I'm confused....are you saying you feel if they THINK you have a gun, they can punt you in the face and break your jaw even if you are complying with their orders? Because he WAS complying with their orders and getting on the ground when he was kicked in the face, and cops can always say they THINK you had a gun, then act violently with impunity.
The radio repeatedly said the man with the gun has a yellow shirt and a hat, this guy has a white shirt and yellow hat, so he doesn't fit the multiple clear descriptions. Why would they think 'the dude had a gun'?
I also note, that was the SECOND time that officer kicked him, the first time was as he tells him for the first time to 'get on the ground' while kicking the suspects legs while the suspect is complying with the original 'get your hands up' command.

Also, they lied about the suspect and falsely charged him with multiple crimes, as noted here....
"The incident occurred at a gas station on U.S. 13. Lateef Dickerson, 30, of Dover, was originally charged with assault, theft and resisting arrest, but those charges later were dropped." Surprise, he apparently also did not have a gun at all!

On top of that, this is the official police statement where they are saying the officer did nothing wrong...
"As the man was in the process of getting on the ground he was kicked in the head once by Webster rendering him unconscious." I can't fathom how they think that's OK, they admit he was complying but kicked him in the face anyway, but that is their position.

They also said "that his actions regarding this incident were outside of Dover Police Department policy and acted accordingly," which seems to mean we know he didn't follow guidelines, and that's just fine (I don't know what else 'acted accordingly' could mean in this case). If they mean to say the department 'acted accordingly', it should be noted that their only reported actions were to give him a paid vacation and a pat on the back.

typical trailer full of marketing bullshit, most of that was pre rendered.. and badly at that. NOTHING was reactive or interactive, and if hey want any kind of multi user environment like a shooter ground or what ever that Role player crap was, you are going to need something a bit bigger then anything they showed in there.

I hate myself for this, but.... the scientist/engineer in me has to complain a bit about the central idea in the plot (I'm pretty sure the plane engines would use more water than you could possibly recover and I don't think any magic gas would make a cloud go bigger). Sorry, I literally had to say that.

Now, that's out of the way, I pretty much love the rest of it.

It's beautifully animated and I love the bits of world building like the submarine. The rendering of the rain starting to fall was particularly impressive.

At the very end, after the playstation logo, it reads "FROSTBITE GAME ENGINE FOOTAGE REPRESENTATIVE OF PLAYSTATION 4. NOT ACTUAL GAMEPLAY.".

So while it may be rendered in-engine, the wording makes it sound like this wasn't done on a PS4 and it's more like an in-engine cutscene, using animations, effects, and physics that won't be seen in the actual gameplay.

Yeah, it's possible, but it's an order of magnitude more difficult than just rendering it.

Part of me thinks this is an elaborate hoax, but if not? This is some *quality engineering

oh well... it actually takes quite a bit. let's start backwards:

posting links to the video on social networks, twitter, promotion and stuff like that: 20 min

adding captions, notes, etc to the video: 10 minutes

uploading the video and filling in the information: 10 minutes

rendering the video: 2 hours

(this week I made an editing mistake and had to change and re-render the video which took and additional 2 hours)

actually editing the video: about 5 hours (which could be less, if the editing preview wasn't so unresponsive and slow now. it makes the work really frustrating and inefficient)

making the animated talking logo in after effects: 15 minutes

recording the voice-over: 1 hour (I make a few mistakes, and then I also need add a few filters to make it sound better, 'cause my mic is kinda crappy)

collecting all the videos I need for backdrops: 2/3 hours

actually writing the script while reading the news I had previously chosen: 2 hours

getting all the release dates down: 1 hour

picking and choosing the news I'll be talking about: 1 hour spread throughout several days.

So on total: around 15 hours. Obviously not all on the same day. Most of it on fridays.

Rumour is that MS's W10 browser will support Chrome extensions. Somehow.

I'm somewhat curious. I'm falling out of favour with Chrome, it gets bogged down and chuggs badly on mobile CPUs when you use font scaling. Also Google's application design is increasingly becoming restrictive and frustrating. Why is there no way to turn off auto-update? Why is the interface locked from modification and I'm forced to squint painfully at the address bar on my 1920x1080p 13 inch laptop?

Firefox has terrible font rendering and I haven't been able to improve it. I originally moved over from it because it seemed to have memory issues where over time videos would randomly freeze for like half a second during play intermittently, dunno if they fixed it.

Opera seems better in these regards but I can't for the life of me get used to browsing without alt + # for switching between tabs.

Meanwhile IE 11 handles font scaling well and memory wise seems fine, but also lacks basic extensions at the moment. If they fix this issue I may switch over at least on my laptop.

The tool used only uses 400mb of RAM according to the YT comments, and I'd imagine memory is the major obstacle to loading an entire games worth of maps all at once. Well, unless you also want an entire games worth of AI running, and you want to render everything without LoD or any consideration for occlusion (but then does it matter what engine you are using?).

Cool. I'd be more impressed with an engine that could render that in game, especially with today's graphical detail... might be possible due to the low polygon count and lower texture quality that something like the Frostbite engine could perhaps pull it off, but with better polygon counts and texture quality would impress me.

Still the amount of work to overlap the maps at the proper spots and seal things off is fairly impressive. Undoubtedly took a great deal of time.

It's not just slavery.

The bible is also big on stoning your kids to death, forcing rape victims to marry their rapists and cutting off the hand of a woman who breaks up a fight. Hell, if you're a dude who's had a really unfortunate accident, no church for you!

As soon as the bible is no longer a perfect guide to morality (which it's clearly not), it is rendered completely useless.

Nothing was moved or interacted with in the entire clip, so it leads me to believe that all of the lighting and shadows were simply pre-calculated and baked on as a texture. In the same manner, the reflections are most likely environmental cube maps, rendered in 3D animation software with all the settings turned up and then saved as an image. If something besides the camera were moving then I would be impressed.

tldr: Actually, games do this all the time, but usually only for water surfaces!

The reason for this is that the way you render a proper reflection is to "flip" the camera to the other side of the reflective surface plane: looking down on a lake, you'd render the water reflection from the point of view of the camera looking up from under the water surface, flipped over. This is called "planar reflection". In order to do this, you render your entire scene again, so it's not cheap. Also, the reflection only works for that one plane: if you had two altitudes of water (or two differently angled mirrors) they'd be on different planes and so you'd have to render a reflection for each one.

You can't render curved surface reflections this way, though. For example it doesn't work on a car (what plane would you flip the camera over?). For that, the trick is called "cubic environment maps". I won't go into the details, but it only really works well for faking reflections on objects since it shows the correct view from a single point. You can create them dynamically for things like racing games, but they require 6 scene renders (one for each face of the cube) for each environment map.

Half Life offered both techniques for water reflections, so one could fire that up and compare them that way.

This demo seemed to use environment maps for the mirrors and I suspect all of the other shiny surfaces.

Note that these techniques are to get detailed reflections: specular lighting (where you don't reflect an image, but instead mathematically simulate simple light bouncing) is easier and cheaper, since it's just math to get a color and strength.

You could do planar reflections for every mirror, but it's a full scene re-render for each one so your frame rate would tank or you'd have to take out other features. Compromises!

Game graphics is all trade-offs and smoke and mirrors: it's our job to fake things and make you think the game is doing sophisticated simulation when actually it's doing as little as it can to get as much as possible.

Yes, semi impressive, but static scenes are in the end static. These are useful for architectural groups and the like to show off stuff. Virtual builds for a client to walk through.

For gaming, one needs to see the scene in a game like state with action going on, and high polygon models moving in and around the high polygon scenery. There is where we get our real test.

Still pretty, and we are getting closer and closer to near realistic stuff rendering out in real time.

Please, if you will read my original comment, it's about tracking the dancers in real time and conforming the effects to their actual movement, also in real time. Somehow that was lost in the replies. I get that the image is being rendered in real time but that's not what I was commenting on. (even though it's pre-rendered video background, the effects and 'camera' and lighting are changed on the fly, but not by the dancers positions/movements)
Your term 'real time interactions' describe what I mean better than I did.

Why would it have to track the dancers to be considered realtime? All my games are real time 3d, they certainly don't track me dancing, or doing much of anything really.

I think the term we want to use is pre-calculated. The particle movements were chosen before hand much like a motion capture, but the rendering is still realtime and thus thing like camera angles can be changed.

I was disappointed because I noticed the lack of true interaction, and when somethings the performers did would effect the particle effects while others wouldn't. That annoyed me even. We have years of tech to monitor actors in real time space, it may have taken a bit more work but this could all have been done with realtime interactions.