You're blocking ads, which pay for BlenderNation. Read about other ways to support us.

Tears of Steel


The Blender Institute presents their latest Open Movie: Tears of Steel.

About the Author

Avatar image for Bart Veldhuizen
Bart Veldhuizen

I have a LONG history with Blender - I wrote some of the earliest Blender tutorials, worked for Not a Number and helped run the crowdfunding campaign that open sourced Blender (the first one on the internet!). I founded BlenderNation in 2006 and have been editing it every single day since then ;-) I also run the Blender Artists forum and I'm Head of Community at Sketchfab.


  1. a great short, but i wish the story had come to completion, i feel the ending was the only bug that the team couldn´t solve properly in favor of the audience experience

  2. AMAZING!!!!!

    Words cannot do here, the level of everything was so polished and professional with a great style and theme throughout.

    Congrats to everyone who had a part in this! My mind just went into watching it the same as a hollywood action sci-fi film like the avengers, and on that level of expectation it not only matched it, the cool points made it surpass hollywoods efforts =D

    Off to watch it a third time, Blender is about to get a huge boost indeed!

  3. Amazing Job! Not shy of the VFX you'd expect in Hollywood cinemas! With more man-hours spent on a project, I'm sure Blender could accomplish many of the effect we see on the big screens today. Great work, and great demonstration. This project is a distinct milestone for Blender and Open Source. Congratulations to everyone involved.

  4. Well, it's obvious now that the only thing that fell short on this project was the trailer :).

    Loved the film; a good concept, and as a demonstration of Blender's capabilities it is an instant WOWer. I don't think anybody will argue that in terms of writing and the traditional side of the craft a steadier hand could have been a nice thing, but I found it very satisfying overall. My only real beef would be that, given the pretty original art direction in the film itself, the IronMan-knockoff-ish credits were a bit of a let down.
    Still, congratulations and mission accomplished!

  5. Some of it was not great, but other parts definitely were. In general the mix of real and CG was well done, and that was one of the project's main goals. Maybe my favorite thing was the effects - either that or the little chainsaw! The plot and dialog were pretty meh to me. The robobeast things looked silly, were not animated very well, they overreacted to small-arms fire and climbed up things that shouldn't support their weight. I was hoping the crashing ship would demolish some buildings. Still the movie is pretty cool and fun (like it says on the box), and I'll watch it again.

  6. Great SFX, weak movie. I think it's poorly written, in places poorly acted and certainly comes across as very amateur. The visual design is good, some of the animation is superb, but this is like a lot of fragments from a film thrown together to tell a story unlike the one the fragments are taken from. I have no idea if this short will help Blender or not, I guess if people concentrate on the SFX aspects then it will, but as a movie it has little to offer.

  7. Effects were great, especially the HUD stuff, the robot arm, the arm guns and the fake Amsterdam. Plot was lacking somewhat, but the credits music was monstrous.

    Congratulations to all involved. You all deserve a hearty meal, several biertje’s and about 6 weeks sleep.

  8. This is going to leave a lot of non Blender Heads asking for more.
    If there were a big banner in the opening credits saying 'made using completely free software' flashing on and off, they'd probably think it was part of the film, not the credits!
    Congratulations on another great project, I'm off to watch it again.

  9. I thought it was good, and you can see there is so much work put into it. The scenes are really full, and believable in terms of everything being placed nicely and in natural poses. The animation is really good. The integration of CGI with video, though, I expected better. I notice the difference between what is filmed and what is CGI, in most shots, and that isn't very good. And lots of lights are very obviously studio lights, and look artificial. That is a sign that there is probably something missing in cycles or the compositor to make it perfect. Or maybe there wasn't enough time to spend hours making it perfect and/or reshooting scenes that could have looked better.
    About the story, it's not brilliant, but it's ok.

    Thank you Blender Foundation!

  10. Great job! I loved it, I laughed, I scratched my head and I watched it again. hehe
    Really nice animations, the color was stellar, wonderful tracking, etc ... Congratulations team and thank you to all the supporters.

  11. I have been totally mind-blown! I felt like my entire being had been dissipated. I was floating in thin air and was lost in an unknown abyss.

    And then I come back to my senses and realize I had just watched 'Tears of Steel', the Blender Foundation's latest Open Movie.

    This probably sums up my feelings towards the film. I'm still unable to digest what I have just seen. It's just too !@#$ awesome!

    Blender's history was etched yet again, this time on a harder stone!

    To the Mango team, to the Blender developers, and to the rest of the Blender community: congratulations!

    I salute you all!



  12. HI Blender team!
    WOW, that project was amazing!

    I am glad Ian Hubert pushed a not completely sad ending, usually preferred by Tom. I am still traumatized with Sintel!

    I am also salute Ton Roosendaal who reached the limit in his killing spree! after killing the protagonist, he killed almost the whole mankind in this movie.

    A whole new level. Nevertheless I still don't understand who you can create the greatest open source in the world, and create those movies that are anything but depressing!

    Please Tom, create something motivational!, Please, please!, You have motivated thousands of people around the world, not only because of the software, but all your contributions. You are really working for a new better world.

    Thank you all!
    Wish you all success! Congratulations,

  13. Ton and the Blender team show again what an incredible piece of software they have created, and the amazing results that are possible. I would score ToS as:

    VFX: 10/10 Superb visuals that perfectly blend reality with fantasy. An excellent advert for Blender and its dedicated development team!
    Sound / music: 9/10 Entirely suited to the film. Nicely noticeable, but never intrusive.
    Acting: 8/10 Some very distinguished looking actors for a short film.
    Script: 6/10 A story as simple as this shouldn't have needed any dialogue.
    Storyline / Direction: 3/10 Overly disjointed and confusing storyline. Tries to be too clever, but in doing so fails to guide the viewer smoothly through the story.

  14. The end title is really amazing. ( Maybe even better than the movie, oops ).
    The descending guy shooting the bars puts the movie in its right perspective, very funny.

    Very cool project, love that it was open and shared throughout the proces.

  15. I had to register today just because I was so compelled to ad my two cents.

    This is a great accomplishment for Blender, worthy of the praise they're receiving! Congratulations to all of the creators!

    I actually had the opposite reaction that many here seem to have had. I really loved the writing and acting, and felt the VFX was where the movie fell short.

    My immediate impression was that the entire movie was meant to exude a certain self-awareness and ironic detachment. The lines "This is pretty freaky" and "I just want to be awesome in space" clearly set the tone. Those calling it "cheezy" dialogue were perhaps expecting a more serious bit of sci-fi. I got the impression that we were supposed to have a little more fun with it.

    As for the VFX. Look, there's not doubt that there is VERY strong work going on here. Certainly the motion tracking, masking, animation, and node-based compositing tools shine brightly. It is OUTSTANDING that such robust tools exist in a completely free platform. I think we have to be honest, however and recognize that the team did not quite reach the goal of being photorealistic. Was cycles not up to the task yet? Could they have "gotten there" with a few more months' time? Who knows. But those comparing the effects to big-budget features are, I fear, being extremely hyperbolic.

    None of those critiques take away from what is a great accomplishment, however. Photorealism is tough. Cycles is in its infancy. I wonder if this movie were to be re-rendered a year from now if the (forgive the pun) blending of the live-action and CG elements would be considerably harder to spot. The great thing is, being open source, we will almost certainly see artists out there who will "take a crack" at improving the photorealism. The results should be amazing.

    Thank you, Blenderfoundation!

  16. Well, I have to be honest here, even though I know most of you won't agree.

    As a film...
    The acting was "meh."
    The story was "meh."
    The action was "meh."
    The art direction was "meh."
    And though I applaud the FX overall, even in some places even the CGI FX were "meh."

    But the film proves the main point of its production: In the right hands, Blender is a serious tool capable of some amazing things. This was still a powerful statement, considering that this indie film was made by a small team of people who generally aren't even in the professional film industry, made in just a few months, with limited a budget, and still developing their main tool as they make the film. For that effort, I do applaud the film.

    As for Blender itself as a tool for the professional industry, though it's not yet on-par with the likes of the monstrously-robust Nuke, and I know most studios will still adopt Blender as their tool of choice (at least, not at this point, anyways), it's definitely got some serious potential, in the right hands. I think those animation professionals would agree.

    I look forward to seeing more and more growth with Blender. It's already proving amply awesome for the smaller indie level, but I hope to perhaps see this great tool reach its way prominently onto the big screen (and perhaps big video games as well) someday. This film was a good stepping stone towards that potential.

    My two cents.

    • Correction: "I know most studios will not still adopt Blender as their tool of choice..."

      And I also meant to add "The script was "meh" to that list, but I think you got the point.

    • I think you are right but I think that they could have done a lot better!
      There are many scenes that are irrelevant for the story.
      And this just by not making it again so long.
      A simple short story good told is totally enough.

      Lets say they had done the starcraft 2 teaser with an actor or something similiar, just keeping it short and having more time to tweak the details. Awesome!

      Another idea could be making a comic book telling a story and like in starcraft doing some animated cuts (1-2 minutes max) throughout the story. There would be the possibility to work width different styles and showing certain new features.

  17. It is an awesome proof of concept and I'm positive that it will turn more heads towards Blender.

    However, I felt that the acting, the script and the robots could have been done better.

  18. This was, as all Blender Foundation productions, so much greater than I expected. Great Job! Blender and Blender Foundation is the greatest!

  19. I'm amazed of this film as a best Blender's DEMO!
    It's not so bad as somebody write here. Isn't ideal, of course. The level of Hollywood directors, great Asian films' directors and others whose achieved a success isn't that easy to grow just in a few films.
    Tears of Steel is VERY ambitious but it's good for the demo sake. If you'll create 80% of great actor's play with smart speeches (by time) like it's suitable for full-length films then here will be not enough time for Blender to show it!
    So actually it would be great (if possible) to remake this movie as a full-length. Hire more actors. More experienced actors. Render those really beautiful scenes with city landscapes in x3 longer (it's very short, I must replay few times to get a real pleasure). Add more different music. Some "ambient" in "no-action" scenes and something different in all action scenes.
    Maybe try to use some crowd simulations to show greater (in comparison with "big" films) moments in this scenario.
    I'm not blaming, really! Just wish to these people (the director, modellers, matte-artist, animators... all crew) to create more "filmish" film :)
    It's a big step! Awesome FX (the most of them), awesome scenes where (in 50% cases) I can't understand what's real and what's not and a pretty stylish end titles :)

  20. Very nice, and a great showcase of what these tools can do in expert hands. I can't wait to pick apart the scene files and see how some of this was done.

  21. Well...technically maybe the best blender movie of the 4 movies, but i really don't get the story. I think this is more a tech-demo than a movie. :(
    Maybe there is alot of hard work behind this and i really respect that. Great Work!
    The VFX is amazing, but vfx is not everything..
    I know that there can't be an oscarlike-acting, but what iam missing in the movie is some kind of heart and soul.
    And a prober Story.
    I am a bit dissapointed, because the story is to complex to be told in 10 minutes.

    Who is the Girl? What are they doing? What is the purpose of the maschines?

    So a enjoyed the visuals, but nothing more.

  22. Some very good and some not so good. I was quite surprised that tracking seemed to be lost at the very end of the scene where her robot hand grabs his face and when the old man grabs the sides of the big robot.

  23. Best Blender open project thus far, by a wide margin. Full Disclosure: I only watched it in low-def, but the compositing and animation looked great. Short of Michael Bay, but easily hitting "Scifi Channel Original Series" quality. As a tech demo, which let's be honest is what it is, they couldn't possibly have nailed it harder.

    I was also quite pleased to see that the film didn't take itself too seriously plot-wise, definitely an improvement over Sintel.

  24. As a visual effects artist myself I am very excited about this project. I feel they did a very good job considering what they have to work with. For a indy project this was very well done. It isn't without flaws however.

    The main flaw for me was the quality of the robot (dog?) animation. To me it almost looked like stop motion animation. There was one scene that reminded me of the demon dogs in Ghostbusters when they bounced around. To me the attacking robot animation looked like old Ray Harryhausen material. This really left me scratching my head because the animated 3D elements attached to people was very well done. Can Cylces not render motion blur yet or was there just not enough time to finish off these creatures?

    Compositing was very well done but I was a bit surprised by some of the keys that were pulled. Some of the shots had softer edges typical of filtered 4:2:0 material and not edges one would expect from a 4:4:4 recording from a SONY F23. They were still very good and this is just a tiny nitpick. Some shots were keyed very well and others ok. To be honest it was much cleaner then the first Matrix movie which has hundreds of glitches. VFX students study it to see how sloppy rotoscoping can be in Hollywood sometimes.

    Acting was fine for what it is. It is very hard to pull off a VFX film mainly due to actors. It takes a massive amount of acting skill to make a made up world seem real. This is one of the most difficult things to do in VFX.

    Story was kind of confusing but I'm not I really care. I happen to like B grade movies and simple but fun sci-fi material. Heck I still love watching Star Trek and this story was way above that. The main problem I could see is that there was too much going on and too many characters to get to know for a short film. Story wise perhaps everything should have been condensed. I think the story would have been pretty clever and interesting if they could have played it out to let it grow for about 20 to 30 minutes.

    Super excited about this movie however. Very well done. It may not be perfect but it opens up a floodgate for other similar projects. Hopefully this movie will encourage others to take a stab at using Blender to create a VFX heavy movie.

    • Good comment about the robot animation, I also felt it looked stop-motion-esque. I thought it was intentional, but maybe it's a Cycles thing...

      I disagree with you about there being too many characters for this length of story - the ones we're not supposed to care about are only given tiny amounts of screen time so you don't invest in them emotionally anyway. It's clear who the main folk are.

      You also make a good point about the believability being as much to do with the actors as it is with the effects - a lot of people don't realise how difficult it is for actors to make you believe something is real when on set they were looking at tennis balls for eye lines and so on. For the most part I think the guys in this film have done a great job. Don't forget, every single one of them was acting in a foreign language, too!

  25. 18 months ago, few people could have seriously imagined a live action short being done entirely with Blender. Today, people nit pick that the lighting / compositing / roto / tracking are at 95% of Hollywood standards.

    Awesome job Ton + team. Your vision has raised people's expectations of what Blender is capable of, even when they don't realize it. :)

  26. I've watched this a few times now and can honestly say I really like it. It's clearly meant to be over the top and stylised. I applaud the Blender Foundation for choosing this project. I know from personal experience as a filmmaker it's a very difficult tone to strike, and when you add all of the sfx difficulties to the usual low-budget filmmaking struggles, it makes me appreciate the film all the more. I actually found it surprisingly emotionally affecting, given the ridiculousness of the story situation. The acting is credible and understated where appropriate, the music is excellent, the effects are very good, and the story gives you enough to let you imagine the backstory without spoon-feeding you irrelevancies.

    My biggest gripe with the effects is with the animation of the robots - or rather, the appearance of the animation of the robots. The animation itself is excellent, but they have a jerky sort of look about them, sort of like ED-209 from Robocop, which was stop-motion animated. I'm guessing it was a stylistic choice, since Blender's obviously able to render smooth animation. I guess I just wouldn't have chosen that aesthetic myself.

    I look forward to seeing more of Ian Hubert's films in future, whatever they may be. Well done to you, sir, and good luck for the future!

    Does anyone else think Ton Roosendaal is becoming something of a Roger Corman for the digital age?

  27. Perfeito, um belo trabalho desenvolvido.
    Parabéns a toda equipe.
    É de emocionar.

    Perfect, a beautiful work.
    Congratulations to the whole team.
    It's a thrill.

  28. Well done! It's amazing what can be done in a short time with a small team, small budget, and Blender.

    Most of what I might say has already been commented on. The best critique comes from @smetvid's, about the VFX. The author noted specific problems with obtainable solutions.

    @JosemariaRRA noted that the lighting looked like studio lights at times, and I think the critique was right, but not the solution. I think Blender needs a light tracking feature in the future . . . creating a light source from projected shadows. It could extend the camera tracking to remove the camera wobble, adding a manual/automatic ray-trace back to the source lighting. At that point, the physical light's characteristics could be used to define color, falloff, and so on.

    @l3clara pointed out that Ton seems to have an urge to kill off characters and even all of humanity . . . as Linda Hamilton said of James Cameron, Ton is "definitely on the side of the machines". What is strange is that I have been watching a lot of disconcerting Scandinavian films recently, and I seem to be getting the impression that for that region, blood and death are equivalent to snow and dim light. I also desire to see a less lurid and desensitized theme for the next movie.

    @Joeri said the ending credits were notable, and I agree (from time 9:50 - 10:30). Whoever put those 40 seconds together created an outstanding piece of art.

    @Earth_or_Bust might be on to something about taking the script too seriously. If that is true, though, imagery and acting is inconsistent with the script. The movie is not over-the-top, like more famous quirky dramas ("Delicatessen", "Hitchhiker's Guide to the Galaxy") and you only see the script's unorthodoxy in bits and pieces, like the post-trailer pile of robots, or the pipe-hanging strap holding on Captain's eyepiece. That witty reality gets lost in the callous grabbing of Old Thom's head, the obdurate meanness of Celia, and so on. So it is hard to tell what the director is really trying to do.

    My apologies to Ian Hubert, but I did not like the film. The plot was inscrutable though it was not meant to be, and I felt like there was not enough information in some of the scenes. This film is punished by achieving the project's mission: an efficient VFX pipeline. The imagery that Blender is capable of now allows the viewer to suspend disbelief . . . there is no visual distinction (besides a few glitches) between what is 3D and what is live footage, much like District 9. When that happens, traditional film elements (plot, acting, framing, etc.) become the critical foci.

    With the "historic" footage being a possible simulation of some sort, it becomes ambiguous to the audience whether the first scene of Celia and Thom is real or fabricated. I also am confused about what the goal of the "corrected" simulation is supposed to be. The simulation is important, I assume, or it wouldn't go on under attack. But I am not even sure why the robot dogs are attacking at all.

    Regardless, thanks for producing this film, and all the hard work that went into it. Working with live footage is exponentially more complex than creating everything from scratch. In 3D, building up levels of detail is the challenge; live acting requires reducing the complexity to fit the ideas presented. With this first project done -- no matter what the reactions are -- it solves a lot of problems for future work, for the Blender group and others. I am overwhelmed with the Blender metamorphosis. Two years ago, I argued unsuccessfully to use Blender over Lightwave; now the argument is simply a matter of cost or preference, and not about functionality or interface.

    Thanks again for this project, and I look forward to future works!

    • HI Delfeld !

      Your comment about Ton being to the machine side, is ratter unsettling. Blender is great as a tool, not by itself.

      You cannot put Blender above humans. Blender is great because it allows persons to fulfill their creativity (I think most of them wouldn't have other options).

      Without the people to use it, Blender is millions of line of code cloned in several machines, as useful as a bunch of sand in the desert.

      Anyhow, I hope next Blender Foundation Movie will be more inspiring.


      • With the quote, I meant to suggest that Ton is not very sympathetic to the humanity of all characters in the films. It seems that someone has to die to emphasize a point, or because they are evil, or to show that a character has been sorely abused. Killing -- as a plot device -- is not necessary, even if it is effective; I would like more broad, rational development of problems and solutions.

    • "I also am confused about what the goal of the “corrected” simulation is supposed to be"
      Someone somewhere on the web (I couldn't find it back) compressed the entire plot into one famous quote: Hell hath no fury like a woman scorned. Old Thom is trying to correct a mistake he made in the past through this simulation in order to try and stop Celia from destroying the world. The brains you see in the beginning of the film are HER brains controlling all those robots. In other words: the robot-arm she developed in her youth has been extended somewhat. At least, that's what I think

      • @JeroenM

        But it is not in the film. A film has to be understood on it's own, without context.

        There can be some discussion about contextual assumptions, such as that robots do exist and could look like dogs, which may not be believable 50-100 years from now. However, if I would make an 1800's period piece with robot dogs in it, the film is obligated to explain how they got there. Likewise, if the brain is a woman scorned, and she is _somehow_ becoming a robot, and now is mostly robot, there is nothing in our 2012 context that explains that, really, so more explanation seems needed. Also not explained are the robot dogs -- are they hers, someone trying to destroy her, or just a simultaneous robot apocalypse? The film also allows the possibility that the "scorning" scenes were never real, even the first one; one context that suggests this is that Hal from "2001: A Space Odyssey" got paranoid from thinking on it's own.

        And so on. A film has to be judged on it's own, not on the story or script it comes from. "The Color Purple" movie was extremely different than the book of the same name, and the author walked out halfway through the shooting because she disagreed with the director; Kubrick removed changed the role of the black man in "The Shining" form a critical role to a trivial plot point, which brought on accusations of racism which the book never had. A film is an explanation of itself, and nothing else can do it.

  29. This was much better than I was expecting, after seeing the trailer. The shortcomings and strengths were covered in previous posts. Might I suggest trying to work with Freddie and Brandon of Rocket Jump/freddiew or Sam & Niko of Corridor Digital. I have been a fan of their youtube work for years. They're very good at creating short concise and often humorous stories. And they often use SFX in their videos. I'd be ecstatic if they collaborated with the Blender Foundation on the next short film. Plus, it would be great for exposure. Both of their channels have a big following. And most who watch are interested in getting into SFX. And with limited budgets they might start using Blender if freddiew and/or Corridordigital used it.

    I enjoyed the story, but it seemed like one that needed to be fleshed out more to be clearer. To complex for this short of a film. I'm excited to see what future possibilities are in store with Blender. It can only go up. Great job guys!

  30. Great Job - i like it.
    About the story,quality of the fx and so on.
    May by you could make a few things better like some of the poster above say.
    An production of an short (or long) movie done with limited resources in time ,money and people is a very complex thing and need an good teamplay.
    Well done Ton (...and the team!)

  31. I'm blown away by the quality and the visuals. I realize it's not perfect and you can find shortcoming, but upon first viewing I hardly found any, I was so impressed and immersed. I didn't quite get the story, I probably need to re-watch it, but the visuals, the style, the music - I just loved it all! It's unbelievable what Blender can accomplish in the hands of some talented people.

  32. i love it but i wish there was more to it like a normal movie has even if it was only like an hr to an hrs and half... i know alot of ppl would love to see the full film story line of this for sure and so would i... i seen it befor and though it was only a teaser but i was wrong....

Leave A Reply

To add a profile picture to your message, register your email address with To protect your email address, create an account on BlenderNation and log in when posting a message.