• Hey, guest user. Hope you're enjoying NeoGAF! Have you considered registering for an account? Come join us and add your take to the daily discourse.

Introducing the next generation of VR on PlayStation

CamHostage

Member
one wire really isn’t a big deal. But I guess vr haters have to have something to complain about. I don’t see these same people praising the quest for its wireless features, or talking about vr in any other threads at all.

Where have you been then? The lack of wires on Quest 2 has made it a phenomenon this Christmas (also it has an easy price.) My sister wanted one, and she hasn't owned a gaming system since PS2.

Technically we're talking only about a secondary function of Quest 2, to tie wirelessly to a PC rigged with a local wifi network, not the main function of it being a stand-alone gaming system. But those who do link wirelessly say that there's no going back to a wire after playing VR untethered. Moreover, people who just pick up and play Quest regularly, no running wires and setting up lighthouses and cameras and all that "work" that goes into VR, they say that it's the way VR should be.

(Most don't know that wireless tethering is possible, even people who could do it; if instead of a Oculus Link Cable they sold a "Oculus Link Base Station" that was just a plug-and-play wifi box for Quest 2 to link to PC, I bet it'd be more clear and a bigger draw to convince users that it's a significant function.)

As for a PS5VR wireless, they haven't even shown the device. Why are people so sure that making a wired version rules out an wireless alternative? What's stopping them from making a dongle to transmit the signal and selling the wireless version for more?

Because a wired version is what they announced.

We can imagine some what-if scenarios if that's of interest, but in reality, Sony plans to release a wired VR headset for PS5 in the future, despite the breakthrough VR device of today being primarily wireless. (And also, affordable, so if the priority is that Sony releases the cheapest device possible, it better be super cheap if the competition has it beat on certain features.)

For a console that could conceivably do wireless VR, from a company that's not know for putting the cheapest, most baseline product possible on the market, that's what I'm questioning: why not wireless? If PS5 can do wireless VR, and if there's a $299 VR product on the market that handled wireless VR in 2020, what is stopping Sony's 2022 product from shooting for that same great functionality? Sony makes sexy products, and wireless VR is sexy.

(However, if somebody knew the specs and costs better and could say more clearly what the hurdle to it being wireless instead of just a guess at theoretical cost, that would be a good contribution to the discussion.)

Just sell a [6GHz WiFi] dongle with a battery pack as an optional extra. There's no need to make the headset more expensive than it needs to be.
Besides, I bet most PSVR2 games will focus on seated VR anyway, where the cable isn't that much of a problem.

If they could sell a wireless VR add-on pack, then fine. It'd be difficult to do and keep it ergonomic/comfortable, that kind of add-on would be something they should plan for ahead of time. (Although a simple halfway solution would be if you would have like a "VR Pocket Pack" and run the wire down to your pocket instead of all the way across the room to your console.)

But god no to seated VR being all PSVR2 can do, why even have VR at that point? There are some good in-cockpit or point-and-shoot games, but generally speaking, if you're not moving around in VR, you're not in virtual reality, you're just playing videogames with a silly helmet on...
 
Last edited:

Romulus

Member
But god no to seated VR being all PSVR2 can do, why even have VR at that point? There are some good in-cockpit or point-and-shoot games, but generally speaking, if you're not moving around in VR, you're not in virtual reality, you're just playing videogames with a silly helmet on...

I actually prefer seated VR sometimes, even when I have the option to stand and play. I love plenty of standing and moving VR games, but I would just as much like to play something like RDR2 seated in VR as I would something where I'm moving around. The motion controls and moving around stuff is like a novelty for me at times. It's cool, but the actual headset is what immerses me most. So yeah, a single wire is nothing, I won't even know it's there.
 

ABnormal

Member
The point is that they do not want a stand alone unit, but a VR system that has the processin power of ps5 (maybe adding foveated rendering, I hope), so it's really difficult (or costly) create a system capable to stream so much video and audio data at very low latency. Even the wireless Vive of Valve, although being a no-compromise costly technology, doesn't do a good job at it. Imagine a low cost unit.
If they would be able to do it wireless, you can bet it would be, or, at least, there would be a wireless version.
 

Unknown?

Member
I would argue that wired VR is rapidly being identified as "ancient technology"...

But yes, higher resolution, higher field of view, brighter, clearer, wider, betterer. That's all fine, but if they're just making a new PSVR to get a new&improved™ regular PSVR product on the market, they will once again fail to upsell PSVR to consumers who have been hesitant to get into VR (much less convincing the consumers who say they have no interest in VR.) Maybe the new input tracking and these comfy new ergonomic controllers will be a plus (they had a hard time trying to convince people that holding lightbulbs in their hands was "the future"...) but better visuals aren't going to do it when you could barely get people to put the headset on back when VR was hot.

If the competition is doing something new and it's working while PlayStation VR is doing more of the same, there will be no competition and ultimately I believe there will be no PlayStation VR.

You said it would be smart for Sony to get a wired version on the market to increase the install base (assuming it's cheaper, which is questionable since the competition's wireless, stand-alone device is the same price as old-and-busted PSVR1.) I'm saying, if they don't put out something that compels the audience (and it doesn't have to be wireless, but hot damn did that ever work for Quest 2, and not only is it a cool gimmick, wire-free is also by virtually all accounts the very best way to enjoy virtual reality,) they'll be chasing an install base of ghosts.
Requiring a FB account will be more of a deterrent than wires.
 
What a silly comment. Been waiting for that RE7 VR for how long now? 4 years and counting? How about Astro Bot? Also, this might be a shock, but not all people wanna game on a PC. Yeah, truly perplexing news, I know. I have a pretty powerful PC (RTX 3080) but the gaming experience sucks so I play on a PS5 instead.

I don't think it's a silly comment at all.
You listed RE7 and Astro Bot. 2 things. Except you can play RE7 VR on PC just not officially, so it's really just one thing.
For every one PSVR game that doesn't make it to PC there's a thousand PCVR games/apps/experiences that will never be on PSVR.
If you were only going to do VR in one ecosystem it's really a no brainer.

Hopefully PSVR 2 is so amazing that I can't help but open my wallet and get a PS5 and PSVR2. We'll see.
 
Last edited:

Gamerguy84

Member
Way to early as we don't know what's included but Kylie gave a good probable.

I wonder what the cost will be. I would assume cheaper than VR1 for reasons like no breakout box.

Plus I would think they want the price as low as possible for entry numbers to jump up. I'm buying either way.

Might even buy a PC VR after I get a 6800xr
 
What do you think for release date? Maybe late 2022 or Spring 2023?

Not sure about what the standards will be by then but the current standards I see are

~ 2000 x 2000 per eye ( this is on the higher end currently )
- They need to come with their own VR centric controllers
- High quality inside out tracking. No more camera or base stations. Good riddance.
- A single wire as an option is fine but wireless should also be an option. Some games like driving or flying or any seated experience is fine with a wire. But for anything standing or moving, wireless is a game changer.

The Quest 2 changed the game and all future VR headsets ( at least on PC ) will need to at least meet this bar. And in all likelyhood the Quest 3 will already be out before PSVR2.

VR headsets are improving quickly, like smartphones used to.
 

mdrejhon

Member
Yeah. You don’t get it. The only processing power the headset would need is to be able to convert a WiFi transmission to a image and sound. Any mobile chipset would be able to do that.
Inventor of TestUFO here -- I have helped some parts of the VR industry reduce display motion blur, so that VR is as comfortable as it is today.

Whether you're doing PlayStation VR or an Oculus Rift, or one of the headset, they all became really good because of some of Blur Busters advocacy to convince the industry to eliminate motion blur to reduce VR nausea -- they built upon each other's ideas, including "low persistence" now being standard in VR.

Now about video-over-WiFi:

The key is perceptually lossless video -- so requires a very good mobile chip with an E-Cinema quality codec at really high bit rates to make it look perceptually lossless.

Blasting 4K digital cinema bitrates (300 megabit H.EVC video) at 4:4:4 chroma with zero compression artifacts, making it look like wireless equivalent of a wired HDMI connection.

It's now possible within the envelope of WiFi 6E. Even Quest 2 WiFi 6 (non-6E) is almost perceptually lossless already sending triple-digit H.EVC bitrates (albeit not at 4K resolution). I currently can do 120 Mbps for 90fps H.EVC, and I can't see the compression artifacts in 99% of the PCVR games played wirelessly to the headset.
 
Last edited:

TheAssist

Member
Talk about dropping an update out of nowhere. Fingers crossed it can interact with PC in some way, I can’t stomach Oculus now that FB is required and Index is too pricey.
My thoughts exactly. Usually I would say there is no dice, but then again, they did put some focus on PC gaming the past few months.
A VR headset that you can just use on the console is a bit meh, I like the freedom of PC (VR) gaming but Facebook had to come and ruin some of the best headsets on the market. I dont even mind the price on the index, its more the fact that I dont want to bother with its set up.

So PSVR2 for PS5 and PC...that'd be swell.
 

Tygeezy

Member
Inventor of TestUFO here -- I have helped some parts of the VR industry reduce display motion blur, so that VR is as comfortable as it is today.

Whether you're doing PlayStation VR or an Oculus Rift, or one of the headset, they all became really good because of some of Blur Busters advocacy to convince the industry to eliminate motion blur to reduce VR nausea -- they built upon each other's ideas, including "low persistence" now being standard in VR.

Now about video-over-WiFi:

The key is perceptually lossless video -- so requires a very good mobile chip with an E-Cinema quality codec at really high bit rates to make it look perceptually lossless.

Blasting 4K digital cinema bitrates (300 megabit H.EVC video) at 4:4:4 chroma with zero compression artifacts, making it look like wireless equivalent of a wired HDMI connection.

It's now possible within the envelope of WiFi 6E. Even Quest 2 WiFi 6 (non-6E) is almost perceptually lossless already sending triple-digit H.EVC bitrates (albeit not at 4K resolution). I currently can do 120 Mbps for 90fps H.EVC, and I can't see the compression artifacts in 99% of the PCVR games played wirelessly to the headset.
I use a very modest 55 mbps with hevc. I find that gives me the best balance between visual quality and latency. My network latency will spike with higher bitrate. Wifi 6e im hoping we can blast 300 + mbps at 1-2 ms network input latency.
 

cyberheater

PS4 PS4 PS4 PS4 PS4 PS4 PS4 PS4 PS4 PS4 PS4 PS4 PS4 PS4 PS4 PS4 PS4 Xbone PS4 PS4
I use a very modest 55 mbps with hevc. I find that gives me the best balance between visual quality and latency. My network latency will spike with higher bitrate. Wifi 6e im hoping we can blast 300 + mbps at 1-2 ms network input latency.
I use similar settings. The image quality is very impressive even if there is visual artifacts. Not that you notice them when you are immersed in the game.
 
VR headsets are improving quickly,
and software is not. Every year since 2016 there's been new VR headsets, but good games come to VR in drips.

all I need from psvr2 is VR mode in major titles, specially shooters and racers. Hardware, as usual with consoles, doesn't matter that much.

I'm betting on March 2022.
 

Romulus

Member
and software is not. Every year since 2016 there's been new VR headsets, but good games come to VR in drips.

all I need from psvr2 is VR mode in major titles, specially shooters and racers. Hardware, as usual with consoles, doesn't matter that much.

I'm betting on March 2022.

I think for the install base size, VR has way more games than I would imagine. You're talking less than 12 million sales and it's got exclusive blockbusters that compare to consoles and it shouldnt have. Alyx has like 3 million people that could potentially play it. That's insane.
 

8bitpill

Gold Member
Give me that Jobe experience!

p91XcHU.gif
 

Gamerguy84

Member
I'm trying not to get too excited this far out. I'll let the hype go when we get a date on PSVR 2. I've never posted here before but I always read this thread.
 
Are you older than 35 years old? Because I still think that movie was top tier amazing!
46

Go rewatch it with more mature eyes and realize how crappy it actually is.

funny thing is, The Matrix was much better a movie. both VR-related movies opened and closed the 90s, respectively. Only in Hollywood VR was a thing for consumers in the 90s...
 
the part of the Nikkei interview that flatlanders forgot to mention:


"The next generation VR system is a very strategic opportunity for PS. We launched PSVR in 2016 and have had time to understand the VR experience for years. What I learned from VR is that the potential market is huge. I want to continue to provide a high gaming experience to keep the community entertained."

huge and immersive.
 

mckmas8808

Banned
the part of the Nikkei interview that flatlanders forgot to mention:


"The next generation VR system is a very strategic opportunity for PS. We launched PSVR in 2016 and have had time to understand the VR experience for years. What I learned from VR is that the potential market is huge. I want to continue to provide a high gaming experience to keep the community entertained."

huge and immersive.

Lets GOOO!!!!!
 


LightField Displays are an exciting area that no doubt will also be the future of tv, but they're not viable now. More likely psvr2 will use OLED microdisplays indeed. But who knows? Niche markets needing specialized micro solutions might be a good space to test tech like LFD - anyone remembers 3DS glasses-free 3D display? Perhaps Sony amuses us, then again micro OLEDs haven't been used either...
 
not official announcement, more like leaks from trusted sources:



foveated rendering with eye-tracking should allow any modern PS5 game to run in VR just fine. Btw, there's something called vibrating mode in the headset and it's not really haptics: there was Sony research into vibrations in the skull to prevent motion sickness. They're gonna win gamers really big.

goodbye, small indie minigames.
 
Last edited:

Audiophile

Member
Eye/Gaze Tracking + Foveated Rendering is my number one ask, it could cut the render budget enormously, seems like a no-brainer in the console space. I really hope they can pull it off, if anyone is willing to push something like this first, I'd expect it to be Sony. Not to mention the eye/gaze tracking can have functional gameplay implications too.

I also hope they stick with RGB-OLED, the SDE (Screen Door Effect) while still ultimately apparent due to overall resolution, felt less so thanks to the conventional subpixel structure. Plus, in regards to OLED itself, I really don't wanna go back to LCD and grey blacks. In addition, 2x resolution along each axis and ~20% more FOV would be nice, a little on the vertical axis would be cool too. One issue with the original display is the sample and hold image retention resulted in a smeared ghosting effect. I doubt we'll see PWM Impulse style refreshes in the OLED space yet, so perhaps a 240hz display with much higher peak brightness and BFI (Black Frame Insertion), or if they can just massively reduce the display response time instead, that'd be great.
 

Audiophile

Member
  • RGB-OLED: 4096x2160 (2048x2160x2)
  • Low Persistence 120Hz Sample & Hold or 120Hz PWM or Hi-Brightness 240Hz Sample & Hold w/ BFI (all for reduced ghosting and smear)
  • FOV: >115-Degree Horizontal, >120-Degree Vertical & >150-Degree Diagonal
  • Low-Abbe Non-Fresnel Lenses (for reduced chromatic aberration and other artifacts)
  • Eye & Gaze Tracking
...Pwease...?!
 
Last edited:

Audiophile

Member
What does foveated rendering mean?

In regards to VR, it rapidly tracks where the eyes are looking and renders/shades the centre of your vision at full resolution/detail then reduces it the further way it is from the centre of your vision.

For eg. You might only need the centre 20% of your vision rendered at 100%, then the next 20% rendered at 60%, the next 40% at 40% and the remaining periphery of your vision at 20%.

If it's done just right (and temporally stable), the full quality image and the reduced quality image should look practically identical.

Rendering anything in our peripheral vision at full resolution is basically an enormous waste and if the headset knows where you're looking it won't have to.
 

mdrejhon

Member
  • RGB-OLED: 4096x2160 (2048x2160x2)
...Pwease...?!
For perfect blacks, I'd prefer a MicroLED display over OLED.
Direct-view MicroLED can get brighter with faster response time.
Or even a 50,000-LED MicroLED backlight behind a LCD for local dimming.

A specially-engineered LCD like the Quest 2 LCD now does less motion blur than OLED.
Original Rift VR OLED had 2ms MPRT
New Quest 2 VR LCD has 0.3ms MPRT.

OLED has a law-of-physics problem getting bright enough to strobe brief enough. Backlights can be fan-cooled, heatsinked, and water-cooled more easily. Discrete LEDs are brighter, and now used in stadiums, and so you can get less persistence with a LED backlight.

Indeed, the past problem is LCD pixel response time is often too slow to fit in the VBI (blanking interval) between refresh cycles, but the Quest 2 succeeded in doing that fully (zero strobe crosstalk top/center/bottom).

It really is noticeably less motion blur during fast head turns. The Quest 2 LCD even has less blur than an average desktop CRT monitor, it's rather impressive how they've completely hidden LCD GtG on the Quest 2 in the dark cycle, making the LCD's pixel response completely unseen by human eyes on the Quest 2.

It is a rare well-engineered strobe-backlight LCD where 100% of the pixel transition is completely fitting inside the black period between strobed refresh cycles. The blacks are crap, but can be fixed with a local dimming backlight. And colors of OLED can be replicated by a NanoSys quantum-dot backlight.

  • Low Persistence 120Hz Sample & Hold
Unfortunately, 120Hz sample-and-hold can never be low persistence.

120Hz requires some form of strobing (impulsing, phosphor, black frame, one-flash-per-frame PWM) to eliminate motion blur.

MPRT(100%) is the refresh cycle time on sample-and-hold.
  • 120fps at 120Hz sample-hold = 1/120sec MPRT = 8.3ms of motion blur
    That's 8.3 pixels of motion blur during a1000 pixels/sec headturn
  • 240fps at 240Hz sample-hold = 1/240sec MPRT = 4.2ms of motion blur
    That's 4.2 pixels of motion blur during a 1000 pixels/sec headturn
  • 500fps at 500Hz sample-hold = 1/500sec MPRT = 2ms of motion blur
    That's 2 pixels of motion blur during a 1000 pixels/sec headturn
  • 1000ps at 1000Hz sample-hold = 1/1000sec MPRT = 1ms of motion blur
    That's 1 pixels of motion blur during a 1000 pixels/sec headturn
Low persistence is defined as anything 2ms MPRT or less.
Therefore, doing "low-persistence" simultaneously with "sample-and-hold" requires at least 500fps 500Hz.

In the absolute best case of GtG=0ms you still have motion blur -- at this point, MPRT(100%) persistence motion blur is exactly equal frame visibility time. As you track eyes on moving images, your eyes are in a different part of a continuously-displayed-static frame, smearing a static frame across your moving retinas.

For an animation demo of this, see TestUFO Animation: Eye Tracking Motion Blur to understand sample and hold motion blur better. It's a side effect of a finite/digital frame rate against your analog moving eyes.

Since these are currently unobtainium frame rates, we can't yet have low-persistence sample-and-hold yet...
  • Frame visibility time is always the guaranteed minimum display motion blur (even at GtG=0).
  • Frame visibility time of impulsed display is the flash length of the frame (length of PWM pulse)
  • Frame visibility time of sample-and-hold is the frametime.
display-persistence-blur-equivalence-to-camera-shutter.png


motion_blur_from_persistence_on_sample-and-hold-displays.png

motion_blur_from_persistence_on_impulsed-displays.png


Some useful reads related to this:
- Pixel Response FAQ: GtG Versus MPRT
- Blur Busters Law: The Amazing Journey To 1000 Hz Displays
 
Last edited:
For perfect blacks, I'd prefer a MicroLED display over OLED.
Direct-view MicroLED can get brighter with faster response time.
Or even a 50,000-LED MicroLED backlight behind a LCD for local dimming.

A specially-engineered LCD like the Quest 2 LCD now does less motion blur than OLED.
Original Rift VR OLED had 2ms MPRT
New Quest 2 VR LCD has 0.3ms MPRT.

OLED has a law-of-physics problem getting bright enough to strobe brief enough. Backlights can be fan-cooled, heatsinked, and water-cooled more easily. Discrete LEDs are brighter, and now used in stadiums, and so you can get less persistence with a LED backlight.

Indeed, the past problem is LCD pixel response time is often too slow to fit in the VBI (blanking interval) between refresh cycles, but the Quest 2 succeeded in doing that fully (zero strobe crosstalk top/center/bottom).

It really is noticeably less motion blur during fast head turns. The Quest 2 LCD even has less blur than an average desktop CRT monitor, it's rather impressive how they've completely hidden LCD GtG on the Quest 2 in the dark cycle, making the LCD's pixel response completely unseen by human eyes on the Quest 2.

It is a rare well-engineered strobe-backlight LCD where 100% of the pixel transition is completely fitting inside the black period between strobed refresh cycles. The blacks are crap, but can be fixed with a local dimming backlight. And colors of OLED can be replicated by a NanoSys quantum-dot backlight.


Unfortunately, 120Hz sample-and-hold can never be low persistence.

120Hz requires some form of strobing (impulsing, phosphor, black frame, one-flash-per-frame PWM) to eliminate motion blur.

MPRT(100%) is the refresh cycle time on sample-and-hold.
  • 120fps at 120Hz sample-hold = 1/120sec MPRT = 8.3ms of motion blur
    That's 8.3 pixels of motion blur during a1000 pixels/sec headturn
  • 240fps at 240Hz sample-hold = 1/240sec MPRT = 4.2ms of motion blur
    That's 4.2 pixels of motion blur during a 1000 pixels/sec headturn
  • 500fps at 500Hz sample-hold = 1/500sec MPRT = 2ms of motion blur
    That's 2 pixels of motion blur during a 1000 pixels/sec headturn
  • 1000ps at 1000Hz sample-hold = 1/1000sec MPRT = 1ms of motion blur
    That's 1 pixels of motion blur during a 1000 pixels/sec headturn
Low persistence is defined as anything 2ms MPRT or less.
Therefore, doing "low-persistence" simultaneously with "sample-and-hold" requires at least 500fps 500Hz.

In the absolute best case of GtG=0ms you still have motion blur -- at this point, MPRT(100%) persistence motion blur is exactly equal frame visibility time. As you track eyes on moving images, your eyes are in a different part of a continuously-displayed-static frame, smearing a static frame across your moving retinas.

For an animation demo of this, see TestUFO Animation: Eye Tracking Motion Blur to understand sample and hold motion blur better. It's a side effect of a finite/digital frame rate against your analog moving eyes.

Since these are currently unobtainium frame rates, we can't yet have low-persistence sample-and-hold yet...
  • Frame visibility time is always the guaranteed minimum display motion blur (even at GtG=0).
  • Frame visibility time of impulsed display is the flash length of the frame (length of PWM pulse)
  • Frame visibility time of sample-and-hold is the frametime.
display-persistence-blur-equivalence-to-camera-shutter.png


motion_blur_from_persistence_on_sample-and-hold-displays.png

motion_blur_from_persistence_on_impulsed-displays.png


Some useful reads related to this:
- Pixel Response FAQ: GtG Versus MPRT
- Blur Busters Law: The Amazing Journey To 1000 Hz Displays
I started to read your post, and I was like dang this guy knows his shit! Then I saw your avatar, and I'm like "Oooohhh the UFO monitor test guy!!!" I love your work and how much you have put into your testing and methodology! Keep up the amazing work!
 
Last edited:

iQuasarLV

Member
Was this blog an announcement for the announcement of the specs for the PSVR2, before an actual announcement of the product and showcase?!
Jesus .
 
Was this blog an announcement for the announcement of the specs for the PSVR2, before an actual announcement of the product and showcase?!
Jesus .

The blog posso was to officially confirm what they've said in countless interviews before: that VR is important to them and psvr will get a sequel.

then they showed the new controllers. Now there are reasonable leaks for the headset specs. We're expecting for the official reveal. You know, pretty much the same reveal in small steps of PS5 itself.

I don't care for hardware, I want a lineup.
 
lots of VR-related patents were published by Sony in previous years. I was pretty sure I had seen before their patent about vibrations reducing motion sickness which is why I immediately realized it was not really haptics. Here are they, uncovered by redditors:




without nausea and big modern games having VR mode, I think it's safe to say psvr2 will be the first real foray into mainstream for VR. more players the better, it means more games and budget for games made for VR...

Btw, awhile ago a patent from them to place ads in VR caused an uproar online. That too depends on gaze tracking obviously, which is why it was a sure bet the tech would be in.
 
Last edited:

Wonko_C

Member
Is Sony going after VR so hard because they know consoles have an end date approaching due to streaming?
They're doing it because it's a disruptive technology that has way more uses than just entertainment. Even more so nowadays because it's going to be useful in this new normal. They're already hiring and doing R&D at Sony, “with a view to five years from now”. (This is clearly not PSVR2 but something further ahead and outside of PlayStation)

 
Last edited:
Is Sony going after VR so hard because they know consoles have an end date approaching due to streaming?
VR and streaming are not mutually exclusive.
there are people right now playing VR games without a pc and without necessarily resorting to mobile chips with PS2 graphics.

both are futuristic disruptive tech still to get mainstream traction... streaming is just a new way to distribute content, VR is how you interact with and experience games from inside rather than looking at a tv.

are you still in the 80s sitting in front of a tv, pushing buttons in your console gamepad and munching grandma's cookies?
 
Last edited:
Top Bottom