In user preferences, goto 'view and controls'. Change your view rotation to 'turntable' and youll probably want to also turn on 'around selection'. Next check 'select with left mouse button'.
I even installed the Spacenavigator thingie, it streamlined stuff a little, but now I can't make a multiple component selction by draging a selection rectangle with the mouse?
Pior, I had to overdose on videos before I saw how easy it actually was to use Blender. Also, I posted a thread link on the last page if you want to follow my questions as a new Blender user, and read responses I got.
I even installed the Spacenavigator thingie, it streamlined stuff a little, but now I can't make a multiple component selction by draging a selection rectangle with the mouse?
Haha at least it was fun, short ride :P
You press B once to get box select, then click and drag.
Reminds me of the time I installed a trial of Max.
*right click right click*
nothing happens
*left click*
there we go
*drag the box around a little bit*
ok thats fun, where is edit mode?
*look in menus*
*google*
screw it
*uninstall*
After 3 years of using blender I can't get along with the max/etc navigation :poly142:
Anyways what about a little hotkey library?
Hmmm Daaark mentioned a few.
F - Creates a face between 2-4 vertices... edges, edges if there are only 2
L - selects every vertex in the mesh
A - selects everything
k - loopcut menu
Shift + middle mouse: pan view
strg + middle mouse: zoom
Oh ok thanks. But it does not 'stick'... doesn't make sense.
Anyways! Thanks for the help guys.
I don't understand. You press B then you drag with left click (yeah, you use left click there... I know.. I know) :poly124:
If you press B again it uses another selection mode I never use.
Also, you can save whatever default GUI and control options you want by hitting CTRL-D.
jrs, last time I used max (4 or 5) there was no 'edit mode' in that sense. You just free form edited everything in place. Also, when you placed a shape down, like a sphere, it wasn't a mesh. It was an object, and you could tweak the number of sides and stuff it had, and when you were happy with it, you converted it into an editable mesh... (which I used to say as edible, much to amusement of others). Then you just pressed the edit button on the sidebar and got all your edit commands.
You want messed up? You should try truespace or carrara. In Carrara you had to edit things in a special room and couldn't see anything else. You want to model a watch on some guys wrist? You have to do it blindly. (That and NOTHING works)
What I meant is that once I was done dragging the selection rectangle, I was already out of 'B mode' - and had to go back to it to perform a selection again. I am sure there is a way around that, but I feel that it will involve hidden shortcut keys and 'clickarounds'. That is something I don't accept from 3Dprograms anymore.
As a matter of fact I find it interesting that programs can have different philosophies to them (I used to be very fine with trueSpace when I first started, and I am sure that Carrarra can be used just as well).
What I find sad, is that developers sometimes fail to realize that proprietary viewport or selection methods are not going to make their procduct more appealing just because it's "smart".
Once you hit a pipeline with other people and other programs manipulating your data, streamlining and conventions are the key. And then, extra customisation options on top of that for the adventurous user maybe.
As a matter of fact I am almost certain that if Blender suddently adopted standard navigation, selection and mesh manipulation systems (on top of the exact same underlying engine and tools code) you would see it's popularity boost skyhigh!!!
Simple, straight to the point video recordings of talented (but not so technical) artists would appear overnight, like cranking out a great basemesh in blender in a matter of minutes, then sculpt it in Mudbox, maybe even send it back to B for rendering ... all that thanks to a free companion app.
Sadly tho I feel like this won't happen anytime soon.
Rectangle select does have to be reactivated every time you use it. If you press b twice it will bring up brush selection mode, which is persistent (left click to select, middle click to deselect, right click to confirm and exit selection, scroll wheel to change brush size).
Switching to a "standard" system might bring in some new people but would be a disaster for Blender. There would be an overnight revolt, probably ending with a forked code base and development crawling to a halt for a long time while the user and development base was reformed. It is possible (Likely?) that official and user made patches and scripts will add alternate systems, but the current general control scheme will be the default for the foreseeable future.
Edit: Ultimately Blender development truly is driven by its users. We picked up a fair number of XSI refugees recently and their influence is already starting to steer development a bit. After they have been around longer I imagine there will be an even stronger push to develop towards the pieces of their old app that they miss the most. Conceivably, if a large number of people from another app jumped into Blender, development could start to move in dramatically different ways in a relitively short time frame, especially if those changes could be justified as allowing an objectively better workflow.
pior - Is the box selection such a big issue? You'll already have your hand on the b, so all you have to do is b, drag, b, drag, and so on. It's no ideal setup, and I would've preferred to see b enter box mode until you press it again, but as it stands, I can press b again before I'm at the mouse position for the next drag. Alternatively, you may want to try out lasso selection, which is done by holding down Ctrl while dragging LMB (or RMB, if you switched). I find it to be more useful than box selection anyway.
That said, I do think Blender should get at least the option of traditional keys, if only to lure new people in... But I wouldn't want to lose my sweet lasso in favour of your boxes .
Well I know it does not look like a big thing, but when you want to add and remove components to your current selection as fast a possible, you really need the fingers to be close to the modifiers keys (ctrl, alt, shift in major apps). From there it really is a slowdown to have to reach for b just to re-enter a mode that I didn't want to quit in the first place.
Is there similar discussions going on on the Blender user forums? It's weird to see that such things went through unnoticed during all those years of fine tuning.
JRS I don't think it would be a disaster at all, especially if the two systems were designed to coexist (they certainly could be). Worse case scenario is that is that no one give a damn about such an update and the user base remains the same, but still it would give an opportunity for Blender hobbyists to be exposed to industry standards. An overnight revolt would be a sad thing to see - are people that dumb really?
Well I know it does not look like a big thing, but when you want to add and remove components to your current selection as fast a possible, you really need the fingers to be close to the modifiers keys (ctrl, alt, shift in major apps). From there it really is a slowdown to have to reach for b just to re-enter a mode that I didn't want to quit in the first place.
You know that you can just shift and ctrl click things out of your selection right? Also note that until 2.5, Blender doesn't work on multiple objects at once very well.
Big fan of lasso selection myself. Always preferred it over box and it's nice that it is so accessible like that, much like with LW/Modo. Even using 'b' occasionally I haven't found it to be a big deal. Same with the nav, which incidentally can either work very similar to Max (middle mouse button driven) or almost exactly like LW/Modo - alt/ctrl/shift and lmb driven, which I've found to be much better when using a tablet than a maya style nav scheme.
I've really grown attached to blender's right click selection, especially how it also doubles as tweak mode. I put it right up there with Silo's tweak mode in terms of ease of use. Once I established some muscle memory with blender's binds I really started to fly with it. I still consider myself a noob with it, but I'm past the point where I open it up and feel any confusion and that's the key to really coming to grips with any software I've found.
Sometimes quirky turns out to be better than standard. I think that's why I enjoy using apps like zbrush, silo and UVLayout so much. They don't do things the way everyone else is and there are benefits to be gained from that. I get that same feeling with blender now.
I found this to be really helpfull too. Nice visual of the hotkey binds. It's slightly out of date, but there's an updated one that isn't as easy to read also floating around. http://wiki.blender.org/index.php/Reference.br/Hotkey_Map
JRS I don't think it would be a disaster at all, especially if the two systems were designed to coexist (they certainly could be). Worse case scenario is that is that no one give a damn about such an update and the user base remains the same, but still it would give an opportunity for Blender hobbyists to be exposed to industry standards. An overnight revolt would be a sad thing to see - are people that dumb really?
A lot of coexisting stuff has already been added (the option for left click to select for example), more is planned (custom keymaps etc), and more is possible in the future (app specific themes and layouts?). However, Blender has evolved its own workflow over the years. Even if all of the commands were identical to, for example Max, pushing the same buttons in the same order would not produce the same results. Adding seperate controls for converts to use as a crutch while they come to terms with the new work flow is considered a good thing. Changing that workflow to match another app simply because it is the standard would not only be an enormous project, it would fundamentally change the way that Blender works and be the end of the Blender that exists today. This would inspire enough nerd rage to drive plenty of developers and prominant comunity memebers into working on their own seperate fork.
The closest equivilant I can think of might be if Autodesk announced tomorrow that 3DS Max and XSI were being canceled but as compensation everyone would get a coupon for 10% off on Maya. Probably a bad example tho, since in Blender's case people would be able to do something about it.
So in conclusion:
alternate control schemes-possible
alternate workflows-possible in a non comprehensive way
replacing current workflows-end of the world
Some opinion stuff with no official sources, this could all just be my overactive imagination:
Getting a spot in studio pipelines is only a short term goal for Blender. The long term the goal is to end the pipeline altogether. It is already possible to take a concept and push that all the way through from base mesh to sculpted high poly to low poly to unwrapped textured rigged and animated model, to rendered composited footage or mockup of a playable game, all without ever leaving Blender. There are still a few rough spots and a lot of places where Blender falls short of specialized apps in this sort of 'all in one' pipeline, but if the current development pace is maintained and development continues to center around complete projects...in 10 years we may look back on the current standard and shake our heads about how silly it was to push a single asset through a half dozen pieces of software before calling it finished.
But why would different branches be a bad thing? Instead of creating a mammoth app able to do everything under the sun, why not using the solid polygon editing source code and create some kind of simple, opensource modeling package out of it. It would be lightweight, fast and free! Same for highpoly sculpting! Same for map baking! Simple self explanatory interfaces DO have benefits over cryptic menus especially since artists are NOT paid to be ubertechnical, they are paid to be good at doing pretty stuff!
Same for animation, rendering, aso. Take it this way : I don't animate, I draw and model. As far as I am concerned I don't need to worry about getting better at animation (I could have that wish, but I don't have the time and will) ; I just need to know the universal, non app-related constraints of character animation to make my models work smoothly with the rigger and animator. In a similar way, the animator does not need to know modeling - to put it simply, I just provide him what he needs.
So basically all these pipeline tasks could very well be separated ... and as a matter of fact they are nowadays! Max, Maya, XSI, Blender can't compete with Photoshop for texture painting. Trying to make them do that can be nice for sketching textures (Maya paint), but it would be a waste of energy to try and implement a full Photoshop clone made from scratch 'inside' them, coded by the same team who writes the 3D part. Some game making teams model in Maya and animate in Max, and it works just fine.
Same for sculpting. For sure Blender now has highres sculpting. But honestly I would rather see the Blender community work on separate killer app for that! Given the importance of that pipeline element, I am sure it would develop extremely fast and would gather a humongous user base. And then maybe you could throw in the shader engine from the main branch to make nice concept images, and later on also bridge this app together with a little modelling app taking care of editing/creating the base cages back and forth...
Take Xnormal as another example. One could think that baking maps is the same thing no matter what app you use (and many who never did any baking themselves believe so). But IRL Xnormal showed that with a focused effort on a very specific task, one can develop a single app doing one thing or two extremely good, AND create a widely recognized industry standard app! All it needed was OBJ as an accepted standard for exchange. Same for Crazybump.
That's basically my point. I'd rather have a few separated apps doing one or two things very well, rather than monster apps trying to do it all. The mammoth scheme is already starting to fail : Maya can do okay polygon modelling, smooth animation, and mediocre non-linear highpolygon editing. Max can do good enough (yet not great by any means) polygon modeling, is quite slow for animation, but is great for highpoly. These guys tried the monster app thing ... it worked for a bit ... and now it's dying. I think it's a bit far fetched to believe that blender could one day do it all, in a way that every 3D artist would agree with.
Also, focused app bring focused thinking : focusing on one problem lets you come up with creative ideas that wouldn't come to life otherwise if buried under many other things. Who would have thought of the importance of editing a falloff brush curve to sculpt properly, 10 years ago? It might have never happen if Z2 and Mud never focused on that.
But that's just my opinion :P
Sorry for the wall of text, Im sure Rob will enjoy it tho :P
And the separation into different apps does happen to some degree, Roadkill for example. Also, there are currently 40 separate branches of Blender with a dozen of those being actively developed. One branch works to change Blender into a modern NURBS modeler, another to give it a modern render engine, another adding Ngons, another upgrading animation tools, and another rebuilding the core event system. They all work independantly and focused on one specialized area, then get folded back into one program when they are ready.
Then you also have projects like Luxrender or Sharpconstruct that are being (or in Sharpconstructs case was being) developed as entirely separate programs which are then added into Blender.
Edit: The advantages of a single all in one app are really pretty staggering. Not only are there no format conversion issues, but it becomes trivial to make even major changes to an asset after it has already progressed past that step. You also end up with the same tools working on multiple levels. The same node system you use to make textures also controls compositing (and maybe animation and modeling in the future), the paint system that you use for 3d textures is also used for sculpting, the commands that you use for polygon editing are the same for UV mapping, etc.
Well it looks like were both editing back and forth there, so Ill just start a new post.
I do agree that specialized apps will have their place for a long time to come. The world Im envisioning is more like the current 2d workflow. You dont have one app to draw lines, another for curves, a third for color, and a final app to export it to a viewable format. The bulk of even professional level work can happen in just one application, but for some tasks a specialized app is still prefered or even required.
For sculpting...Im very curious where that will go. In my experience Blender will handle around 10% of the polycount you can get out of Zbrush. Some optimization is possible, but the underlying principals of how Blender and Zbrush work (which is too technical for me to understand) make it improbable that Blender will ever be a match in terms of raw power to handle gigantic numbers of polygons.
On the other hand, sculpting is supposed to be a major focus of project Durian after 2.5 is finished. Im not sure where they are planning to go with this but I am sure that they have a pretty good idea (or they wouldnt have anounced that as a major goal for the project). I know that there was already some experimental work done on micropolygon displacement and addaptive subdivision...but I dont really know what they have in mind.
I see blender not as the most powerful polycrunching tool out of the bunch, but definately the most flexible and lightweight of all of them, and I still can't find better uw tools out of the box in any of the other applications.
I wouldn't try to bring in any heavy meshes into blender for baking though
hopefully that's a part of the goals for the next big project.
The Blender UI communicates badly yes, and has enough issues to work further on. However, there's really a lot of innovative and good design concepts behind it, validated and refined in a day-to-day situation with artists. Most people who get past the learning curve then find it intuitive, pleasant, and very fast to use.
Blender uses a lot of concepts similar to one of the world's leading UI designer Jef Raskin (although I didn't know of him until 5 years ago). The old geek has died 3 years ago, but his son Aza has taken over his work with great results. He's now lead UI design at Mozilla and heads up an own startup too. Check on this stuff for example;
I've just found these pages today, and he's fully aligned with our ideas. More over, the ideas we've defined in past week's workshop you can find above as well.
To be clear; it's not that "Blender has been based on Raskin's ideas", but that these concepts are quite obvious from a UI designer perspective, more people have developed similar concepts and use it with a lot of success. Modo is a good example of this; although they claim they never looked at Blender, their core concepts are similar, and were realized in a great way, in many areas better than Blender did.
Many of these concepts go back to the beginning of Blender, 15 years ago. It's still cutting edge, and although people can disagree or prefer to rather follow other conventions, it's definitely not coming from "software developers who are bad in UIs by definition".
Note that I - as the original creator of Blender - have a background in industrial design, and worked for many years as graphics & 3D designer. I've only learned coding to be able to improve and design tools to use ourselves.
Software architecture or other abstract engineering stuff I find typically very difficult to grasp, unless it's usable and benefits creating content!
Even now, I still prefer working with creatives, which goes much easier for me than managing developers. Keeping this focus on our users and the best artists is why we realized Elephants Dream, Big Buck Bunny and YoFrankie. And with more challenging movie projects coming soon!
It should work but I believe that you have to have 32 bit Python installed as well. Also keep in mind that 32 bit RAM limits will probably also be in place.
Thank God it's not April 1st today, or you wouldn't have believed this - the Blender Foundation have decided to release *one* more 2.4 release. If everything goes according to plan, we should see 2.49 at the end of April!
Ton Roosendaal writes:
On popular request (users banging on our doors!) but mostly because current sources have seen much improvements, we challenge fate by releasing the very last available number before 2.50.
Features of 2.49 include (from the top of my mind):
* Texture node editor
* Live Video Texture in Game Engine
* Improved API to talk to logic bricks in GE
* Etch-a-ton, sketching armatures
* Bullet Physics update
* Jpeg 2000 and EDL support in sequencer
* Everything I forgot
* And plenty of bugfixes of course
Most of the branches remain 'branch' for the time being. So no volumetrics, no BMesh, no new Nurbs, no Freestyle, although all these progress well. Dome rendering for the GE we look at to include though.
Within a week we define when the first Release Candidate becomes available, which is probably around April 15th. About 10 days later the real release then could follow. Just wait for further announcements.
Im not sure if this is a late April fools joke or what, but if its true the Blender Game Engine just got a huge boost.
Entertainment Arts Research Using Blender Game Engine for Production
April 6, 2009- Atlanta, Georgia--Entertainment Arts Research Inc. and its subsidiary Twilight 22 Studio both based in Atlanta, Georgia USA are in pre production on an original third person action thriller IP called Fire Wire District 22 driven by artistic creativity and advanced technology.
Fire Wire District 22® is using the Blender Game Engine. EARI is using the Blender Game Engine by adding additional real time shaders, as well as for scène graph and culling. Benoit Bolsee, one of our programmers, has been able to speed up the game engine 5 times in just the very first optimization that was done, stated Nate Nesler. Nesler is the Artistic Technical Director for Fire Wire District 22®.
Twilight 22 has released screen space ambient occlusion, light scattering, improved depth of field, chromatic aberration, and cross shaped bloom to the community for the Blender Game Engine based on the work done by the artist Martinsh Upitis, Mike Pan, and Dalai Felinto for Twilight 22. Future releases to the community will also include subsurface scattering skin shaders, advanced water shaders and effects, level of detail with background loading, and multi-threading for multi-core systems with better over all acceleration of the Blender Game Engine.
Professor Joseph Saulter, CEO and Chairman of Entertainment Arts Research Inc., commented, We look forward to forming strong partnerships with the Blender community. Saulter continued, We are also open to creating alliances with other gaming companies to create open source game development tools.
Development Team:
Joseph Saulter
Johathan Eubanks
John Walker
Nate Nesler
Harsh Borah
Chris Saunders
Dalai Felinto
Benoit Bolsee
Mike Pan
Martinsh Upitis
Campbell Barton
About Blender. Blender is the open source software for 3D modeling, animation, rendering, post-production, interactive creation and playback. Available for all major operating systems under the GNU Public License.
About Entertainment Arts Research Inc. EARI an Atlanta-based video game publisher and interactive entertainment virtual world developer founded in 2000 by Professor Joseph Saulter. Under the leadership of CEO Professor Joseph Saulter, Entertainment Arts Research Inc. is the only African American founded, authentic creator of urban lifestyle gaming entertainment content traded publicly. Board Member Jonathan Eubanks has over 60 video games produced on the open game market, and is the Executive Producer for Fire Wire District 22®. Entertainment Arts Research Inc is traded on the OTC under the ticker EARI.
This summary of Electronic Arts Research Interactive Inc. (Electronic Arts Research or the Company) in its entirety and all attachments hereto are confidential. Neither the document nor the information contained in the document may be reproduced or disclosed to any person under any circumstances without express, written permission of Electronic Arts Research. By reading the enclosed documents you agree to be bound by the conditions of the Electronic Arts Research Non-Disclosure Agreement. Forward Looking Statements: This document contains or may contain forward-looking statements. These statements include, among others, statements of Electronic Arts Researchs plans, objectives, expectations, and intentions, including forward looking revenue and income projections. These statements involve uncertainty and uncertainties. Actual results may differ significantly from those projected or suggested in this document. Assumptions are made about the market, the competitive environment, and other factors that may impact the outcome of these forward looking statements. Other risks include the ability of Electronic Arts Research to obtain adequate financing, and assumptions about the U.S. Federal, The Uninet Kingdom FSA regulations and other International regulatory processes. Electronic Arts Research does not undertake any obligation to release publicly, any revisions to these forward looking statements to reflect events or circumstances occurring after the date hereof or to reflect the occurrence of unanticipated events.
Im really not sure. My first guess would be that they were once known as Electronic Arts Research Interactive until some EA lawyers got ahold of them and made them change the name. That last part looks like standard corporate fluff and it sounds plausible to me that nobody remembered to update it.
Ive been digging a little bit more. The company is apparently real and some of these promised upgrades are already being implemented. Im not sure how much will make it into 2.49 (major features are already locked out and a full feature freeze is only days away).
Edit: for some reason I hadnt really looked through that list of developers very closely. Theres a couple of big name Blender devs on the list that have been making major changes to the BGE code for the past couple weeks, so Im going to guess that most of this will be in 2.49.
Edit 2: Some official clarification on the Electronic Arts thing.
To clear up some confusion we are not part of EA, Electronic Arts. Our umbrella company name use to be EAR, Electronic Arts Reseach, but it was changed to Entertainment Arts Research because of the EA made a new division named EAR so it was too close to our company name back when. Our PR copied the release for the bottom from an old press release and we did not catch it when I posted it. I have since corrected it. So to be clear we have no connection with EA, Electronic Arts.
Also we are planning to go to console but right now we are planning for PC release first. Not sure if the Wii will be one of the target consoles. That was an old piece of information someone found somewhere. I have no idea how old it that other information could be.
Cheers,
Nate Nesler
Artistic Technical Director
Twilight 22
Man that is going kick some serious ass for the BGE. Nice to see that it will be getting just as much love as the rest of the software in the near future - and from commercial developers no less!
I think that hybrid commercial/enthusiast/full project development is what the Blender Foundation is going for, so with any luck well see more projects like this in the future. I certainly wasnt expecting the game engine to be the beach head* for this as it is generally considered a peripheral part of Blender.
I would guess that the Peach project had something to do with it. The Crystal space devs spent months screwing around with their engine trying to get a game together and finally the Blender devs just did it themselves in BGE...in a single weekend. Apparently somebody noticed that and saw an opportunity.
*not to say that this is the first commercial development of Blender since it went opensource. The current fur and particle system is largely thanks to commercial development, as well as the upcoming volumetics and GI systems.
Hmmmmm...custom UI panels through simple python scripting. It's not exactly a "fix" for Blender's traditional interface. (since it doesn't change the fundamental structure of how the program is organized" But it does significantly open Blender up for extensive customized UI modifacation.
The real benefit here is that it will be easier for developers who use Blender to streamline the program for their particular content pipelines. Users will have the option of consolidating their individual workflow into a single panel. Certain functions that can be accessed through Python can be automated into a basic UI panel. This will be especially good for the Blender Game engine. It will make it possible to create custom UI "tools" that will automate a lot of Blender's Game engine options. Imagine having a simple panel that will generate and costomized a tile-based room or level for a pre-existing python game script.
Hmmmmm...custom UI panels through simple python scripting. It's not exactly a "fix" for Blender's traditional interface. (since it doesn't change the fundamental structure of how the program is organized" But it does significantly open Blender up for extensive customized UI modifacation.
No, the new UI rewrite changes the traditional interface.
No, the new UI rewrite changes the traditional interface.
(scratches head) But i'm a long-time Blender user. That isn't necessarily a good thing for me. I'm already familiar with Blender's various UI conventions. But then, I suppose the traditional Blender workflow will be similar enough that it won't pose any problem for me. After all, if I could get used to the old Blender interface, I can probably familiarize myself with ANY UI.
(scratches head) But i'm a long-time Blender user. That isn't necessarily a good thing for me. I'm already familiar with Blender's various UI conventions. But then, I suppose the traditional Blender workflow will be similar enough that it won't pose any problem for me. After all, if I could get used to the old Blender interface, I can probably familiarize myself with ANY UI.
Click the green text that says FIRST at the bottom of the page.
(goes off and reads the rest of the posts in the thread)
Aaaand, my mind is blown. I am especially excited by the re-tooling of Blender's animation system. I've always like the way that Blender "blends" different animation methodologies. But I've always found the way it stores those animations to be a bit tedious and unintuitive. Stroring animations entirely in the Actions editor is a good idea, as well as tying the IPO editor to a scene manager. (so that you are always clear on what curves you are editing, and what they are attached to)
The UI revisions make it sound like everything I had proposed would be possible. Very exciting.
I don't care about using python to change the UI as it is/will be. I quite like blender's UI as it is and I'm sure it will only get better with 2.5. I'm more interested in the custom rigging/control panels I'll be able to build that will pretty much duplicate all the rigging tools I've built for myself in XSI. I very much want to use blender as my main content/pipeline tool so all this stuff, from the new UI to the new animation tools and better, more robust python integration are most welcome changes.
yup same for me - though not for animation but rather modeling/uv or game export stuff. It would be very nice to easily add new panels to the existing application to extend the features in a way you need it.
GSoC grants have been announced. Hope that they work out well.
This year's Google Summer of Code projects has just been announced. Blender got 6 proposals accepted!
Blender Light Paint using Spherical Harmonics - Jingyuan Huang (mentor: Martin Poirier)
Inspired by Illumination Brush, this project aims to create an interactive tool for Blender to preview, modify, and create image-based lighting environments specified by HDR light probe images. Artists can change the HDR light probe by directly painting colours on the models. This tool will be integrated with the existing image-based rendering routines in Blender in order to help artists determine the final renderings.
Collada Support Improvements For Blender - Chingiz Dyussenov (mentor: David? Letwory)
COLLADA is a flexible and versatile digital assests exchange format
created by the Khronos Group consortium. It is supported by the top
commercial DCC tools like Maya, 3ds Max, Lightwave 3D, Cinema 4D and
Unreal Engine. This project is about improving COLLADA import/export support in Blender.
Implementation and Validation of Python Import/Export API - Arystanbek Dyussenov (mentor: Campbell Barton)
Blender 2.5 Project has been in active development past 6 months.
Many improvements have been made and it is a very big step towards
robust architecture. In the light of these great changes I propose a
small help to the Blender team. This project will help implement
Python import/export API.
Integration of OpenGL VBOs and vertex arrays for faster 3D Viewport rendering - Lukas Steiblys (mentor: Joseph Eagar)
One of the biggest weaknesses of Blender is the inability to render high polygon counts while modeling. I believe the way blender uses OpenGL API is the biggest cause: it does immediate mode rendering. Using VBOs and vertex arrays, sometimes called the retained mode, would speed up the rendering substantially and provide a user with a more responsive interface.
Raytrace Optimization - Andre Susano Pinto (mentor: Brecht Van Lommel)
This proposal aims at optimizing blender raytrace.
Including:
* Making it faster
* Support for render instancing
* Added support for generic objects (used to implement instancing but later can be expanded into LOD)
* Script to measure performance
* Removing worst-cases
* Make it easier to try tweaks and different data structures.
Refactor of Non-Linear Animation (NLA) System in Blender for 2.5 - Joshua Leung (mentor: Matt Ebb)
This project aims to refactor the Non-Linear Animation (NLA) system in Blender to function in conjunction with the new "Animato" animation system core in Blender 2.5. A more extensible framework for evaluating animation clips to layer and mix them will be developed, allowing for the addition of more powerful features at a later stage. This will be presented by a more modern and usable UI for the creation and editing of such data, which clearly exposes the true nature and power of NLA workflows.
google does a great job with the Summer of Code, other big companies should take an example on it.
It is also great to see that blender get's so actively supported especially code wise in the community.
Roughly the Blender 2.5 project has the following phases:
Redesign of internal window manager and event/tool/data handling system.
Porting over old code to clean-up and match the new design.
Bringing tools and editors back using new event system and 'operators'.
Redesign and implement UI context, workflow and visual representation.
Design and implement the new Python API.
The first two phases are nearly complete, the third phase is about 70% done. The March-July period will be mostly devoted to steps 4 and 5.
More information can also be found in our wiki: wiki.blender.org/index.php/BlenderDev/Blender2.5
Other current projects
Currently work is being done on many areas in Blender;
Volumetric rendering
New mesh system, supporting ngons
NPR rendering, using FreeStyle
New dynamics based IK controllers
Sculpting speedup, support for animated multiresolution meshes
It's getting better version after version, but for 2.50 we have to wait a lot!
jrs100000, do you know what is Projection texture painting?
I know Blender is able to paint on mesh and on UV, but what's this new feature for?
Thanks.
I think (and this is just a guess based on what every other app calls it) projection texture painting works like mudbox or zbrush where you can take a texture and project/paint it onto the mesh. So lets say you have a few lizard scale tiles (soft under belly, mid range and rough back) and you want to cover parts of your mesh, you line them up and brush paint each into certain sections. Or a skin imperfections sheet, or rough cracked or burned skin tiles to paint in some different kinds of areas.
Replies
I even installed the Spacenavigator thingie, it streamlined stuff a little, but now I can't make a multiple component selction by draging a selection rectangle with the mouse?
Haha at least it was fun, short ride :P
Check this series out.
http://www.blendernewbies.com/tutorials/swf/pretzels/part1/pretzels_part1.html
http://www.blendernewbies.com/tutorials/swf/pretzels/part2/pretzels_part2.html
http://www.blendernewbies.com/tutorials/swf/pretzels/part3/pretzels_part3.html
And yeah, that hidden options menu is a mindfuck at first.
Anyways! Thanks for the help guys.
*right click right click*
nothing happens
*left click*
there we go
*drag the box around a little bit*
ok thats fun, where is edit mode?
*look in menus*
*google*
screw it
*uninstall*
After 3 years of using blender I can't get along with the max/etc navigation :poly142:
Anyways what about a little hotkey library?
Hmmm Daaark mentioned a few.
F - Creates a face between 2-4 vertices... edges, edges if there are only 2
L - selects every vertex in the mesh
A - selects everything
k - loopcut menu
Shift + middle mouse: pan view
strg + middle mouse: zoom
btt:
I am really lookin forward to that release
If you press B again it uses another selection mode I never use.
Also, you can save whatever default GUI and control options you want by hitting CTRL-D.
jrs, last time I used max (4 or 5) there was no 'edit mode' in that sense. You just free form edited everything in place. Also, when you placed a shape down, like a sphere, it wasn't a mesh. It was an object, and you could tweak the number of sides and stuff it had, and when you were happy with it, you converted it into an editable mesh... (which I used to say as edible, much to amusement of others). Then you just pressed the edit button on the sidebar and got all your edit commands.
You want messed up? You should try truespace or carrara. In Carrara you had to edit things in a special room and couldn't see anything else. You want to model a watch on some guys wrist? You have to do it blindly. (That and NOTHING works)
As a matter of fact I find it interesting that programs can have different philosophies to them (I used to be very fine with trueSpace when I first started, and I am sure that Carrarra can be used just as well).
What I find sad, is that developers sometimes fail to realize that proprietary viewport or selection methods are not going to make their procduct more appealing just because it's "smart".
Once you hit a pipeline with other people and other programs manipulating your data, streamlining and conventions are the key. And then, extra customisation options on top of that for the adventurous user maybe.
As a matter of fact I am almost certain that if Blender suddently adopted standard navigation, selection and mesh manipulation systems (on top of the exact same underlying engine and tools code) you would see it's popularity boost skyhigh!!!
Simple, straight to the point video recordings of talented (but not so technical) artists would appear overnight, like cranking out a great basemesh in blender in a matter of minutes, then sculpt it in Mudbox, maybe even send it back to B for rendering ... all that thanks to a free companion app.
Sadly tho I feel like this won't happen anytime soon.
Switching to a "standard" system might bring in some new people but would be a disaster for Blender. There would be an overnight revolt, probably ending with a forked code base and development crawling to a halt for a long time while the user and development base was reformed. It is possible (Likely?) that official and user made patches and scripts will add alternate systems, but the current general control scheme will be the default for the foreseeable future.
Edit: Ultimately Blender development truly is driven by its users. We picked up a fair number of XSI refugees recently and their influence is already starting to steer development a bit. After they have been around longer I imagine there will be an even stronger push to develop towards the pieces of their old app that they miss the most. Conceivably, if a large number of people from another app jumped into Blender, development could start to move in dramatically different ways in a relitively short time frame, especially if those changes could be justified as allowing an objectively better workflow.
That said, I do think Blender should get at least the option of traditional keys, if only to lure new people in... But I wouldn't want to lose my sweet lasso in favour of your boxes
Is there similar discussions going on on the Blender user forums? It's weird to see that such things went through unnoticed during all those years of fine tuning.
JRS I don't think it would be a disaster at all, especially if the two systems were designed to coexist (they certainly could be). Worse case scenario is that is that no one give a damn about such an update and the user base remains the same, but still it would give an opportunity for Blender hobbyists to be exposed to industry standards. An overnight revolt would be a sad thing to see - are people that dumb really?
I've really grown attached to blender's right click selection, especially how it also doubles as tweak mode. I put it right up there with Silo's tweak mode in terms of ease of use. Once I established some muscle memory with blender's binds I really started to fly with it. I still consider myself a noob with it, but I'm past the point where I open it up and feel any confusion and that's the key to really coming to grips with any software I've found.
Sometimes quirky turns out to be better than standard. I think that's why I enjoy using apps like zbrush, silo and UVLayout so much. They don't do things the way everyone else is and there are benefits to be gained from that. I get that same feeling with blender now.
http://wiki.blender.org/index.php/Reference.br/Hotkey_Map
A lot of coexisting stuff has already been added (the option for left click to select for example), more is planned (custom keymaps etc), and more is possible in the future (app specific themes and layouts?). However, Blender has evolved its own workflow over the years. Even if all of the commands were identical to, for example Max, pushing the same buttons in the same order would not produce the same results. Adding seperate controls for converts to use as a crutch while they come to terms with the new work flow is considered a good thing. Changing that workflow to match another app simply because it is the standard would not only be an enormous project, it would fundamentally change the way that Blender works and be the end of the Blender that exists today. This would inspire enough nerd rage to drive plenty of developers and prominant comunity memebers into working on their own seperate fork.
The closest equivilant I can think of might be if Autodesk announced tomorrow that 3DS Max and XSI were being canceled but as compensation everyone would get a coupon for 10% off on Maya. Probably a bad example tho, since in Blender's case people would be able to do something about it.
So in conclusion:
alternate control schemes-possible
alternate workflows-possible in a non comprehensive way
replacing current workflows-end of the world
Some opinion stuff with no official sources, this could all just be my overactive imagination:
Getting a spot in studio pipelines is only a short term goal for Blender. The long term the goal is to end the pipeline altogether. It is already possible to take a concept and push that all the way through from base mesh to sculpted high poly to low poly to unwrapped textured rigged and animated model, to rendered composited footage or mockup of a playable game, all without ever leaving Blender. There are still a few rough spots and a lot of places where Blender falls short of specialized apps in this sort of 'all in one' pipeline, but if the current development pace is maintained and development continues to center around complete projects...in 10 years we may look back on the current standard and shake our heads about how silly it was to push a single asset through a half dozen pieces of software before calling it finished.
But why would different branches be a bad thing? Instead of creating a mammoth app able to do everything under the sun, why not using the solid polygon editing source code and create some kind of simple, opensource modeling package out of it. It would be lightweight, fast and free! Same for highpoly sculpting! Same for map baking! Simple self explanatory interfaces DO have benefits over cryptic menus especially since artists are NOT paid to be ubertechnical, they are paid to be good at doing pretty stuff!
Same for animation, rendering, aso. Take it this way : I don't animate, I draw and model. As far as I am concerned I don't need to worry about getting better at animation (I could have that wish, but I don't have the time and will) ; I just need to know the universal, non app-related constraints of character animation to make my models work smoothly with the rigger and animator. In a similar way, the animator does not need to know modeling - to put it simply, I just provide him what he needs.
So basically all these pipeline tasks could very well be separated ... and as a matter of fact they are nowadays! Max, Maya, XSI, Blender can't compete with Photoshop for texture painting. Trying to make them do that can be nice for sketching textures (Maya paint), but it would be a waste of energy to try and implement a full Photoshop clone made from scratch 'inside' them, coded by the same team who writes the 3D part. Some game making teams model in Maya and animate in Max, and it works just fine.
Same for sculpting. For sure Blender now has highres sculpting. But honestly I would rather see the Blender community work on separate killer app for that! Given the importance of that pipeline element, I am sure it would develop extremely fast and would gather a humongous user base. And then maybe you could throw in the shader engine from the main branch to make nice concept images, and later on also bridge this app together with a little modelling app taking care of editing/creating the base cages back and forth...
Take Xnormal as another example. One could think that baking maps is the same thing no matter what app you use (and many who never did any baking themselves believe so). But IRL Xnormal showed that with a focused effort on a very specific task, one can develop a single app doing one thing or two extremely good, AND create a widely recognized industry standard app! All it needed was OBJ as an accepted standard for exchange. Same for Crazybump.
That's basically my point. I'd rather have a few separated apps doing one or two things very well, rather than monster apps trying to do it all. The mammoth scheme is already starting to fail : Maya can do okay polygon modelling, smooth animation, and mediocre non-linear highpolygon editing. Max can do good enough (yet not great by any means) polygon modeling, is quite slow for animation, but is great for highpoly. These guys tried the monster app thing ... it worked for a bit ... and now it's dying. I think it's a bit far fetched to believe that blender could one day do it all, in a way that every 3D artist would agree with.
Also, focused app bring focused thinking : focusing on one problem lets you come up with creative ideas that wouldn't come to life otherwise if buried under many other things. Who would have thought of the importance of editing a falloff brush curve to sculpt properly, 10 years ago? It might have never happen if Z2 and Mud never focused on that.
But that's just my opinion :P
Sorry for the wall of text, Im sure Rob will enjoy it tho :P
If anyone can do it without bloat Blender can.
And the separation into different apps does happen to some degree, Roadkill for example. Also, there are currently 40 separate branches of Blender with a dozen of those being actively developed. One branch works to change Blender into a modern NURBS modeler, another to give it a modern render engine, another adding Ngons, another upgrading animation tools, and another rebuilding the core event system. They all work independantly and focused on one specialized area, then get folded back into one program when they are ready.
Then you also have projects like Luxrender or Sharpconstruct that are being (or in Sharpconstructs case was being) developed as entirely separate programs which are then added into Blender.
Edit: The advantages of a single all in one app are really pretty staggering. Not only are there no format conversion issues, but it becomes trivial to make even major changes to an asset after it has already progressed past that step. You also end up with the same tools working on multiple levels. The same node system you use to make textures also controls compositing (and maybe animation and modeling in the future), the paint system that you use for 3d textures is also used for sculpting, the commands that you use for polygon editing are the same for UV mapping, etc.
I do agree that specialized apps will have their place for a long time to come. The world Im envisioning is more like the current 2d workflow. You dont have one app to draw lines, another for curves, a third for color, and a final app to export it to a viewable format. The bulk of even professional level work can happen in just one application, but for some tasks a specialized app is still prefered or even required.
For sculpting...Im very curious where that will go. In my experience Blender will handle around 10% of the polycount you can get out of Zbrush. Some optimization is possible, but the underlying principals of how Blender and Zbrush work (which is too technical for me to understand) make it improbable that Blender will ever be a match in terms of raw power to handle gigantic numbers of polygons.
On the other hand, sculpting is supposed to be a major focus of project Durian after 2.5 is finished. Im not sure where they are planning to go with this but I am sure that they have a pretty good idea (or they wouldnt have anounced that as a major goal for the project). I know that there was already some experimental work done on micropolygon displacement and addaptive subdivision...but I dont really know what they have in mind.
(Was not meant to be, the damn thing didn't save...)
I wouldn't try to bring in any heavy meshes into blender for baking though
hopefully that's a part of the goals for the next big project.
Ive been digging a little bit more. The company is apparently real and some of these promised upgrades are already being implemented. Im not sure how much will make it into 2.49 (major features are already locked out and a full feature freeze is only days away).
Here is some of the stuff they have released already
http://blenderartists.org/forum/showthread.php?t=152343
Edit: for some reason I hadnt really looked through that list of developers very closely. Theres a couple of big name Blender devs on the list that have been making major changes to the BGE code for the past couple weeks, so Im going to guess that most of this will be in 2.49.
Edit 2: Some official clarification on the Electronic Arts thing.
I would guess that the Peach project had something to do with it. The Crystal space devs spent months screwing around with their engine trying to get a game together and finally the Blender devs just did it themselves in BGE...in a single weekend. Apparently somebody noticed that and saw an opportunity.
*not to say that this is the first commercial development of Blender since it went opensource. The current fur and particle system is largely thanks to commercial development, as well as the upcoming volumetics and GI systems.
http://blenderartists.org/forum/showpost.php?p=1356291&postcount=1304
The real benefit here is that it will be easier for developers who use Blender to streamline the program for their particular content pipelines. Users will have the option of consolidating their individual workflow into a single panel. Certain functions that can be accessed through Python can be automated into a basic UI panel. This will be especially good for the Blender Game engine. It will make it possible to create custom UI "tools" that will automate a lot of Blender's Game engine options. Imagine having a simple panel that will generate and costomized a tile-based room or level for a pre-existing python game script.
(scratches head) But i'm a long-time Blender user. That isn't necessarily a good thing for me. I'm already familiar with Blender's various UI conventions. But then, I suppose the traditional Blender workflow will be similar enough that it won't pose any problem for me. After all, if I could get used to the old Blender interface, I can probably familiarize myself with ANY UI.
(goes off and reads the rest of the posts in the thread)
Aaaand, my mind is blown. I am especially excited by the re-tooling of Blender's animation system. I've always like the way that Blender "blends" different animation methodologies. But I've always found the way it stores those animations to be a bit tedious and unintuitive. Stroring animations entirely in the Actions editor is a good idea, as well as tying the IPO editor to a scene manager. (so that you are always clear on what curves you are editing, and what they are attached to)
The UI revisions make it sound like everything I had proposed would be possible. Very exciting.
http://www.blendernation.com/2009/04/17/interview-with-ton-roosendaal/
It is also great to see that blender get's so actively supported especially code wise in the community.
http://www.blender.org/development/current-projects/changes-since-248/
Partial list of new features:
Also updates on the progress of 2.50
If I can get this with configurable mouse/keyboard shortcuts, I'll switch to blender in a heartbeat.
jrs100000, do you know what is Projection texture painting?
I know Blender is able to paint on mesh and on UV, but what's this new feature for?
Thanks.