Yeah that makes sense. Seen this in any software team I have worked with.
Yes. And this would go I think to all of it. Not just a traditional animation story telling but also designing VFX or CGI animated diagramatic sequences for documentaries or museum displays. Get it nailed down in the very early stages.
Animation either 2D or 3D is a lot of people time and money. Donāt go charging off building assets and systems you will never need. Only build what you are sure you will need and nothing more. Sadly Iāve seen this not happen on several CG productions in the past. Assets built or purchased that were not suitable or were never used. Experimental directions tried that could easily been tried out with clear boarding and planning early on that ate massively into time and budget. And then a mad panic and cobbling stuff togther at the last minute.
Part of that is I think it is often tempting in CGI to want to see stuff done on the computer right from the get go. But if I am looking at the boot up interface of Max Maya or Blender with no plan I canāt make anything. You canāt just magic a production out of thin air.
Well actually you can with boards and animatics. 
3 Likes
There are a few exceptions to this. If you know the artist well enough and what they are capable of, they are able to take a theme and use references to build ideas from scratch with modeling tools.
I donāt do this often. But I did have a project where I knew my top Zbrush artist could hit it out of the park. I gave him the parameters of the character and the story and then let him go to town.
Then I had another artist who I knew if I let him go with the theme would come up with the character perfect for the other side of that story.
It actually work as I expected.
But I have worked with these artists over a number of years with and without concepts and I know more or less it would come out right. I also knew that their imagination would surpass the concept artist.
These were the two main characters I knew would never change.
I also did this with the vehicle for the main character.
For the rest of the project it was concepted and we are following the animatic.
But this allowed me to keep this project moving in the background while we finished up another project, and also finished the boards and animatic.
So in this case it was a smart and efficient use of time and resources overall.
1 Like
So while this thread has drifted somewhat, just going to post to put the drift into perspective.
So the question was along the lines of the current state of the āOpen Movieā as both a community concept and a production reality.
What has plagued the community concept as a production reality leads us to these discussions about production process. And how that failed from the inception of pretty much every effort outside of the Blender Studio.
I donāt see that it has changed much.
I am a firm believer in an objective and honest look at any situation that you want to resolve.
So the first observation is that while not dead as a concept it has yet to actually succeed, coming from the community. There was one outside of the Blender Studio. And it was funded. It was rather private and the people doing it were paid.
There is nothing wrong with that. In fact I will make the point that I think this is still the only working model. Private operation, publicly or privately funded. Tight knit group, well organized, community contributions tightly controlled or non existent. Headed by a paid and experienced producer and a director or directors. Internal creative leads following clear direction and a production schedule kept and project completed.
That is the model in a nutshell.
That does not mean a new model could not emerge.
But while there might be promise of a new working model, it cannot be considered working until completed.
Now part of the reason for starting this thread as you might have guessed is because I have been considering turning one of the projects at my studio into an open project.
Something I have thought about a lot. But I had decided if I would do it, I would follow some form of the only working model I know of. And that would be the Blender Studio.
It would not rely on community contributions but would not fully exclude them either. And this is going back to some kind of version of the pre-BS model.
Anyway enough for now. More if I have an update on that.
3 Likes
Does anybody know what happened to āWires For Empathyā (formerly known as āTubeā)?
It once seemed to be about the most promising non-Blender Studio endeavour towards an open movie by far, but apparently has stalled for soon to be a decade?
It seems they ran out of money and then it fell into hibernation or somethingā¦
greetings, Kologe
2 Likes
I do vaguely remember that one. Nice production value.
It would be very interesting and useful to see more behind the scenes stuff but I donāt just mean the general updates (which while interesting) donāt really provide any āproductionā information.
Iām talking more about issues and problems faced and how they were specifically solved. Getting all the final assets etc will be nice of course, but that doesnāt really tell you how it was all put together and more important why any specific technique was chosen.
Like the example of the models that didnāt cut it to start with, it would be great to see said models, what the problems where and why they would be a problem and how to fix it. Thatās all very useful and in the context of the film very practical information. Something that one doesnāt get from just looking at the final model in Blender.
Or as another example, the use of Kitsu, how you went about setting it up and configuring it, along with practical examples of its actual use in the production. I donāt mean the server setup or installing it, Iāve even covered that in a thread on this forum myself, but a detailed look at itās actual day to day use.
As @tischbein3 said, one of the best parts about working on the project, would be the professioinal feedback and insights into the full process and details of a 3D production. It would just be very nice if it was possible for some of that information to be shared outside of those working on said production.
2 Likes
That is an interesting point. I think this breakdown is also missing at times even within a studio from artist to artist. The solution to this is documentation and training materials. Even if it seems like a no-brainer to you, it might be new to someone else. And also if you are figuring stuff out, you have it all sorted. But then come back later, and chances are you are going to forget half of it.
So myself, I find making tutorials for my team and documentation. And encourage the same in the team.
This is why it is important to have a lead in each area. And part of that job is documentation of best practices, workarounds or certain quality guidelines, so that the work throughout the team is consistent.
Collation, editing and publishing of this documentation to the broad user base of a community is another thing entirely. It takes another level of effort to make that happen.
But I agree, for myself anyway, that this would be invaluable to the community.
3 Likes
Yup, thatās why one of the things in my latest video (https://www.youtube.com/watch?v=--9-HC_zXnA) was Production Notes.
Nothing worse then working something out and then forgetting how to do it 3 months later and having to work it out again.
Of course that doesnāt mean one had to work it out in the first place, which may well mean re-inventing the wheel that someone else has already done but you just donāt know it at the time.
2 Likes
Yeah exactly correct. Our first go-to is tutorials and docs. But then also many times there are a few sources and it helps to document it so you donāt have to sort through a whole tutorial and search around again for the sources.
And then sometimes things are so specific or, tools so finicky you have to sort a few things yourself for your particular case.
And then of course sometimes a doc is simply pointing to existing tutorials.
1 Like
So far we didnāt run into any big issues except the hair sim. We are waiting to get it implemented in the new hair system. We are also having problems with feathers but thatās another thing we put on the side for now. So at this point itās mostly modeling and lookdev so there are no real challenges. I wonāt show the very problematic models because they were very bad and the artists preferred that I help them in private instead of in the general chat on discord. That wouldnāt be cool of me to make a clip to show how bad their work is.
Itās hard to document everything because we are not in a studio environment. We donāt even see each other. This movie is being done by messages!
2 Likes
Yeah in the end we decided to design polygon hair for our current project, and use more stylized characters. We rig it with dynamic bones. And then bake then edit the bone chain as needed.
Some funny stuff going on with the hair texture I gotta get fixed. But the general idea.
5 Likes
Iāve been working with live-action feature films for many years, as a director, screenwriter, teacher and editing commissioner, and the linear waterfall processes have always bothered me(ex. reshoots are just too expensive). Having all elements and all processes inspire each other and, especially, letting the knowledge gained along the way influence and improve on elements which should have been locked-for-production a long time ago. Or in other words, keeping all elements open and changeable, until there is nothing more which can be improved. This was one of my hopes for the Blender, but these days seem tools like Unity much more suitable for such non-linear workflows. The Blender VSE has simply been neglected for too long, and the Blender data structure makes filmmaking with unrendered(to disk) clips in the sequencer just too tedious and unperformant to work with. In spite of this, I havenāt given up hope of using Blender for more open and creative film/story development, which is why I have developed several tools for that process. Ex. screenwriting in the Blender Text Editor, which can turn the script into a timed VSE sequence with dialogue, descriptions and populated med key-word assigned assets: https://github.com/tin2tin/Blender_Screenwriter Last week Iāve added chatGPT, so you can bounce off script ideas with that and even get it to deliver correctly formatted screenplay scenes. Also, Iāve developed an add-on for text strips to speech, so you can get voice on your dialogue. And another add-on where you can edit all the text strips and the order of them, which with very little effort could be returned to screenplay format. With tools like these you should be able to move with ease between writing, drafting, trying out stuff, make mise en scĆØne with assets, do lighting, audio etc. in a non-linear workflow, where everything can be changed at any time. And the development can be let growing more wildly, but also based on what actually works, instead of just assumptions of what might be working in the screenwriting phase. You should, as a filmmaker, find yourself in a process which will feel much more creative by developing the film through exploring if just the tools would allow it.
4 Likes
My approach to it is not really novel necessarily. But certainly simplified to the idea of reducing the necessary stuff you have to push around to get the story.
The relative concept in the real world would be on one extreme, you, your phone and an actor.
And on the other extreme, you, a 35 person crew, 5 production trucks, a big cast, extras and so on.
You have a certain amount of mobility with your large production on the set. Or course, some degrees of freedom.
But if you could figure out the movie on your phone you might eliminate or add locations and scenes quite freely.
So in 3D this process is all about previz and working with proxy models and so on. Starting at the lowest point, small thumbnail sketches to full boards to animatic blocking.
As you move toward the final stages of production, Unreal Engine provides a great artist experience in immediate feedback. But that is another aspect to talk about more fully as well.
Well, that answers a comment I had on a previous video then. I expect we will be waiting till v4.0 at the earliest for any initial hair sim system.
Fair enough about those models, would still be interested in anything else, like Kitsu that I mentioned. Are you using the Blender Studio add-on for it, I tried playing around with it a bit, but from what I could tell it is somewhat hardcoded to the Studio setup/file structure and Iām not a coder that can fix that.
But even that aside, just how you setup users/departments/tasks, etc would be interesting to know.
1 Like
Yes, having a huge team on a live-action shoot seriously limits the wiggle room, though I have been rewriting scenes in the car driving to the shoot, and those scenes made it into the finished 4 mill. $ film.
What Iām thinking about is having tools which enables you to develop the film through the elements which truly inspires you. Iāve been writing screenplays for years, and I must admit I find it very boring and uninspiring. So, developing the film through the elements which really inspires you, will properly result in a more energized and original end result.
Letās say there is a certain visual area which inspires you, and you find or create assets and move them around, adding time to the elements, camera angles, throw it into a timeline and then convert the timeline into a screenplay, while keeping everything rough until you feel you have the timeline worked out, and then you can start improving on all the elements. And the script will mainly serve as inspiration and tool for ex. voice actors.
When the starting phase of developing a film is words, you quickly end up with stories which are more focused on the word and the structure instead of how the film medium can engage emotionally by all other means.
Looking at the Charge previz process, it seems to be very waterfall-ish ex. previz is rendered to frames instead of using scene strips: https://studio.blender.org/blog/charge-previsualization/ and it is mainly used at a video reference for the further work.
If the Blender scene strips and their performance was improved upon, there should be no need to rendering pixels to disk at all in the development phases, and from the vse timeline you should be able to directly edit any element in any scene without any re-rendering to disk, or having to search through folders to locate files, do the change, render to disk, and manually replace that in the strip. Imo, it should all be handled dynamically from the VSE real-time.
However, in the lack of performant scene strips, and workflows of endless scene switching(bc. vse data is embedded in the scene data), we can only do work-arounds like rendering the scene strips to disk, like this add-on can do: https://github.com/tin2tin/add_rendered_strips but the advantage of this is you can still at least access the scene directly and do changes in it.
1 Like
Sounds like it could be a fun alternate way to work.
I like the old school process of a script.
So for now I am ok wit it.
AFAIK there never was any huge community involvement in the open movies beyond community/crowd funding. Ton likely always hoped that with the open content model after release community members might do something to add to the works, or make derivatives perhaps. On that note with Sintel I seem to recall someone doing a game project and I think someone else trying to make some open content comics off of it.
2 Likes
Sintel had quite a program set up for community to build props.
There are probably threads here somewhere.
Yes, I do recall there being something towards that but IIRC there were few assets of sufficient quality and fit in terms of creative direction so as to get included, sadly.
2 Likes