performance I think we should, collectively, be somewhat embarassed.

[Henri Beauchamp]

Then without ALM:

Notice the much sharper picture (the sign, the lanterns, the leaves on the ground, etc).

True, you got less (spot) lights than with ALM on, but frankly, the difference is not as dramatic as what you pretend !!!

Edited November 3, 2022 by Henri Beauchamp

[Andred Darwin]

51 minutes ago, Henri Beauchamp said:

Then without ALM:

Notice the much sharper picture (the sign, the lanterns, etc).

True, you got less lights than with ALM on, but frankly, the difference is not as dramatic as what you pretend !!!

Dishonest? Pretend? Really? lol, we are better than that! You are also comparing the landing point, with a different viewer.... not the same area...  you know very well there are differences between viewers, machines, operating systems. Here is mine and anyone with the same settings and hardware will get very similar results :

CPU: 11th Gen Intel(R) Core(TM) i7-11700KF @ 3.60GHz (3599.99 MHz)
Memory: 65361 MB
OS Version: Microsoft Windows 10/11 64-bit (Build 22621.674)
Graphics Card Vendor: NVIDIA Corporation
Graphics Card: NVIDIA GeForce RTX 3060/PCIe/SSE2

Windows Graphics Driver Version: 31.0.15.1659
OpenGL Version: 4.6.0 NVIDIA 516.59

Window size: 1920x1009
Font Size Adjustment: 96pt
UI Scaling: 1
Draw distance: 128m
Bandwidth: 3000kbit/s
LOD factor: 4
Render quality: 5
Advanced Lighting Model: Enabled
Texture memory: 512MB
Disk cache: Max size 3993.6 MB (100.0% used)

SL Official viewer (Second Life Release 6.6.7.576223 (64bit)), Use Shared Environment enabled, teleported from one my locations, in fairness I do have a custom preset enabled,  that I use while seeing areas from the Destination Guide (I will reset all settings and post a new screenshot later on).... and did exactly what a regular user would do to test.... Turn off Advanced Lightning Model under graphics.... I don't think the average user cares or know about FXAA, or SMAA or whatever... The new improvements, ALM On, EEP,  rendering speed can and will help SL continue for many more years, staying stuck with the old will not (maybe short term) .... 

Edit: I just reset the settings on the viewer and here is a new screenshot from that same area with ALM off... also... that area seems to have a light changing script.... between blueish, white, green... so its not same as I posted before. 

ALM ON Above, Off below.

Edited November 3, 2022 by Andred Darwin
New Snapshot

[Istelathis]

ALM gives the environment more depth, at least in my opinion.  If I could leave it on all of the time, I would do so but the frame drop on my viewer is significant and I usually only use it for a short time.

In the Genesis viewer, without ALM:

This is great for exploring, due to the limitations of my laptop having a 1050 ti and an older CPU.  But it looks a bit flat, the scene is missing a feeling of depth.

 

With ALM:

Much better in my opinion, I feel drawn in by the light reflecting on the trees ahead , the shadows give the depth I was looking for, and the lights suspended above me really pop.  If only I had a newer PC to be able to wander around without a significant FPS drop.  

I need to download LL's viewer, and see all of the changes that have been made to it.  Regardless of ALM or not, the region is gorgeous.  

[Henri Beauchamp]

33 minutes ago, Andred Darwin said:

Dishonest? Pretend? Really?

Yes, because you choose a scene with many spot lights and then zoom out in a way that nobody would normally do, to encompass even more spot lights, meaning the difference between deferred and non-deferred modes becomes ”dramatic”.

What I am pointing out is that in normal outdoor scenes conditions, (e.g. when you move around in a sim or sail near coasts, or fly over land) ALM is more an annoyance than anything else; first and foremost it blurs details in textures themselves with LL's FXAA shader, and even with SMAA (which fixes this issue), it causes ugly artifacts in (sometimes not so) distant grillage meshes, rails, windows, etc. Then it causes a higher memory usage for textures (since it loads the ”normal” and ”specular” channels textures in excess of the ”diffuse” texture), which in turn can cause the viewer to use a degraded level of detail on the objects textures to fit them all in VRAM, causing even more blur (and the annoying alternating blur/un-blur effect on some objects).

Do not take me wrong: I am not against progress and would definitely love to see all the ALM shinies without all the blur and artifacts, but it is just not how things (currently) work, thus why I almost never use ALM myself while my computer renders it even faster than the forward render mode in most scenes.

Edited November 3, 2022 by Henri Beauchamp

[Andred Darwin]

19 minutes ago, Henri Beauchamp said:

Yes, because you choose a scene with many spot lights and then zoom out in a way that nobody would normally do, to encompass even more spot lights, meaning the difference between deferred and non-deferred modes becomes ”dramatic”.

You are right, I picked that area because it has many lights, I wanted to showcase ALM in action.... I'm not sure what you are referring as "normal" scene conditions, but most if not all areas in the Destination Guide are like that, be that night, be that sunshine, whatever, those are the areas users that like to explore, new users, visit the most (Flickr is a huge example of it, and yes, people that like landscaping as an example do zoom out... and a lot...may not be your personal experience though...neither needs to be, you do whatever you want ) my goal was to highlight that ALM is awesome and way more "interesting" compared to simply turned off and development of the viewer should not be held back because of old hardware (as I mention at another post, in my opinion should not be ignored either), the comment I hear the most from users coming back is how nice SL looks now with EEP, Mesh and ALM.... not FXAA or old computers running SL... The computer I took the screenshot is not by any means the fastest or the latest or expensive, its not old yet either, and I since the viewer update I have not seen a region that would drop it below 60 fps, which is awesome ( although I don't really visit any region with more than, lets say 10 avatars in view ).

Edited November 3, 2022 by Andred Darwin

[Henri Beauchamp]

18 minutes ago, Andred Darwin said:

my goal is to highlight that ALM is awesome and development of the viewer should not be held back because of old hardware

You should read again what I wrote about progress and regressions. I am in no way adverse to progress (this is actually the exact opposite, being an eager early adopter), but I cannot stand regressions and should some new shiny feature appear that got drawbacks, then I want to be left the choice to enable that feature or not, depending whether the said drawbacks are becoming unacceptable in some specific conditions or use cases.

I am not asking to hold back ALM and PBR changes, but just that the final user is left a choice, since the picture quality and the required computing power are both drawbacks that can impact badly the said users...

I want choice, and will never, ever accept being shoved negative things down my throat. That's why I made the Cool VL Viewer, by the way... 😛

Oh, and about:

Quote

I'm not sure what you are referring as ”normal” scene conditions,

Please, re-read: ”when you move around in a sim or sail near coasts, or fly over land”... That is, with a ”normal FOV”... Not zoomed out 200m away !

Edited November 3, 2022 by Henri Beauchamp

[Andred Darwin]

46 minutes ago, Istelathis said:

Much better in my opinion, I feel drawn in by the light reflecting on the trees ahead , the shadows give the depth I was looking for, and the lights suspended above me really pop.  If only I had a newer PC to be able to wander around without a significant FPS drop.  

That's exactly my point, and I hear the same from many others... I hope they will continue to improve the viewer ( the last update made it go, in my case from an average of 50 to 60 fps, to 130 to 200fps depending on the region on the same hardware! ), a newer pc can help depending on the configuration you choose, sometimes a small update on the current pc (assuming is not a laptop, like in your case... ) can make a huge difference (not only for SL) and not be, hopefully, too heavy on our pocket. I really like the official viewer (specially since the performance update), and that is the one I use when playing,  Henry's Cool Viewer is probably the fastest, using the v1 UI/Layout.... Firestorm seems to not be as fast for me compared to the other two, but the dev team is getting there (if not already... ), I just don't like too many options on all windows... feels clunky to me (even though many are very useful and handy....)

Edited November 3, 2022 by Andred Darwin

[Coffee Pancake]

On 11/2/2022 at 4:04 PM, Wulfie Reanimator said:

You could afford him a tiny bit of good faith, it doesn't ruin your own argument.

I don't mean to sound overly dismissive .. I have a lot of hardware that will "technically" run SL. It's not impressive or some kind of gotcha, none of it's a good experience in real world use and none of it should have any place in SL's future going forward. 

If it's over a decade old, it's time to put it out to pasture and move on, especially when it's actively holding the platform as a whole back.

Oh .. and suggesting that the use of point lights is somehow disingenuous is laughable.

[animats]

We need directional non-shadowing lights with range and cone angle. Then you could light interiors properly. Lights that throw shadows slow down rendering too much, and point lights go through walls and ceilings.  But large numbers of stationary non-shadowing lights can  be handled.

Edited November 4, 2022 by animats

[Coffee Pancake]

We need a complete overhaul of the lighting model. 

Lighting is what lets SL down graphically, we don't have the tools to do it well and people rarely use what we do have because it defaults to a having a day star in your house.

[animats]

Here's my personal list of viewer-side improvements. These are things I've done in my experimental Rust viewer, and many of them are now appearing in the LL viewers in C++.

• Load textures at roughly one texture pixel per screen pixel. So, if someone has a 1K texture on a small object, it never gets loaded unless you get really close and zoom in. LL now has it in the performance viewer.
• Physically based rendering. LL now just released the first PBR project viewer today.
• Switching to Vulkan. At Creator User Group, it was announced today that Vulkan is on the schedule for 2023.

It's encouraging to see LL moving forward.

Stuff I've done experimentally but LL hasn't done yet:

• Fix texture loading order. Load in order of screen area covered by the texture. This fixes that constant annoyance where you want to read a sign or a vendor or get a close-up look at something and you want a minute or two for loading. The viewer has to reorder the loading queue on the fly as the viewpoint moves. The bare beginnings of this showed up in a Linden project viewer, but it was never finished.

All that's in my demo videos. Much more is possible.

 

[Henri Beauchamp]

2 hours ago, animats said:

Physically based rendering. LL now just released the first PBR project viewer today.

And from the release notes:

Quote

Known issues:

• Degraded performance
• Degraded stability
• .../...
• Occlusion culling does not work

Just one word: OUCH !

I am afraid that, despite my warning ( ”I do really hope the same mistake will not be done with the future PBR viewer, i.e. do take your time” ), LL will be trying to push PBR too soon to ”release” status, with the same ill effects seen as when they pushed EEP while its renderer was still full of bugs that only got solved with the performances viewer, months later... 🙄

Dave Parks replied to my worry with:

Quote

The current plan is to get the PBR viewer into a public alpha and gather more feedback, and spend the next six weeks optimizing, fixing regressions, and responding to feedback.

Seeing as they release a ”public alpha” now already, I have some doubts that only six more weeks will be enough...

*sighs*

[Andred Darwin]

5 hours ago, animats said:

Switching to Vulkan. At Creator User Group, it was announced today that Vulkan is on the schedule for 2023.

That's great to hear! Thank you for sharing, also sharing  the release of the new alpha viewer!

I played with the new alpha viewer this morning on the main grid, not to test PBR (not available at the main grid yet), just out of curiosity to see how it renders (considering known issues, water, etc..) compared to the current one, what called my attention on the notes was "Improved VRAM Utilization".

From the release notes: 

Quote

Improved VRAM Utilization
In order to better facilitate the increased number of textures associated with GLTF PBR materials, improvements have been made to the viewer texture streaming and VRAM utilization. Textures should load more quickly and should downres less often, and Second Life should make full use of available VRAM.

This is really noticeable compared to the current official viewer, like water to wine, lol, even after going back and forth to quite a few regions and let it load completely (cache full)! 

Rendering seems darker than the current viewer.
Looking at FPS, it takes a hit even without the new features in-world, average of 10% to 18% (Main grid, no reflection probes, pbr materials, etc), probably due to Occlusion culling either being disable or just not working (notes mention it does not work in this version), however it feels smother.

I thought it would be really slow on an event area with occlusion disabled, but didn't seem at all... the perception of not seeing "texture downres", and objects fully “rezzed” while turning or "going up a stairs between floors" was a much nicer experience than the current behavior (I'm sure that perception will vary between machines…).

Even taking a hit on performance, I wonder if there is an option to set the distance for occlusion culling take effect on the current official viewer? 
Example: Draw distance set at 256m, but no object occlusion within 64m (or 128m)

What is sort of a bummer, is the LOD factor, set and reset to 1.125 every time you switch between viewers (alpha and official) or reinstall... the "thing" sets itself to Ultra, and LOD at 1.125? 

Sincerely, I think that is one of the main reasons, a lot of people just "ignore" the official viewer ( That first impression... ), the default could be bumped a bit (2.0 like Firestorm?) and the maximum allowed at the graphics preferences window could also be bumped (Maybe 4.0, Ultra, lol).

 

 

Edited November 4, 2022 by Andred Darwin

[Henri Beauchamp]

59 minutes ago, Andred Darwin said:

Improved VRAM Utilization

This has been the case, for years, in TPVs, which all increased the VRAM usage beyond the 512 MB LL stubbornly kept as a limit in their viewers for years...

[Lyssa Greymoon]

8 minutes ago, Henri Beauchamp said:

This has been the case, for years, in TPVs, which all increased the VRAM usage beyond the 512 MB LL stubbornly kept as a limit in their viewers for years...

To placate people who demand SL run on e-waste.

[Andred Darwin]

15 minutes ago, Henri Beauchamp said:

This has been the case, for years, in TPVs, which all increased the VRAM usage beyond the 512 MB LL stubbornly kept as a limit in their viewers for years...

I know, I'm just glad it changed, I use the official viewer day to day and "texture downres" was is quite annoying (probably necessary... but still annoying) !

It doesn't seem to fully use memory available (or maybe it does, and I am confused about what VRAM really is, or the amount that is available to a program... or the info below is just the current amount used... ), the graphics card I use has 12GB of memory, and this is what comes up at the "About Second Life" window:

Quote

 

CPU: 11th Gen Intel(R) Core(TM) i7-11700KF @ 3.60GHz (3599.99 MHz)
Memory: 65361 MB
OS Version: Microsoft Windows 10/11 64-bit (Build 22621.674)
Graphics Card Vendor: NVIDIA Corporation
Graphics Card: NVIDIA GeForce RTX 3060/PCIe/SSE2

Windows Graphics Driver Version: 31.0.15.1659
OpenGL Version: 4.6.0 NVIDIA 516.59

Window size: 1920x1009
Font Size Adjustment: 96pt
UI Scaling: 1
Draw distance: 128m
Bandwidth: 3000kbit/s
LOD factor: 4
Render quality: 5
Advanced Lighting Model: Enabled
Texture memory: 4095MB
Disk cache: Max size 3993.6 MB (22.6% used)

 

Texture memory changed from 512mb to 4095MB.

Hopefully, they will also change the default LOD when set to Ultra! 

 

Edited November 4, 2022 by Andred Darwin

[Andred Darwin]

13 minutes ago, Lyssa Greymoon said:

To placate people who demand SL run on e-waste.

I guess... pretty much! 😛

Take a look at the comments/exchange between Henry and Dave at https://bitbucket.org/lindenlab/viewer/commits/aaf7b17db047f0cb2630b479d5468062e6ca815e#comment-12384200

Of course Henry has a point, and to an extent, it's nice that he puts it forward (also very knowledgably), at the same time it's nice to see LL is going after, not only improving the viewer meaningfully (like the performance improvement)  but quite fast!  (TPV's will continue to improve it and maybe even better!)

In my opinion Forward Renderer should be "kept" on its own viewer fork (like it happened in past many times... v1, Snowglobe, v2), and receive improvements when needed, for a period of time, but removed from the current development... so the current work can move forward easily... opportunities to improve or mistakes will happen for sure, but that can be worked out... 

[Wulfie Reanimator]

4 hours ago, Andred Darwin said:

Take a look at the comments/exchange between Henry and Dave at https://bitbucket.org/lindenlab/viewer/commits/aaf7b17db047f0cb2630b479d5468062e6ca815e#comment-12384200

Thanks for sharing that discussion.

[Henri Beauchamp]

4 hours ago, Andred Darwin said:

In my opinion Forward Renderer should be ”kept” on its own viewer fork (like it happened in past many times... v1, Snowglobe, v2), and receive improvements when needed, for a period of time, but removed from the current development...

This makes no sense to maintain two different branches of their release viewer for LL.

And no, it's not just about old computers compatibility, but also about most modern (non gaming) notebook computers (that only got an iGPU or APU to render, not a discrete GPU, and are therefore no better than a GTX460, or even weaker), as well as future mobile support (think ARM-based devices, Chrome-books, etc).

The cost for simply freezing the forward renderer as it is and keeping it along the evolving PBR renderer is close to zero: just a matter of keeping some shaders and adding some branching in the code to cope for C++-level differences. I know first hand, since it is exactly what I did for the v1.28.2 branch of the Cool VL Viewer which had both the WL and EE renderers (and switchable on the fly, by just checking a box in the graphics prefs too) !

Edited November 4, 2022 by Henri Beauchamp

[Henri Beauchamp]

6 minutes ago, Wulfie Reanimator said:

Thanks for sharing that discussion.

I shared it already in several links that you will find in my above posts, in this very thread...

[animats]

OpenGL version usage stats, from LL, from the commit log. (Open image in bigger window to read.)

"The Y-axis here is unique account id, the x-axis is GL version by month, going back to January 2018. So the first cluster of 4 bars shows the counts of GL major verison 1-4 for as they were in January 2018, last 4 bars are September 2022. Note that this image is cropped to not show the trend in overall agent numbers (there’s a place to talk about that trend, and this isn’t it)… the ‘4' count is actually higher than shown in the image."

Simplified version

Based on this, Linden Lab wants to drop support on OpenGL versions 1 and 2, and only support 3 and 4.

[Wulfie Reanimator]

38 minutes ago, Henri Beauchamp said:

I shared it already in several links that you will find in my above posts, in this very thread...

That you did!

I hadn't checked many/any external links during the week since I've been reading this thread on my phone during commutes.

[Andred Darwin]

1 hour ago, Henri Beauchamp said:

This makes no sense to maintain two different branches of their release viewer for LL.

And no, it's not just about old computers compatibility, but also about most (non gaming) notebook computers (that only got an iGPU or APU to render, not a discrete GPU, and are therefore no better than a GTX460, or even weaker), as well as future mobile support (think ARM-based devices, Chrome-books, etc).

The cost for simply freezing the forward renderer as it is and keeping it along the evolving PBR renderer is close to zero: just a matter of keeping some shaders and adding some branching in the code to cope for C++-level differences. I know first hand, since it is exactly what I did for the v1.28.2 branch of the Cool VL Viewer which had both the WL and EE renderer (and switchable on the fly, by just checking a box in the graphics prefs too) !

I respect what you say.

Sincerely,  "iGPU or APU to render, not a discrete GPU,  ARM-based devices, Chrome-books" really? Maybe a super expensive Jetson TX2, AGX and alike (to that level) would have a chance to run SL with all it takes in way that is actually playable, and that in a very distant future! iGPU's, APU's, Chromebooks all have its pros, but in any of them SL fits... sorry, could get better, but is never going to work! I understand that sometimes people will use what they can or what they have, but for SL, in this case, it's the same as to "like to feel pain".

It was the official dev team that came up with that idea to remove Forward Rendering and concentrate on a improved Deferred Rendered, and Dave's arguments make a lot more sense to me, I also believe they will do a great job based on what they were able to achieve with the latest performance improvement... there is also the ROI standpoint.... 

I think its the same as you keeping the v1 UI on the Cool Viewer, its cool to see and remember the old school SL but many can't stand it (its ugly to them compared to the current UI on Official, Firestorm, Black Dragon, etc...), and sincerely, some only use it cause you did a great job keeping it up, otherwise, it would only come up as "memories" from old times, and just because you did, does not mean a company looking for profits should, you very talented, you like SL, you like the Cool Viewer and have a lot of time to keep up with the Cool Viewer (and kindly shares it for free!)

Actual state of Forward Rendering and continue to invest on it,  in my opinion, is the same as keep spending money on that old car that breaks all the time! Deferred Rendering is the new "car" in this case (and seems in 2023 powered by Vulkan "motor" lol )... looks great, much better! And based on Dave's arguments, the old car, probably should be "retired"... 

The thought of having 2 viewers, a light one specifically tailored for iGPU's and APU's would allow users of those machines continue to use what they are used to, also allow passionate and talented people like yourself to continue to improve it with improvements specific to those machines... (and maybe as a base for ARM, etc... which is the only segment that still uses Forward Rendering, Vulkan is gaining ground...)... the same would happen with the discrete gpu's, actual gaming pcs... as of cost... if 30% of SL users need it,  that cost would also be close to nothing for LL, many companies do that, keep one or two developers taking care of the current version, while the team works on the next big thing.. the cost of not moving forward will be much higher... competition, although pretty far away, will get there pretty soon, SL is the best in all metaverse as a whole today... and will continue to be the best with the improvements pursued by Dave and team!

Edited November 4, 2022 by Andred Darwin

[Henri Beauchamp]

1 hour ago, animats said:

Based on this, Linden Lab wants to drop support on OpenGL versions 1 and 2, and only support 3 and 4.

This got nothing to do with preserving the forward renderer, even though the latter would allow an easy path towards ARM and mobile support thanks to the fact it does run with OpenGL v2.1. Reread this.

Note also that OpenGL v1.x and 2.0 are not supported any more, even in current viewers: it has been years their support got dropped. OpenGL v2.1, on the other hand, is what most ARM64 platforms support (including Chromebooks)... Preserving the compatibility with it would be an opportunity to get a real viewer (and a free viewer too, like in free beer) running ”soon” on mobiles, and even perhaps under Android...

AGAIN, the forward renderer advantages over ALM are:

• Easy on GPUs: it can work decently on any modern computer, including notebooks without a discrete GPU. This keeps the entry level for SL as low as possible, and prevents people without a gaming computer to give up on it on the first login.
• Better, even on modern gaming computers, for moving around SL (flying, boating, driving) across sims, with a waaaaaay better AA quality, less memory used (important when you need to push the draw distance to 512m, while flying or sailing, for example), most often better FPS rates (yep, even with a modern GPU, in these circumstances, i.e. with a large DD and few close avatars around), faster textures rezzing (less textures to download while moving around: super-important, when you move fast !).
• Lower power consumption (= larger autonomy for notebooks), lower heating (= no throttling on cooling challenged notebooks, e.g. Macs M1/2), lower fans speed (= much less noise).

Edited November 4, 2022 by Henri Beauchamp

[animats]

The real problem isn't old computers. It's new, low-end ones.  This is what US$89 buys today.

Gateway 11.6" Ultra Slim Notebook, HD, Intel® Celeron®, Dual Core, 64GB Storage, 4GB RAM, Mini HDMI, 1.0MP Webcam, Windows 10 S.

This is the era of the $1000 phone and the $100 laptop.

(That thing runs Windows 10S, the locked-down version that can only run programs from the Microsoft Store. Which is why the computer is so cheap. Second Life can't be in the Microsoft Store without routing all payments through Microsoft, which takes a 15% cut. Unlocked computers start around $136. What's the lowest-end laptop in current production that runs SL at all?)