Adding to Cart…
Licensing Agreement | Terms of Service | Privacy Policy | EULA
© 2024 Daz Productions Inc. All Rights Reserved.You currently have no notifications.
Licensing Agreement | Terms of Service | Privacy Policy | EULA
© 2024 Daz Productions Inc. All Rights Reserved.
Comments
LOL.
There's actually quite a few things I like about DS. Like alot in fact. Never mind I have no time to start dealing with migrating my product library. Which is a reason why Octane plugin for DS is no go for me. Complicated software having a step back is just a thing once in a while. Dependency on Nvidia and their slow IRAY development (how long has RTX been out?) does not help matters. Having DS 4.12.0.86 as a fallback works fine for me.
Still to do: check out Unreal vs Unity for realtime feedback on animation. Scrubbing timeline without iray preview is incomplete feedback. With preview is too slow.
Have the alembic plugin to keep all the morph effects. Will see how well it works with UE & time costs. Unity alembic plugin sounds more a WIP yet.
Also maybe Maya - Daz plugin coming. That was being worked on. Not sure of status now.
Daz developers did not remove the option, Nvidia did. The enable/disable checkbox in Daz Studio no longer affected the Iray render, other than causing errors in the log file. You need the newer drivers because of the version of Iray in Daz Studio. You could roll back to an earlier driver, but only if you also went back to an earlier version of Daz Studio. Daz Studio 4.12 will not render with the older drivers. In my case, it wouldn't even drop back to CPU. The resulting "render" was a blank, transparent, image.
3D-GHDesign! Throw out your 980 Ti GPUs and buy an RTX 2060! (you may not buy the latest GPUs, because these items get very hot). Install the newest driver from Nvidia and and forget about Scene Optimizer, OptiX and its checkbox. NVidia has already thought of everything for us. Just render and enjoy the speed. RTX 2060 are not that expensive, but it’s quite a pleasure. RTX is not a NVidia’s marketing move, this is an inevitable future of the 3D.Thanks to DAZ for the timely transition to this technology!
I noticed that, during animations, copy-pasting tcb keys brakes the last tcb key before the pasted one.
- Sometimes the values of the effected bones/properties become float.NaN, which causes the viewport to go "all gray" during that broken key period of the animation. Going forward-back in timeline *might* cause an infinite loop somewhere and causes daz studio to freeze.
- Sometimes they go the most extreme values (limits) of the bone/property. During the period of broken frame, bone/property sticks to the extreme value and when broken key is finished, it returns to the normal.
Workaround I found to the both of the cases are to select affected keys, make them linear keyframes, then make them back to tcb again (for triggering re-calculation). *Sometimes* this also doesn't work (for the first case above) and it always calculates as broken, so I have to leave those keyframes at linear.
I think this was happening until the last beta before 4.12, but these were fixed in 4.12. With the latest beta, they are ALL triggered again and broken.
4.12.1.16 is a very bad release.
Look at the process that eat up all resources and even stay running in the system after you close Daze3D. The process exit after a long time.
The process dont have a name but use the same symbol as Daze.
Rendering suffer, working with Daz3D suffer from it.
Sorry, but I'd rather go AMD.
My main bugbear with Daz is it's render solution locks me into Nvidia for relatively fast renders. My CPU does actually do a decent job, but even so, my 980ti is about 3 time faster than my CPU; of course, I never have it drop FROM cpu. :)
I've been messing around with transfers to Blender, it is just time-consuming, sadly.
@fableman, I've noticed, using GPU-Z, that closing Daz Studio doesn't release all of the memory on the video card. Using Task Manager, I've been able to locate Daz Studio where it's migrated from Applications to Processes. (Or maybe Services. I don't have The Beast running to check it. That's Windows 10. This computer is Win7 and the Task Manager is somewhat different.) While it's holding onto about 2GB on the video card, it's also holding onto about 6-7 GB of System RAM. This is based on my last image, Aftermath. The numbers may be different with other scenes, I haven't had a chance to check it out yet.
Anyway, I'm seeing this behavior with 4.11, as well as with 4.12.
This is lonmg-standing behaviour for me, it will - after a hard-working session - take quite a while to free up rsources, using one thread (CPU ~13% on my system). It shuts down quickly if it hasn't doen much, or loaded a figure like one of the Genesises with lots of morphs, but it does finish after a bit. If it's not clearing it may be an issue related to security software or soem other system-level tool.
Just thought I'd try 4.12 (Beta) again. First I ran an animated scene image series render and, true to form, it fell back to CPU on the second frame (GPU-Z showed a slight increase in VRAM usage but not enough to push the limits). Then I loaded a simple scene, just the base G8F and tried to adjust the frame rate settings on the timeline. DAZ Studio froze with the spinning blue circle - had to crash out. Loaded again, tried the same process, same result - spinning blue circle when trying to alter the frame rate settings (I was going for 24 FPS, 14 frames total, frames 0-13).
4.12 must be the most bug-ridden release yet. 4.11 does both of those scenes without complaint although rendering a single frame is slower and the viewport is much slower than 4.12. So I create scenes in 4.12 beta and, if there is an animation, I load them in 4.11 to render the image series.
Nvidia driver version: 431.60 for GTX 1070 8GB GPU.
I never noticed it before, but my last two images, (linked in my sig if any one wants to check them out,) really pushed Daz Studio and my GTX 1080 to the limit. I had geoshells with snow shaders up the wazoo in Heart of Winter, and Aftermath has 9 figures, if you count Dog 8, and six of those are in the foreground. Instancing, lots of reflective surfaces, and a large volume of water didn't help, either.
I've just been using the Task Manager to End Task. Is that a bad idea?
I work on ssd and never experienced a long delay freeing resources. But I also usually work with small projects, at least on daz studio. What I guess is it's freeing the swap file, that's the commit size in the task manager, this may get fairly large especially when using lots of 4K textures.
As for ending it with the task manager I guess it's a bad idea. A system restart should get it fine though.
The framerate change freeze happens to me too. It's pretty randomic, though, some times it works fine some other it freezes as soon I hit the Return key. I guess it depends on the scene content, somehow.
Also I noticed that the timeline doesn't manage aliases, even with all the filters (those TRSOAH options) activated if I select and delete/move a keyframe all the Aliases are left back. Every time I have to go to the Parameters tab and manage my aliases from there.
It was ok for me before this release, now it is kinda terrible when you exit the application, Moving around in scene feels alot slower.
Render is using the GFX cards CUDA
CPU at 30%
I use latest Nvidia studio driver and optix and got a 2080 card.
But I still able to play computer games same time I render aslong the render stick to CUDA and leave my CPU alone. :)
Just terrible bad when exiting Daze3D.
Since installing Daz3d beta v4.12.1.16. I too have experianced falling back to CPU rendering in the viewport even with a very simple scene
My system specs: Intel I9 9900K@5GHz, Nvidia RTX2080ti, 32Gb RAM, Windows10(fully updated) Nvidia STUDIO drivers v431.86
I allow Install Manager to retain all files downloaded, but the previous beta is not in those files, I'm assuming IM overwrites the older beta files with the latest
Can anybody provide advice on how I obtain the previous 'working' beta version please ?
Force quitting will stop DS from saving any changes to the UI and so on.
Yes, you would need to make a back up of the old installers before downloading the new.
All i can suggest is a support ticket
Why would you do that? That was one of the things I liked the most about my workflow, being able to work on multiple scenes at once. it says (by default) so how do I change this back?
It's a WIP for the next Beta release.
If you look back through the thread you will see that this was step one in fixing the issus caused by having the existing code run more than one instance - the change log shows that we will in future be able to run multiple instances of the same release channel, but we will need to use a customised shortcut (on current form) which gives each instance a unique name so that it can keep its data and setting separate from other instances.
That hasn't been an issue, but I could have just been very lucky. I'll not do that in the future! Thanks, Richard.
I noticed similar issues with certain scenes only Looks like your hardware configuration does not matter here because I have 8gb 1080 card
Daz reverts to CPU leaving many errors
"an illegal memory access was encountered (while de-allocating memory)"
GPU memory remains locked at the same level and there eis is no ways to reset it without reloading Daz
Would any of you guys who have time, please file a bug report about this.
I have a bug report in, and they are taking me through the usual check this & that steps. No mention they had other such reports. So if more of you reported then perhaps this issue could get better clarified.
To start, clear your log, start a new DS session, do whatevs you do to trigger the "drop to CPU" and then submit a bug report with that log.
Probably would help alot. Thanks.
You all need the new drivers from nvidia (not Nvidia STUDIO !). The latest driver is NVIDIA GeForce Graphics Drivers 441.08 WHQL for Nvidia cards and AMD Radeon Software Adrenalin 2019 for AMD cards.
See this links: https://www.techpowerup.com/download/nvidia-geforce-graphics-drivers/ AND https://www.techpowerup.com/download/amd-radeon-graphics-drivers/ And pay attention to which version of windows each specific driver for.
Why is that? Looking at changelog, don't see anything pertaining to IRAY.
Why are you posting links for drivers to non-nvidia & non-amd sites?
1. Because the driver controls the entire video system of the computer. And IRAY is its software subset.
2. Because this site explains the features of Nvidia and AMD drivers better than the Nvidia and AMD themselves.
I made an attempt to submit a bug report on that. but it is unlikely they will fix it as it is a pretty common issue with Iray.
At best, maybe we could beg Daz developers to implement some Iray plugin reloading function or else there will be no end of this CPU fallback suffering.
The log file is not that interesting It was already posted here many times over. it looks like a pure software bug because I even downclocked my GPU and memory to the min to be sure that it is not some hardware fault. and the scene is only using half of GPU memory max
2019-11-01 01:37:59.503 WARNING: ..\..\..\..\..\src\pluginsource\DzIrayRender\dzneuraymgr.cpp(304): Iray [ERROR] - IRAY:RENDER :: 1.10 IRAY rend error: CUDA device 0 (GeForce GTX 1080): an illegal memory access was encountered (while launching CUDA renderer)
2019-11-01 01:37:59.503 WARNING: ..\..\..\..\..\src\pluginsource\DzIrayRender\dzneuraymgr.cpp(304): Iray [ERROR] - IRAY:RENDER :: 1.10 IRAY rend error: CUDA device 0 (GeForce GTX 1080): Failed to launch renderer
2019-11-01 01:37:59.503 WARNING: ..\..\..\..\..\src\pluginsource\DzIrayRender\dzneuraymgr.cpp(304): Iray [ERROR] - IRAY:RENDER :: 1.2 IRAY rend error: CUDA device 0 (GeForce GTX 1080): Device failed while rendering
2019-11-01 01:37:59.503 WARNING: ..\..\..\..\..\src\pluginsource\DzIrayRender\dzneuraymgr.cpp(304): Iray [WARNING] - IRAY:RENDER :: 1.2 IRAY rend warn : All available GPUs failed.
2019-11-01 01:37:59.503 WARNING: ..\..\..\..\..\src\pluginsource\DzIrayRender\dzneuraymgr.cpp(304): Iray [WARNING] - IRAY:RENDER :: 1.2 IRAY rend warn : No devices activated. Enabling CPU fallback.
2019-11-01 01:37:59.503 WARNING: ..\..\..\..\..\src\pluginsource\DzIrayRender\dzneuraymgr.cpp(304): Iray [WARNING] - IRAY:RENDER :: 1.2 IRAY rend warn : Re-rendering iteration because of device failure
2019-11-01 01:37:59.503 WARNING: ..\..\..\..\..\src\pluginsource\DzIrayRender\dzneuraymgr.cpp(304): Iray [ERROR] - IRAY:RENDER :: 1.2 IRAY rend error: All workers failed: aborting render
2019-11-01 01:37:59.504 WARNING: ..\..\..\..\..\src\pluginsource\DzIrayRender\dzneuraymgr.cpp(304): Iray [ERROR] - IRAY:RENDER :: 1.2 IRAY rend error: CUDA device 0 (GeForce GTX 1080): an illegal memory access was encountered (while de-allocating memory)
The issues are fixed in newer versions of drivers. What is the version of your driver?
Interesting, I had a driver which was just a week older than this and it was crashing all the time just updated to 441.08 and now torturing Iray for an hour already still did not crash
It needs more testing but the newest driver may actually work properly.
I will tell more about results later
My system disk in Windows 10 is a 223 GB SSD and when I do DAZ Studio renders (I'm on 4.12+ but it's been the case with prior versions as well) the disk space gets used up as swap in the DAZ render process I suppose by Windows 10 swapping to disk. While some of it's recovered when I finish the render, and then more when I exit DAz to get most of it back I have to reboot. And their always seems to be some that 'automagically' permanently disappears unless I go through the nagging hassle of clean installing Windows 10 once again.
Just tried it. No luck for me. Still same error after upgrading to 441.08 from the 1 week earlier version (and before that was on 440.97)
This issue ONLY happens on 4.12.1.16. 4.12.0.86 continues to work fine.
Same error Info as onix.
2019-11-02 13:33:06.648 Iray [INFO] - IRAY:RENDER :: 1.2 IRAY rend info : CUDA device 0 (GeForce RTX 2080 Ti): Scene processed in 2.008s
2019-11-02 13:33:06.648 Iray [INFO] - IRAY:RENDER :: 1.2 IRAY rend info : CUDA device 0 (GeForce RTX 2080 Ti): Allocated 1.688 GiB of work space (2048k active samples in 0.000s)
2019-11-02 13:33:06.694 WARNING: ..\..\..\..\..\src\pluginsource\DzIrayRender\dzneuraymgr.cpp(304): Iray [ERROR] - IRAY:RENDER :: 1.3 IRAY rend error: CUDA device 0 (GeForce RTX 2080 Ti): an illegal memory access was encountered (while launching CUDA renderer)
2019-11-02 13:33:06.694 WARNING: ..\..\..\..\..\src\pluginsource\DzIrayRender\dzneuraymgr.cpp(304): Iray [ERROR] - IRAY:RENDER :: 1.3 IRAY rend error: CUDA device 0 (GeForce RTX 2080 Ti): Failed to launch renderer
2019-11-02 13:33:06.694 WARNING: ..\..\..\..\..\src\pluginsource\DzIrayRender\dzneuraymgr.cpp(304): Iray [ERROR] - IRAY:RENDER :: 1.2 IRAY rend error: CUDA device 0 (GeForce RTX 2080 Ti): Device failed while rendering
2019-11-02 13:33:06.694 WARNING: ..\..\..\..\..\src\pluginsource\DzIrayRender\dzneuraymgr.cpp(304): Iray [WARNING] - IRAY:RENDER :: 1.2 IRAY rend warn : All available GPUs failed.
2019-11-02 13:33:06.694 WARNING: ..\..\..\..\..\src\pluginsource\DzIrayRender\dzneuraymgr.cpp(304): Iray [WARNING] - IRAY:RENDER :: 1.2 IRAY rend warn : No devices activated. Enabling CPU fallback.
2019-11-02 13:33:06.694 WARNING: ..\..\..\..\..\src\pluginsource\DzIrayRender\dzneuraymgr.cpp(304): Iray [WARNING] - IRAY:RENDER :: 1.2 IRAY rend warn : Re-rendering iteration because of device failure
2019-11-02 13:33:06.694 WARNING: ..\..\..\..\..\src\pluginsource\DzIrayRender\dzneuraymgr.cpp(304): Iray [ERROR] - IRAY:RENDER :: 1.2 IRAY rend error: All workers failed: aborting render
2019-11-02 13:33:06.694 WARNING: ..\..\..\..\..\src\pluginsource\DzIrayRender\dzneuraymgr.cpp(304): Iray [ERROR] - IRAY:RENDER :: 1.2 IRAY rend error: CUDA device 0 (GeForce RTX 2080 Ti): an illegal memory access was encountered (while de-allocating memory)
2019-11-02 13:33:06.694 WARNING: ..\..\..\..\..\src\pluginsource\DzIrayRender\dzneuraymgr.cpp(304): Iray [ERROR] - IRAY:RENDER :: 1.2 IRAY rend error: CUDA device 0 (GeForce RTX 2080 Ti): an illegal memory access was encountered (while de-allocating memory)
Mmm. You're just stating what a driver does. The changelog should show changes. It is possible that transparency of changes isn't there. But Nividia is usually pretty good. But based on my test, still have same issue after updating to 441.08.
That site does provide a more compact overview, which can be very helpful if one is short on time. Not sure I would agree with better. Risk that they didn't include everything.
Will be interested to hear how you make out. I found with my issue that it was only certain intense changes usually triggered drop to CPU. Did end up with a a scene setup that triggers the issue right away, which is great, as was a pain to trigger before when wanting to test issue again.
Just tested that scene setup again with new 441.08.
Result: Drop to CPU during Iray viewport preview in 4.12.1.16. And 4.12.0.86 still works fine.