Dreams is upcoming software (in early access now) for the playstation4.
Its from the makers of “little big planet” (Media Molecule) which has a great level designer in it. Dreams has taken the concept of level design to new heights and is just a full on game engine. 3d modeling, Rigging, Animation, Materials, Realtime NPR renderer, Visual logic/programming, and a full on music creation suite.
Early access launch trailer:
I’ve been using it for a while and although its control system is very hard to get used to, I really think its one of the most amazing things made in a long time. It’s so cool its crazy. I can’t imagine what kids growing up with these tools today will make when they are our age.
Anyway, for a very basic comparison to Unity:
I created the following GFX synced to midi (using the OPZ) as a test in unity:
and here is a Dreams version I made of the same thing (I have not done any music yet, just leaning the Dreams tools and I got a bit carried away… its work in progress).
It was much faster to create the Dreams version, and the Dreams version is more fleshed out. Not having to learn 2+ software packages and how they work together and jump between them with import/ export issues is a huge savings in itself. On top of that, if you are patient enough to learn the dreams ways of doing things, (using a game controller or the motion controls and all the software quirks) its quite streamlined for what it is.
I will go on with some other things but just a first post to get started.
Am I a bad person for liking the Unity version a lotttt better ? Also the OPZ track was nice.
Other than that what I’ll be curious about with Dreams will be what kind of gameplays emerge from it. I haven’t seen very interesting things on that aspect yet, mostly crazy visual stuff which is really cool to look at, and from people who apparently have 0 knowledge of 3D creation which is absolutely crazy considering how hard it is to learn how to do that with current software offer. I should try it sometimes see how far you can go into personalization vs assets and texture packs already made out for you. Edit : I should add : How far can you go into personalization “without it becoming as complex as any other game engine to get there”.
hah, no you can like whatever one and thanks for that!
the dreams version is just a test scene made from scratch as I learn the tools. I do believe I could recreate the exact same thing in dreams as that simple unity scene, but I got side tracked along the way trying different things.
I have not used any pre-made assets yet, although I think that is the answer to having people with no experience be able to use the tool.
I don’t have enough experience to make any judgments, but so far I am really impressed at the quality level and the range of looks that you can create – enough to have your visual style come through and to create what you have imagined. That is enough of an achievement (imho) to be considered a truly great one, and everything else is an interesting bonus.
Making noisy / low pixel count stuff PS1 style / that animates well and behave like pixel art in Dreams would actually be an interesting use of the medium since most of what I see goes into the complete opposite direction of “Hey I can make AAA graphics alone!” (no you probably can’t ^^) ! Yeah I’m curious too I’ll see what comes out of it ! I wasn’t fond of what came out of the “little big planet” level builder to be completely honest, but this seems like a completely different thing. We’ll see if the timing’s right too, because it looks an awful lot like microsoft’s “Sparks” which failed on such a massive scale for coming out too soon with this idea.
I read an article about this program this week, I’ve had to avoid thinking about it because I can’t invest in another game console. Why isn’t this on Switch or PC…gaaaaah.
Otherwise, it looks exactly like what I want a creative VR world to look like. I’m really bummed it’s not something I can get into right now.
You are in luck, the same post filter i used to pixelate in unity is available in Dreams
I have definitely noticed that (incoming generalization) most of the discussion on reddit does center around ‘can it be photorealistic’ as the only bar to measure the quality of the render.
Personally, the non-photorealistic render is much more interesting to me because it does not exist (within easy reach of non programmers) in the typical render pipelines available to casual users. Even larger animation studios would have to spend quite some time and money to achieve the NPR look of the dreams engine.
To me it is very much like film/game concept art, in that it’s very rough, but beautiful, and can be created relatively quickly to get an idea across in a very visually appealing way.
The visual programming is also pretty elaborate, with basic nodes for math ops, but also advanced nodes for ray tracing or movement/velocity data. When used to drive music/visuals or when sroven by music visuals, it can be pretty versatile!
Thanks for the more precise answer ! I actually think Dreams as a 3D engine for live audio performance integration would be quite fantastic since its ease to create visual landscapes easily would shine and the gameplay part would be more in terms of “how do this react to sound” which would avoid the whole game design aspect and replace it with simpler concepts. (Because I think game design is still a very complicated and tough to master art no matter what the tools are, proof of that is how apparently complicated it is for even Media Molecule to make a full blown “example game” out of their own engine). Otherwise there’s often this misconception / focus about the fact that what makes a game hard to create is the for the graphic part to be super technically impressive, and although it its true and a great thing to try and tackle that particular challenge, I really don’t think that’s the hardest part (at all).
Actually that may be the biggest single bad thing about it : an open ended game engine tied to a single physical proprietary videogame machine is absolute nonsense and pretty sad.
It’ll definitely, absolutely, completely be on PS5, and if the games catches up steam with its community and have some nice creative experiences in it, they’ll make a big deal out of 1.0 being on PS5, I’m actually pretty convinced that’s why they only released an “early access” on PS4 in the first place. They use it to build up the community, maybe let media molecule make their long awaited “example game” out of it that they seem to have such a hard time with, and they’ll release the full blown package on PS5.
I guess media molecule make playstation video games, and this is there ‘be the creator’ creation philosophy come to full completion, so of course its on ps4.
…but I completely agree, it seems absolutely crazy, considering the amount of work that went into it and how many other uses there are for an engine like dreams, for it to be locked to playstation, and not even to have an import/export at least; a reaction that comes from total amazement of what they have created.
There have been rumors mostly in the form of cheeky non-answers and winky wink expressions, that PC may be an option in the future… but of course nothing real.
Here’s a look at the visual programming, I made a bare bones flocking setup:
The birds in the test scene at the end are pretty hard to see — they get lost in the video compression. Still working on how best to get content out of the ps4.
I would love to do a ‘skill trade’ with someone who might be able to write an OSC app on norns. Just pipe dreaming…
and thanks for that!
