Support NeoGAF

[Eddie-Griffin]

https://www.techgoing.com/meta-quest-vr-headset-now-supports-unreal-engine-5/

Meta Quest announced that Meta engineers have recently completed the necessary work to transition to Unreal Engine 5, providing official support for Unreal Engine 5.1.

Meta Quest says that v49 will be the last update to include new features for Unreal Engine 4, and v55 will contain the last patch for Unreal Engine 4 (currently scheduled for release in June 2023).

After the release of the final Unreal Engine 4 patch for v55, VR Support will no longer provide work orders related to Unreal Engine 4. There will also be no new features or routine bug fixes released for Unreal Engine 4 after v55.

Meta Quest recommends that developers do not upgrade their Unreal Engine version for now and can still use Unreal Engine 4 to distribute their apps to the Meta Quest Store or App Lab. If app development upgrades to Unreal Engine 5, architectural changes may result in minor app performance issues. Critical bug fixes will be available for Meta Quest until the end of 2023.

The announcement that Meta Quest 1 will no longer be supported, however, Meta plans to provide an upgrade path for developers.

Quest 2 will now shift support over to Unreal Engine 5 going forward, making the last Unreal Engine 4 update the final one the company will give.

Quest 1 will be left behind and discontinued, although they promise an upgrade path.

The fact they are making the full shift to Unreal Engine 5 most be to get a head start on the Quest 3, which is probably going to be the real beneficiary of Unreal Engine 5. While Quest 2 and Quest Pro (I assume) will support it, I expect it to be limited with much of the support for it being for Quest 3. It seems that they are aware they need to catch up for better quality and produced game titles if they want to still have a competitive product, along with keeping up with the industry and making it easier for devs who are also transitioning.

I expect that the Ventura headset (which will be cheaper than Quest 3) will replace the Quest 2 with more modern tech and features to be the real low-end headset supporting Unreal Engine 5 with Quest 3 at the top.

One thing you can say is that they want to address the output quality and retention issues by near-mandating the new standard instead of what people expected, which was for them to add (if at all) limited support and drag older tools. By doing this they will raise the minimum bar for game output ambition and quality.

 

[Ozzy Onya A2Z]

Wasn't the Quest 1 released in 2019? Only 4 years support before sunsetting? That's pretty short. Also Quest 2 dropping UE4 support already? I doubt Unreal are dropping support for UE4, pretty average developer support really.

It seems like the double hit bad news was tucked away and rolled into a PR release about UE5. Perhaps I'm off base?

 

[Eddie-Griffin]

Wasn't the Quest 1 released in 2019? Only 4 years support before sunsetting? That's pretty short. Also Quest 2 dropping UE4 support already?

I mean it's effectively hardware wise, sunsetted earlier last year, but with software support going on. Now it looks like that is on the verge of ending.

But yeah, that was an issue about these VR headsets, they are basically having to drop the previous versions or pull support if the next one ends up growing form where the previous left off to capitalize on it. I suspect by end of 2024, Ventura will fully take over Quest 2 and that will be dropped too in only 4 years.

It raises question marks for people thinking about, or who were already in VR as both groups will have to wonder that if they get in 2-3 years later how much support will they get before they are left in the dust for another headset?

I doubt Unreal are dropping support for UE4, pretty average developer support really.

It seems like the double hit bad news was tucked away and rolled into a PR release about UE5. Perhaps I'm off base?

No I wouldn't say you're entirely off base, they already had some devs, including Zenith, stop updates on PSVR1 and Quest 1, but for Quest 1 specifically, they didn't just drop that but also the ability to play the game, while the PSVR1 version will be playable but stuck forever on the last update with no support.

Quest 2 has more active users and sales so you would think that's an odd decision, but when you think about it it's likely that Facebook has been hinting at killing the Quest 1 off for awhile.

Like I said, I expect them to do this with the Quest 2 next year (otherwise the Ventura headset would be pointless) and people should expect 3 or max 4 years of support for headsets.

At the same time though, this news does mean Quest 3 will be modern and will have a higher minimum for game production and quality. Which may make Quest 3 seem more appealing.

 

[Ozzy Onya A2Z]

I mean it's effectively hardware wise, sunsetted earlier last year, but with software support going on. Now it looks like that is on the verge of ending.

But yeah, that was an issue about these VR headsets, they are basically having to drop the previous versions or pull support if the next one ends up growing form where the previous left off to capitalize on it. I suspect by end of 2024, Ventura will fully take over Quest 2 and that will be dropped too in only 4 years.

It raises question marks for people thinking about, or who were already in VR as both groups will have to wonder that if they get in 2-3 years later how much support will they get before they are left in the dust for another headset?

I looked up Unreal support and it seems 4.27 support has ended but hotfixes might still be a thing, where Meta has just ruled that out altogether. Pretty rough for devs to be looking down the barrel of a UE4 game and upgrade to UE5 post launch or mid dev. That's not going to happen, which really sort of raises the question of timing for hardware sets vs Unreal support. I guess we're not going to see GaaS style games in VR at all or meaningful updates to games for a few years yet; in terms of DLC etc. Perhaps I'm reading too much into it and UE5 will kick off another decade of fresh support with fresh headsets and titles.

The install base and specifics of optimising for different headsets compounds the already complex and risky development issues for VR. Tough gig at the moment. Hopefully Unreal 5 makes it a lot easier of VR devs going forward but checking Unreal forums it appears most devs prefer the stability of UE4 and UE5 is a memory pig so far.

 

[Eddie-Griffin]

I looked up Unreal support and it seems 4.27 support has ended but hotfixes might still be a thing, where Meta has just ruled that out altogether. Pretty rough for devs to be looking down the barrel of a UE4 game and upgrade to UE5 post launch or mid dev. That's not going to happen, which really sort of raises the question of timing for hardware sets vs Unreal support. I guess we're not going to see GaaS style games in VR at all or meaningful updates to games for a few years yet; in terms of DLC etc. Perhaps I'm reading too much into it and UE5 will kick off another decade of fresh support with fresh headsets and titles.

The install base and specifics of optimising for different headsets compounds the already complex and risky development issues for VR. Tough gig at the moment. Hopefully Unreal 5 makes it a lot easier of VR devs going forward but checking Unreal forums it appears most devs prefer the stability of UE4 and UE5 is a memory pig so far.

It depends what tools Quest 3 will give devs, or other headsets. PSVR2 will have some support for it, the Firewall game coming out later is made with it. But we still don't know what tools Sony has to help devs utilize it for VR, and we don't know that for Quest either.

Then you have VR like Shitfall Mega Man, which aren't even giving devs tools just providing the headset that should support the engine spec wise.

It seems that VR is heading toward acceleration onward without caution and leaving the devs behind that are still in transition. Even more so since Quest is the market leader.

 

[Ozzy Onya A2Z]

It depends what tools Quest 3 will give devs, or other headsets. PSVR2 will have some support for it, the Firewall game coming out later is made with it. But we still don't know what tools Sony has to help devs utilize it for VR, and we don't know that for Quest either.

Then you have VR like Shitfall Mega Man, which aren't even giving devs tools just providing the headset that should support the engine spec wise.

It seems that VR is heading toward acceleration onward without caution and leaving the devs behind that are still in transition. Even more so since Quest is the market leader.

Hopefully the focus isn't just on headset tech advancements and dev pipeline really comes to the party too. If great tools aren't there the hardware potential will largely be irrelevant in what manifests in the games themselves. Another tough point on VR is having to support specific features of one headset and not another, I suppose that's largely why we don't see mass support of big titles across a number of headsets. The other thread you had about open standards is crying in the corner on this point.

 

[UltimaKilo]

Wasn't the Quest 1 released in 2019? Only 4 years support before sunsetting? That's pretty short. Also Quest 2 dropping UE4 support already? I doubt Unreal are dropping support for UE4, pretty average developer support really.

It seems like the double hit bad news was tucked away and rolled into a PR release about UE5. Perhaps I'm off base?

Quest 2 is woefully underpowered, but compared to Quest 1, it was being held back significantly. It will be a few years before Qualcomm releases something far ahead of the XR2, unfortunately.

 

[Stitch]

Cool I guess.

Most Quest games are made with Unity.

 

[poppabk]

Most devs will probably stick with UE4 anyway, but Quest 1 as a standalone is on its way out anyway. Not really worried about Quest 2 being dropped both because of the install base and 95% of my stuff is on Steam anyway.

 

[Ozzy Onya A2Z]

Quest 2 is woefully underpowered, but compared to Quest 1, it was being held back significantly. It will be a few years before Qualcomm releases something far ahead of the XR2, unfortunately.

It's nuts how many devices have Qualcomm chips, I mean they were always popular but in recent years/devices they've been everywhere. I wonder what Apple's headset thingy is going to be running, proprietary like their M1/2 chips?

Cool I guess.

Most Quest games are made with Unity.

Didn't know that. Cheers.

 

[Eddie-Griffin]

It's nuts how many devices have Qualcomm chips, I mean they were always popular but in recent years/devices they've been everywhere. I wonder what Apple's headset thingy is going to be running, proprietary like their M1/2 chips?

Apple is using the M2 chip in their headset in combination with other things we don't know yet.

Samsung will be using Qualcomm though given that they announced a partnership with them for their upcoming headset this year.

No idea what TCL is going to use, but apparently it won't be Qualcomm.

Those are the big 3 people have their eyes on, though there's like 40 other headsets releasing this year, or already have.

 

[Minsc]

Apple is using the M2 chip in their headset in combination with other things we don't know yet.

Samsung will be using Qualcomm though given that they announced a partnership with them for their upcoming headset this year.

No idea what TCL is going to use, but apparently it won't be Qualcomm.

Those are the big 3 people have their eyes on, though there's like 40 other headsets releasing this year, or already have.

So Apple's headset by itself could actually be faster than a PS5 console? That would be a decent start. Given that Apple likes to iterate on products quite often, I'd expect to see ~3 headset revisions per console generation, so there's a good chance gaming-wise it would be able to run everything current gen, each generation, assuming it is even targeted towards gamers - but given their revenue split on iOS, they'd be foolish not to.

 

[Alexios]

Quest 2 is woefully underpowered, but compared to Quest 1, it was being held back significantly. It will be a few years before Qualcomm releases something far ahead of the XR2, unfortunately.

There's already a new range of snapdragons with stuff like ray tracing, it remains to be seen if Quest 3 has dibs on an XR version of those or something else altogether but it can't be just more XR2 beefed up like seen from Quest 2 onwards in all similar devices if they promise a decent spec bump.

 

[Buggy Loop]

There's already a new range of snapdragons with stuff like ray tracing, it remains to be seen if Quest 3 has dibs on an XR version of those or something else altogether but it can't be just more XR2 beefed up like seen from Quest 2 onwards in all similar devices if they promise a decent spec bump.

From SadBrad, Quest 3 supposedly has the XR2 Gen 2 Adreno 740, so with ray tracing.

Adreno 730 variant had 2.4 TFlops so it should hover in the range of 3TF which is not bad. 2.5-3 times performance increase over previous chipset and a lot more GPU modern features. Tensor accelerators for AI. Not GPU but Wifi 7 (airlink upgrade!), support for AV1 codec (airlink upgrade again!)

Should be a really solid upgrade for standalone and also raise the bar for PC streaming

Only one on the block that can beat this chipset is probably Apple's M2 GPU that is rumoured to be in their AR/VR headset, but at what cost, it won't even be close in price.

 

[Eddie-Griffin]

So Apple's headset by itself could actually be faster than a PS5 console? That would be a decent start. Given that Apple likes to iterate on products quite often, I'd expect to see ~3 headset revisions per console generation, so there's a good chance gaming-wise it would be able to run everything current gen, each generation, assuming it is even targeted towards gamers - but given their revenue split on iOS, they'd be foolish not to.

It will raise the bar for the developers get to come into their ecosystem, they'll probably have some decent tools as well.

The OS will probably iterate the same with the hardware in line with Apples other products like Iphone, since the OS is supposed to be compatible with iphone and ipad with continuity, and it has hand-off functions as well.

 

[Minsc]

It will raise the bar for the developers get to come into their ecosystem, they'll probably have some decent tools as well.

The OS will probably iterate the same with the hardware in line with Apples other products like Iphone, since the OS is supposed to be compatible with iphone and ipad with continuity, and it has hand-off functions as well.

Hmm, not sure how a desktop/mobile OS would work on a headset without physical controls or a screen to touch. I guess maybe if the camera is super great it could impose a touch screen in VR and project the UI there and let you run all the Mac OS or iOS apps on the headset?

I guess there is too much unknown. I was thinking though, I'm the only one I know with a VR headset (which isn't too surprising, they're not that popular), but once Apple releases their VR headset, I'm pretty sure I know 4 or 5 people who would pre-order it right away, even at $3000. So maybe it will be more of a success than other brands.

 

[DenchDeckard]

Pretty cool.

Looking forward to quest 3. Hurry up and announce it already

 

[Eddie-Griffin]

Hmm, not sure how a desktop/mobile OS would work on a headset without physical controls or a screen to touch.

I said it has it's own OS that is compatible. Which is why it has continuity and Hand-off.

As for compatible apps, as was described in the other thread, you can shift your games and documents across the devices to and from. Some of which may use the "in-air typing" of the VR headset, although I'm sure they will allow keyboards too since there's a report that "in-air typing" may not be "perfect" at launch with some issues.

I'm curious however how the interactive software apps games or otherwise like say 3D modeling apps or etc, will work on this headset using continuity.

the UI there and let you run all the Mac OS or iOS apps on the headset?

I guess there is too much unknown. I was thinking though, I'm the only one I know with a VR headset (which isn't too surprising, they're not that popular), but once Apple releases their VR headset, I'm pretty sure I know 4 or 5 people who would pre-order it right away, even at $3000. So maybe it will be more of a success than other brands.

I would bet that they will probably have compatibility with the headset on an app by app basis, based on what extra requirements are needed for the app to work. basic apps that can all be used the same way will be available at once, but for others they may need to make it so that the App on the headset is used differently when moved to the headset.

But yes many people are suspecting even before they show it, the Apple headset to change the game, and a growing interest in Samsung and TCL as well, which they are waiting to on more info on.

But already we have had numerous companies react to Apple. In fact, a couple are rushing to release to beat them, like Qinyu (IQIYI brand). Which I think is supposed to show off their headset today and it will release in April.

 

[Fess]

Pretty cool.

Looking forward to quest 3. Hurry up and announce it already

Hopefully it will happen tomorrow at GDC, apparently they will talk about the future of the Quest ecosystem.

 

[CamHostage]

The fact they are making the full shift to Unreal Engine 5 most be to get a head start on the Quest 3, which is probably going to be the real beneficiary of Unreal Engine 5. While Quest 2 and Quest Pro (I assume) will support it, I expect it to be limited with much of the support for it being for Quest 3. It seems that they are aware they need to catch up for better quality and produced game titles if they want to still have a competitive product, along with keeping up with the industry and making it easier for devs who are also transitioning....
By doing this they will raise the minimum bar for game output ambition and quality.

So, have you read the Meta blog on the UE5? It's not all "Oh hell yeah, UE5 is awesome and now it's powering Quest!" news...

https://developer.oculus.com/blog/updates-developers-unreal-engine-5/

In reading your other posts, I don't think you're being unrealistic about the benefits (because there are benefits,) but just for those who only caught the headline and don't know what it means when a platform "supports" a complex development suite like Unreal Engine 5, here's some of the bulletpoints which may be relevant to gamers or gamers into development process...

• No Nanite on Quest. Epic has not brought Nanite to mobile-powered devices and although there's some external experimentation with that from other mobile device providers on next-gen mobile chipsets, Meta says "Nanite is not available on Meta Quest" (which I would take to mean the entire line, including the future Quest 3.)
• No Lumen on Quest. Epic says it has "no plans to develop a dynamic global illumination system for the mobile renderer" (which makes sense, given how heavy it is; Epic has said that even in high-end devices, "While VR can be supported, the high frame rates and resolutions required by VR make dynamic global illumination a poor fit."
• And then other little developer-oriented stuff, like...
  • OpenXR is the only VR API supported by Meta in UE5. (I believe Vulkan integrates with OpenXR though? Otherwise it's weird that Meta touts Vulkan as a great solution for UE projects in a fairly recent previous blog post.)
  • Late latching is not being ported from UE4 to UE5, which is worrisome since Oculus once called this a "free lunch" in "increasing GPU performance without increasing latency" and recommended in projects as recent as 2020, but maybe the need has passed?
  • UE5 removed "CPU occlusion culling", which can be worked around by developers with their own custom occlusion culling; otherwise the GPU occlusion culling default is heavier on the system, and will also have certain actions like passing through the geometry of a wall a frame late.
  • PhysX is being deprecated, but Chaos Physics works on mobile so that's probably fine unless a developer has an outdated project. (Some still like PhysX but Chaos has apparently made strides since release.)
  • With UE5 moving onto Nanite and Lumen as default yet not supporting low-end devices, hardware geometry tessellation and Unreal Lightmass automation are gone. (Developers can find their own way around those gaps.)

https://www.roadtovr.com/meta-quest-unreal-engine-5-support-lumen-nanite/

Particularly without Nanite and Lumen on Quest, a lot of the stars in people's eyes about UE5 fades away. There are still exciting new features (and handy developer functions) which should make for cooler games or other enhancements due to its more modern architecture, but for the people psyched that Quest is getting "one-better Unreal Engine", don't miss the fine print.

 

[Eddie-Griffin]

So, have you read the Meta blog on the UE5? It's not all "Oh hell yeah, UE5 is awesome and now it's powering Quest!" news...

https://developer.oculus.com/blog/updates-developers-unreal-engine-5/

In reading your other posts, I don't think you're being unrealistic about the benefits (because there are benefits,) but just for those who only caught the headline and don't know what it means when a platform "supports" a complex development suite like Unreal Engine 5, here's some of the bulletpoints which may be relevant to gamers or gamers into development process...

• No Nanite on Quest. Epic has not brought Nanite to mobile-powered devices and although there's some external experimentation with that from other mobile device providers on next-gen mobile chipsets, Meta says "Nanite is not available on Meta Quest" (which I would take to mean the entire line, including the future Quest 3.)
• No Lumen on Quest. Epic says it has "no plans to develop a dynamic global illumination system for the mobile renderer" (which makes sense, given how heavy it is; Epic has said that even in high-end devices, "While VR can be supported, the high frame rates and resolutions required by VR make dynamic global illumination a poor fit."
• And then other little developer-oriented stuff, like...
  • OpenXR is the only VR API supported by Meta in UE5. (I believe Vulkan integrates with OpenXR though? Otherwise it's weird that Meta touts Vulkan as a great solution for UE projects in a fairly recent previous blog post.)
  • Late latching is not being ported from UE4 to UE5, which is worrisome since Oculus once called this a "free lunch" in "increasing GPU performance without increasing latency" and recommended in projects as recent as 2020, but maybe the need has passed?
  • UE5 removed "CPU occlusion culling", which can be worked around by developers with their own custom occlusion culling; otherwise the GPU occlusion culling default is heavier on the system, and will also have certain actions like passing through the geometry of a wall a frame late.
  • PhysX is being deprecated, but Chaos Physics works on mobile so that's probably fine unless a developer has an outdated project. (Some still like PhysX but Chaos has apparently made strides since release.)
  • With UE5 moving onto Nanite and Lumen as default yet not supporting low-end devices, hardware geometry tessellation and Unreal Lightmass automation are gone. (Developers can find their own way around those gaps.)

https://www.roadtovr.com/meta-quest-unreal-engine-5-support-lumen-nanite/

Particularly without Nanite and Lumen on Quest, a lot of the stars in people's eyes about UE5 fades away. There are still exciting new features (and handy developer functions) which should make for cooler games or other enhancements due to its more modern architecture, but for the people psyched that Quest is getting "one-better Unreal Engine", don't miss the fine print.

Again, until we see the Quest 3, which is going to be the likely main beneficiary of the move not the Quest 2, we can't really say anything.

Though that is expected to change today according to insiders.

 

[ChuckeRearmed]

In 5 fps