This!! I panicked when I first read that as if I DC it’s so handy having my world camera in the same place so I don’t have to fidget with it again mid set. Smoother experience for viewers too.
I’ve used this dirty hack for 4500+ hours and in that time I’ve never come across a world or shader that was affected by it. It was perfectly functional and practical, even if it’s not a clean implementation.
well trying to make sure performance FPS is running so smooth in running game possible to fix it somehow.
While I can understand why the devs are hesitant about FSR, I still think implementing it as say, an advanced option (with a pop-up warning that it could potentially break worlds) or as a command line option would still be a good idea. Yes, there are issues, but it doesn’t change the fact that many people used it before the EAC update and found it useful enough to offset the issues that it does have.
It seems to me like an imperfect, optional solution is better than refusing to add any solution whatsoever. The community doesn’t share the view that a solution is only worth implementing if its a one-size-fits-all. And when so many people who have played the game with FSR are telling ya’ll that they still want FSR despite the issues, its baffling that ya’ll seem to be leaning towards ignoring those voices.
The justification has been stated and that’s how it is. From a technical perspective, their reasoning makes perfect sense, even as someone who previously relied on it. I can’t really say your anger is justified. That being said, usually the reason people rely on this so much is due to a GPU bottleneck. Try disabling Hardware Accelerated GPU Scheduling (HAGS) in your Windows Graphics settings to remove some overhead, that was single-handedly what made playing the game without FSR possible for me. If this isn’t your solution, then it really is time to consider an upgrade.
Waiting for opportunities:
- Add button “Disable post-processing”.
This may cause nausea or pain in the eyes.
There may be problems with this, so I propose to prohibit disabling postprocessing for game worlds by default (“game” tag), as well as enabling the author of the world through the button on the website to prohibit disabling post-processing, let’s say just adding to the world tag’s “do not disable post-processing” and vrchat menu, after reading this tag, will not allow it to be turned off.
If you put some relax shader on top of it, it will only lead to an additional drop in performance and as a consequence to nausea. - The solution to the “avatar stats” problem.
Loading “very poor” avatar causes freeze’s in the game of checking the avatar’s performance statistics on the client side.
Button to disable processing of this data would be very useful for stability. - Add the ability of Avatar haptics to work with OSC, to output information, if possible.
- Fullbody mirror
Is the same as “face mirror” but in full growth (2d projection on “hud”), yes I know that soon we will be able to attach “personal mirror” to the review, but in that case it might to be block by extraneous objects or avatars, the option with a 2d projection on “hud” is the most convenient, perhaps as part of “face mirror” setup, which in this case will be called “hud mirror”.
I have been using FSR and what it makes a difference is that it sharpens the edge of avatars, making it MUCH MORE beautiful from a distance. This always happen when you are in a group of over 4 people, standing in front of the mirror.
I access VRChat with 3090, through Oculus Air link and SteamVR. FSR improves framerate, making avatars more beautiful and brings bad effects on world environment objects, like rocks. Despite the cons, I will be willing to have FSR on. Please reconsider the implement of FSR!
Glad to see you are examining an alternative from NVidia, if it could sharpen avatars, FSR is no use, though.
Believe me I tried to push them to offer a toggle for post processing but too many think there are ‘better’ ways to solve it so it’s not being added. They may be right or the provided options may make things worse. We will see
If the options they create don’t help me in the long run I’m going to be a lot more aggressive in pushing for it
Sharpening the FSR is important even if scaling is not available.
isnt there any talk on offering a more privacy focused anti cheat? I certainly don’t trust eac to scan through all my files and send the data to god knows who. Could you not develop your own form of anti cheat? I know bloons td6 has a light anti cheat that detects dll files in the games install folder that shouldn’t be there, and doesn’t require root access to do.
I did not get a chance to comment on the features shown off in the Aug 23 dev update, but this is actually related to that- will the new Main Menu/World sorting/categories screen allow us to organize our favorites?
Specifically, I added a bunch of worlds to my different Favorites folders when I first started using VRChat. Now I would like to reorganize them by moving them between folders (for example, moving worlds that were in my Cool Stuff folder into my Hangouts folder and Games folder).
As far as I can tell, this functionality does not exist in VRChat at all at the moment, unless I manually unfavorite/refavorite each world. It’d be nice to be able to drag-and-drop worlds into different Favorite folders. Even if it’s something that has to be done through the webpage UI, it’d be better than nothing.
Hey, nobody is perfect, ah ha… Usually I’m more thorough before blindingly filing feedback. I just made an assumption this once (and look who assumed wrong!)
And I oop.
Why not natively implement AMD FidelityFX Contrast Adaptive Sharpening (CAS). this is postprocess and has immense benefit for visual clarity, especially in VR. If implemented, it should be configurable (contrast adaption and the sharpening amount). It doesn’t improve performance, but it does improve visual clarity to a notable degree.
(also please fix the embed details for the canny so it shows the post title and such)
Better call saul
Good to hear! Always happy to know about coming performance updates for avatars - there’s a lot to expand upon, especially to help creators make more performant avatars!
Will there be a point that Poiyomi will be compatible with Quest? Or even just some kind of update to the avatar SDK for quest?
The problem is that a significant performance impact factor is shaders (ie poiyomi, which used to be the bane of my existence on an RX580), and they cannot measure their performance realistically without sandboxing, and even then that will measure render time already under load from the game itself.
Otherwise, there’s a lot of misinforming values, like how stingy it is on polygons, and the pointless inclusion of bounds as somehow being “non-performant” (most of the time larger bounds are only for a specific animation or utility, nothing to do with the rendering of the skinned meshes). It also fails to isolate realtime pixel lights from vertex lights (vertex lights don’t cost anything because they’re always being calculated even if none are present).
I’d personally prefer this system just be removed entirely, or at least the “rating” value removed, leaving only raw data for people to make their own decisions on how bad it is based on their setups and experience. And ofc the whole system have a toggle to disable because it causes lag on load.
I understand that quest doesn’t gets much love but when will we get a performance update just a little love for the quest vr users?