subdiv it is lagging. Now i export this model with that finall subdiv to modo and with no problem i can rotate and do all stuff. Let’s go more i exported this model to zbrush and subdived it 2 times. And still no lag. So plizz blender performance is pretty bad compared to the standard on the market.

33

Personally this is what I really would like to see improved.
I would like to have great plugins in Blender,paying for them it’s not a problem if there were the quality.
Particles,deformations,shaders,export for external renders,these things should progress even without the agreement of the Blender foundation,but the way development works now is too much dependent on what they want(which often it’s not what advanced users want).
Anyway,I know this will never change and I have allready accepted it(it’s the price to have an open source project so good),but everytime asking for “what want professionals” start to sound a bit naive.

34

Bf decided what kind of license to use,so I don’t think it’s fair to say “this has little to do with BF”.

35

I’d really like Subsurf performance fixed. On the other hand, there’s OpenSubdiv on the horizon. Which will probably not magically make stuff superfast, but it should at least bring Blender’s subsurf in line with Modo. On the other hand, for that to happen they need to solve the manifold, half-edge problem so it might take a nice long while considering Pixar has very little interest in that. And then there’s the fact topology changes (modelling) will be slow because computing their FAR is pretty expensive no matter what you do. So I dunno.

It’d be very nice if we could get a ‘static’ checkbox to make it use VBOs, and if all meshes weren’t double sided by default (who needs that?).

36

No, you didn’t mention that. If you know these “tricks”, then why are you complaining about bad performance instead of complaining about having to work around it? Are you trying to troll me? Performance does not suck if you follow the advice, that’s the point!

The license was chosen over ten years ago. Today, the copyright is owned by many individuals, a lot of which only have licensed their contributions under the GPL. So, even if the BF didn’t favor the GPL (which it does) there isn’t a lot it can do about it today.

37

I never said that BF should change the license now,but the fact that the license can stop(or make harder) some external development I think is something that also you can agree with.

38

Of course I can agree with it, I brought that up in the first place! I miss practically no opportunity to rile on the GPL on these forums.

39

Has Blender really made zero progress to attract professional since 2.4x?

What Zalamander does n’t count as pro, Ace how can you ask such a question?

40

Zalamander

point is that if YOU turn all thos VBO and other stuff it is still shit compared to modo zbrush etc.