🪰 Viewer Bug Reports

I've become aware of a texture rendering issue that impacts textures when on a face. It's not so much an issue that only impacts PBR textures, but it is exponentially worse and makes certain designs as mastered in Substance unusable as PBR. I'm not enough of a rendering brain to understand what is going on here, but it's my belief this is not correct and is a bug. I hope that this will get in front of the right people to be evaluated and hopefully fixed. This is being demonstrated with a 0-25% gray bi-linear gradient, however anything that uses color and gradients near this will exhibit the same obvious defects. A single example using real-world art is included for reference. Texture UUIDs Dark gradient lossy: cac83520-08b2-500f-b683-1ede13d4e052 Dark gradient lossless: bd66f42c-6b56-780e-c183-3307151f8292 Raw files, non-anotated screenshots, and sample items shown are available upon request but I don't think they will be necessary. As I was sharing this around with creator friends moments before submitting this I was made aware this may be or have been a known bug, and can be alleviated entirely on the material side by setting the alpha mode selector to blended. Masked and opaque do still exhibit the shown issues. This made me frown. There is no alpha channel on the textures shown.

I am not sure if this is the right place to post this so if it's not please somebody recommend a more specific form recently we had both of our regions converted over to web RTC. We did this understanding that this is the future and we might as well get used to it now and avoid the rush when the world converts. However there are a few things that I have noticed that are concerning. There are times when you're in world and ones voice can sound quite robotic almost like it's dropping packets. The temporary fix seems to be unclick your voice key scroll way out from the distance hit the escape key to return to normal view and then re toggle on your voice. I have no idea why this works but it seems to fix it temporarily. Web RTC also seems to use a very different calibration of voice levels whereas when you can hardly hear somebody their voice brackets are in bright red in in a regular SIM would be over modulated. I am not sure are these things that LL knows about or who to talk to regarding these concerns. The individual individual volume control for other specific users do not function. The Voice waves are not correct

I've been testing the 2025.03 beta on different machines, and on this particular machine equipped with a Ryzen cpu, a 3080m gpu and 32gb of ram, i observed intense frame inconsistency, that gets worse the more i wander around. It seems worse when getting inside and outside of a pbr probe. A side note: I also played with the max number of probes slider, i observed that if set to lower values, the rendering is completely messed up, looks a disco psychedelic show :) So i set this value back to 256. Note 1 : After i observed this frametime inconsistency, i hard reset the laptop internal power settings, i reset the bios, i re-installed the nvidia driver (this time i used the studio driver ..) and the problem persists. This doesn't occur on latest FS beta, and didn't occur in the previous SLV version. Note 2 : I also tested the viewer on an older windows laptop with a 1660ti gpu (Ryzen cpu, 16gb of ram), and on this one, the viewer does marvels. It's perfectly smooth and performance is impressive, as it's actually much better than the last time i tested it on the previous viewer. So i'm not sure what's going on, but i thought i would share :)

If I begin with a fresh Avatar floater, fresh from a login. I filtered for "Tori" and got a complete list of the outfits with "Tori" in the name or contents: https://gyazo.com/5e829bdd1789d74865d31b83f48845b6 If I clear the textbox using the X tool at the right end of the box and type in "!L", the result will be the complete list of the 100 or so outfits including those characters in the name (or contents). If I then type over the text in the now yellow textbox (not clearing first with the X tool), typing "Tori" I get at subset of the complete list. A list containing only those items in the previously filtered list. https://gyazo.com/64c893f0eb2f1c56918d852b039f3226 This is an example of a case that is easily reproduced. However there are numerous cases I cannot explain. For example when starting with a new filter after clearing the textbox and not getting the complete list of known instances. This has also been occurring for months in previous releases of the SL viewer.

I see my dot - so I am connecting  too the server or it's a false connection ? I see no sound waves for myself - other avatars are in the RED; they can not hear me and I can't hear them. Was told to wait for a region restart-  I did nothing changed. Was sent the Wiki on sl voice trouble shooting - even though it did not apply too me, its trouble shoots the Vivox  system .     BUT humor them anyway and went through each step 1. by 1. too no avail Things I have tried on my own and with help from Tech friends Both Firestorm Viewer , current and Beta  The Current SLV viewer  tried them all . White listed the recommend files  in Windows Defender installed a different Anti virus , no change  ( Avast - and the firestorm is white listed in there) reinstalled all viewers-  from scratch , deleted everything off my pc! ( a long shot)  did not work !  i didn't think it would :) Tried other accounts on my PC  same as my main . Tried 2 other pc's in the house and worked fine.  so its not blocked in my router/firewall was told it used Port 443? then is open , if that is the port its using and was blocked why would it work on other pcs? I restarted the router and unplugged it and restarted pc ... I have been to the testing regions but it doesn't work.   the old voice system does work just fine its just the WebRTC that is the problem I was thinking it was My Pc where the problem is - but having talked to support again this am.  seems too think it may not be? I  have sent my logs to support . Which is where I am at now. submitting this to form as requested by support

Location to reproduce: http://maps.secondlife.com/secondlife/Han%20Loso/72/190/59 Steps to reproduce: Touch 3-4 perimeter buttons on the circular dialing control in front of the gate. Observe the illuminated chevrons on the gate ring itself This bug seems to present on fullbright surfaces using the legacy bumpiness options. I've isolated the component of the object that demonstrates this bug and attempted setting it to bumpiness:none and back to bumpiness:bark as it should have, and it looks much better. Changing my selection away from the item however makes it revert to the broken state. Items should not change how they look based on if they are selected or not, and legacy content should not look this different, the right example in my isolated example screenshot is completely acceptable if it can be achieved without it resetting on de-selection. Additional wacky aspect of this comes in when scaling the item: https://youtu.be/qB7tVtKFMOQ This scaling issue does not at-all repro on FS 6.x

When editing an object, going to the features tab, and selecting a physics preset (Stone, Wood, Etc.) it changes it accordingly. When you open different objects that have linked parts, it shows the last preset you selected in the grayed-out field. When you edit linked, select the first link, it will then show you the true existing preset selected for that object. Whatever you selected in a previous edit always shows for all linked objects until you change it. If you build, for example, a cube, but delete that cube before you close the build menu, that preset field will be blank for all linked objects until you edit any object and change it. When you edit linked and go to the first link, it now sets whatever was in that preset field to that instead of the blank. Relogging does not reset this behavior. It appears to save that information in the system because I have logged into the SL Viewer and the Firestorm viewer and determined that state persists. I have attached screenshots from both viewers showing the same thing. This is real problem for landscapers because when I am changing those physics settings for many different linked objects, for each linked object I have to edit linked, go to the first link to see if I have already changed it or not. It should show the first link's existing setting, not the last one used to change something else.

Issue: An item rezzed by a rezzer, or rezzed by dragging out of a prim, can be grouped and coalesced back into the object, incorrectly updating the last-selected object as a coalesced one. It reasons that each selected which were rezzed separately be placed back in the object separately. Now that we can pick up objects as a group but still separated, the Build > Object > Save Object back to Object Contents should respect the form it was rezzed in and put it back as singles of the same name as well. Why? Right now we can make really simple rezzers that just let us organize stuff willy nilly but to save updates made (like recoloring, relinking, etc) we have to Save Back to Object Contents, one at a time. If we could do it as a group, we could select a whole build and choose the option, any anything which has an object_rezzer key still around can be updated in one stroke. This would be phenomenal for keeping builds organized and updated! Note: This may create issues with mixed coalesced and linked items. In that case, it should be prudent to take the coalesced items first, and then go through the rest of the selected list. If this is unfeasible, a warning is in order if saving a previously coalesced set back. Reproduction: Rez 3 prims, name 2 of them uniquely Pick the named ones up separately Place them in the object inventory Drag them out of the inventory to rez them, one at a time Select both of them and Build > Object > Save back to object contents Attempt to drag each out again. The last selected one before saving is now coalesced as both, and the other is not updated.

In the latest build, 7.1.12.13550888671 (64bit), enabling FXAA or SMAA divides the performance by 2 on my macbook Pro M2 (it used to have a negligible impact).

The effect of this seems to be easiest to reproduce by having a larger inventory, mine is 155K+. For me it reproduces on two different computers, running either Firestorm or the SL Viewer. Log in, open an inventory window (Ctrl + I) Open Windows' Task Manager, monitor your RAM usage Open a bunch of extra inventory windows, a few at a time (5, 10, 15, 20) (Ctrl + Shift + I) See how RAM usage increase, and if you just let it be, the RAM usage continues to go up for some time... I don't know how common it is to have a bunch of inventory windows open at once, but it would seem to me that there might be room for optimization here, with all the windows not having unique data but are views into the same data I think that it shouldn't be duplicating the data in RAM for each window like it seems to do???