• iOS/Safari issues when launching the site from mobile home screens should now be fixed! A2HS button compatibility on Android Firefox has also been fixed! Add some purple to your home screen today.

Everything You Need to Know about Dreams.

Oct 26, 2017
396
Germany
My guess would be to sell MOVEs. In an answer to this very question that I sent to Alex via Twitter, he said they haven't used it because the cable gets in the way, as do your hands when using the triggers. I sat there for quite some time, biting my tongue, wondering whether I should post a follow up response but I didn't. What I was just told was not the truth. The very reason the light is on the controller IS so it can be tracked by the camera, and the reason it's positioned the way it is IS so that the cable and your hands don't interfere with it when using the triggers. Many games have been using this feature successfully, so in conclusion, it HAS to be to sell MOVEs.
I think a more logical reason is that the don't want the main input to depend on the camera. If you think ahead a little bit (and they already kind of hinted at a PS5 version of Dreams) it would make sense to design your main control option in a way that works without peripherals/accessories that might not be available on a new system. And I think it's save to assume that PS5 will not be using a camera for tracking any more.
 
Oct 25, 2017
4,535
I think a more logical reason is that the don't want the main input to depend on the camera. If you think ahead a little bit (and they already kind of hinted at a PS5 version of Dreams) it would make sense to design your main control option in a way that works without peripherals/accessories that might not be available on a new system. And I think it's save to assume that PS5 will not be using a camera for tracking any more.
They could make it better with the camera while still working without.
 
Oct 25, 2017
4,487
Even after that thermometer explanation, I'm still having trouble understanding why stretching costs a lot. And I think it's very unnatural for the user. Like why not make the stretch tool just clone a bunch of copies behind the scenes, but look like a stretch to the user?
 
Oct 25, 2017
6,908
Can you export music on this? Hopeful to get in the Beta today.
theres no real export other than make a music file from a video file.

Even after that thermometer explanation, I'm still having trouble understanding why stretching costs a lot. And I think it's very unnatural for the user. Like why not make the stretch tool just clone a bunch of copies behind the scenes, but look like a stretch to the user?
haven't watched but was he stretching during sculpt mode?

during sculpt mode it retains the fleck size. but if you were to stretch that outside that mode it would make the fleck size bigger but shouldn't eat up thermo. stretching that inside the sculpt mode eats up the thermo.
 
Oct 25, 2017
6,864
Celle, Germany
Someone answered me about the missing codes in the beta forum:

MM said on the stream to contact Sony, because they can't do anything about it.
I contacted Sony - they said to contact MM.
Twitter Sony support also told me to contact MM. Well, fantastic, guess I can at least "enjoy" the beta forum. 🤷‍♂️
 
Oct 30, 2017
103
Even after that thermometer explanation, I'm still having trouble understanding why stretching costs a lot. And I think it's very unnatural for the user. Like why not make the stretch tool just clone a bunch of copies behind the scenes, but look like a stretch to the user?
Because that would only work with hyperrectangles and only for specific sizes?

Yes it is a bit counter-intuitive that simple shapes can take more thermo than seemingly more complex ones. Their optimization tool to reduce detail for shapes that don't need it because e.g. they were created the wrong way by stretching in create looks neat though.

Also many games in the stream showcased level linking which is nearly instantaneous. Making big dense games will not be a problem I think.
 
They showed some cool stuff. I want to do an animation test with dynamic (moving) trigger zones driving keyframe animation responses on a player controlled puppet. So just for example: an enemy attack (melee, projectile, whatever) is a moving, dynamic trigger zone, and if the player presses the right inputs while contacting that trigger, it will result in a dodge or something. If that works, then it's time to follow it down the rabbit hole. I already have some ideas about how that could work on a puppet.
 
They showed some cool stuff. I want to do an animation test with dynamic (moving) trigger zones driving keyframe animation responses on a player controlled puppet. So just for example: an enemy attack (melee, projectile, whatever) is a moving, dynamic trigger zone, and if the player presses the right inputs while contacting that trigger, it will result in a dodge or something. If that works, then it's time to follow it down the rabbit hole. I already have some ideas about how that could work on a puppet.
Could also work the other way around? Make the player a triggerzone and anything that enters (depending on what) could trigger an action?
 
Oct 30, 2017
103
Polish of the games have gone significantly up since last stream! I am very relieved to see characters with very neat animations and tight gameplay.

That platformer with imp as the main character <3

Also that game with the Greek columns was very funny!

What a thrill it will be to dream surf and be surprised by all these cool ideas.
 
Could also work the other way around? Make the player a triggerzone and anything that enters (depending on what) could trigger an action?
Yep. Could be applied either way. The enemy attack thing was just a quick hypothetical. What actually got me thinking of this was wondering how someone might do the iconic stealth kill camera shots in Tenchu. Those are relatively static, but then I was like, "Well if that worked... what else would".
 
Last edited:
Oct 25, 2017
4,487
Because that would only work with hyperrectangles and only for specific sizes?

Yes it is a bit counter-intuitive that simple shapes can take more thermo than seemingly more complex ones. Their optimization tool to reduce detail for shapes that don't need it because e.g. they were created the wrong way by stretching in create looks neat though.

Also many games in the stream showcased level linking which is nearly instantaneous. Making big dense games will not be a problem I think.
Yeah I'm in the beta, I understand the workarounds. I just think the current workflow is unhelpful in terms of these issues. A good UI should dissuade things which are bad decisions and offer guidance towards good decisions, but Dreams kinds of lets you do whatever and then you have to figure out why you've spent half your thermo on a mailbox.

I think for them it's second nature after several years working with it, but for new users it's like ..."u wot mate". I think there could be some UX improvements there.
 
Dec 26, 2017
166
So stretching in sculpt mode = bad. Stretching outside of sculpt mode not as costly. I kind of get that.

But they were also alluding to increasing the size of a sculpt in sculpt mode would have a similar effect as stretching....like combining shapes to make a more complex shape since that makes the sculpt larger. Is there a way around this? Plop down a bounding shape that is larger than what you anticipate your final sculpt to be and then cut into that? If you then add shapes to that cut sculpt and stay within the original bounding shape would you not experience this "virtual paper size duplication" they mentioned in the stream that causes the thermo to increase so much when stretching?
 
Oct 30, 2017
103
Well I know that whenever I get my hands on Dreams I will not stress about thermo use and stretching in/out of create mode. I'll just use the optimization tool whenever the thermo is full. Perhaps using clone tool is the only thing to keep in mind.

Unless someone in the beta says otherwise, stretching in sculpt mode then reducing detail with the optimization tool or stretching outside of sculpt mode directly should end up being more or less the same.

I think a good analogy is say you create a square in Dreams, see that as creating a 32x32 image. Then you stretch it in create mode this is like making it a 512x512 image which takes up a lot more memory/thermo space. Whereas if you stretch it outside create mode this is like keeping your 32x32 image but just rendering an up-scaled version. The upscaled 32x32 image is a lot less detailed than the native 512x512 image but since it is just a square you don't see a big difference between the upscaled image and the hi-res image. The hi-res image is overly detailed for what it is and needs to be compressed. This is the optimization tool that shows you which object consumes most of the memory. It is up to you to decide whether it is justified or not.

TLDR; Don't worry about stretching. Use the optimization tool when your thermo is full.
 
Oct 25, 2017
4,487
Well I know that whenever I get my hands on Dreams I will not stress about thermo use and stretching in/out of create mode. I'll just use the optimization tool whenever the thermo is full. Perhaps using clone tool is the only thing to keep in mind.

Unless someone in the beta says otherwise, stretching in sculpt mode then reducing detail with the optimization tool or stretching outside of sculpt mode directly should end up being more or less the same.

I think a good analogy is say you create a square in Dreams, see that as creating a 32x32 image. Then you stretch it in create mode this is like making it a 512x512 image which takes up a lot more memory/thermo space. Whereas if you stretch it outside create mode this is like keeping your 32x32 image but just rendering an up-scaled version. The upscaled 32x32 image is a lot less detailed than the native 512x512 image but since it is just a square you don't see a big difference between the upscaled image and the hi-res image. The hi-res image is overly detailed for what it is and needs to be compressed. This is the optimization tool that shows you which object consumes most of the memory. It is up to you to decide whether it is justified or not.

TLDR; Don't worry about stretching. Use the optimization tool when your thermo is full.
It really depends on the subject matter. If you're doing a soft, dreamy look then using the optimization tool is hardly noticeable. If you're going for very precise shapes with sharp, well-defined edges, you have to be careful about overdoing it and with how you build things.

That being said, so far I haven't hit any big thermo walls, so as John said it's a bit more guidance than hard limits. But I do wish the UI made those decisions more obvious to the user.
 
Oct 25, 2017
4,535
Yeah I'm in the beta, I understand the workarounds. I just think the current workflow is unhelpful in terms of these issues. A good UI should dissuade things which are bad decisions and offer guidance towards good decisions, but Dreams kinds of lets you do whatever and then you have to figure out why you've spent half your thermo on a mailbox.

I think for them it's second nature after several years working with it, but for new users it's like ..."u wot mate". I think there could be some UX improvements there.
the key point they made on the stream is that scaling things up *in sculpt mode* increases thermometer. That makes sense as sculpt is the raw data you’re working with so scaling up is effectively ‘adding more clay’ rather than stretching your existing material across a larger area. Outside of sculpt mode it is actually stretching the original so shouldn’t increase detail/thermo too much

The slightly confusing part was sculpting the one piece item as that is still large and therefore more detailed.

We could really do with more detailed or specific streams for tutorials. At the end of a general stream probably restricts how in depth they can go so the result can fall between two stools
 
Dec 26, 2017
166
The house someone from ERA made is a true work of art. Good job!
Yes, that was incredible. I hope there will be more from NinjaTrouiLLe in the same style. I'd love to walk around a themed environment like that in VR when PSVR support is added.

Congrats to all from ERA who had some of their work showcased on the stream. I got a big grin on my face when they mentioned that "Don't Sneeze" level.

Speaking of themes and styles. The characters people are creating are pretty cool. However, I would wish for characters to be published in collections so that if you are using them to make a game that they would all share a similar art style. Picking parts from multiple creators might make for some pretty non-cohesive levels/games.